4. Sanity Checking Our Expected Goals Models And Final Thoughts.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sanity Checking Our Expected Goals Models and Final Thoughts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import FCPython\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following lines of code allows us to view entire dataframes (or extremely large amounts of it) when called, thus making our lives a bit easier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option(\"display.max_columns\", 100000)\n",
    "pd.set_option(\"display.max_colwidth\", 20000000)\n",
    "pd.set_option(\"display.width\", 100000000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we run the same code from the previous notebook to set up our two models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "shots_model = pd.read_csv('shots_design_matrix.csv')\n",
    "shots_model = shots_model[shots_model['Distance']<75]\n",
    "y=shots_model['Goal']\n",
    "x_rf=shots_model.drop(['Goal', 'Unnamed: 0', 'firstName', 'lastName', 'label', 'venue', 'date', \n",
    "                       'eventSec', 'matchPeriod'], axis=1)\n",
    "x_logreg=x_rf.copy()\n",
    "x_logreg = x_logreg.drop(columns=['X', 'Y', 'side_of_field', 'left_foot','prev_touch', \n",
    "                                  'prev_pass', 'C', 'in_swinging', 'out_swinging',\n",
    "                                 'rebound'])\n",
    "x_logreg['strong_distance'] = x_logreg['strong_foot']*x_logreg['Distance']\n",
    "x_logreg['strong_angle'] = x_logreg['strong_foot']*x_logreg['Angle']\n",
    "x_logreg['free_kick_angle'] = x_logreg['free_kick']*x_logreg['Angle']\n",
    "x_logreg['distance_angle'] = x_logreg['Distance']*x_logreg['Angle']\n",
    "x_logreg['duel_header'] = x_logreg['prev_duel']*x_logreg['header']\n",
    "x_logreg['distance_squared'] = x_logreg['Distance']*x_logreg['Distance']\n",
    "x_logreg['strong_distance_squared'] = x_logreg['distance_squared']*x_logreg['strong_foot']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_logreg_train, x_logreg_test, y_train, y_test = train_test_split(x_logreg, y, test_size=0.2, random_state=42)\n",
    "x_rf_train, x_rf_test, y_train, y_test = train_test_split(x_logreg, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now set up our model parameters:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "logistic_regression = LogisticRegression(random_state=42, max_iter=1000)\n",
    "\n",
    "random_forest_xg_model = RandomForestClassifier(random_state=42, \n",
    "                                                n_estimators=500, \n",
    "                                                criterion='gini', \n",
    "                                                max_depth=8,\n",
    "                                                max_features=8, \n",
    "                                                min_samples_leaf=5, \n",
    "                                                min_samples_split=250\n",
    "                                               )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this next section of the code, we will predict an xG for each datapoint in the dataset for both the Logistic Regression and Random Forest models. However, we want an out of sample prediction of the datapoint. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you think of something like Michael Caley's xG graphics, he is generating xG predictions immediately after the match. He is not re-training the model, but he is using his model to predict new datapoints, thus these maps are \"out of sample predictions.\" For our analysis, we want to do something similar, where each datapoint is predicted based on a model trained on other datapoints."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fortunately, ``sklearn`` has a nice function called ``cross_val_predict`` that does this. The way this function works is we set a number of folds (in our case 8), we then split the data into 8 parts. We trained the model on 7/8 of the data, and predict on the final eighth. We then repeat this 8 times until we have an out of sample prediction for revery datapoint in the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Note: </b> This code takes some time to run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf_hat = cross_val_predict(random_forest_xg_model, x_rf, y, cv=8, method='predict_proba')[:,-1]\n",
    "logreg_hat = cross_val_predict(logistic_regression, x_logreg, y, cv=8, method='predict_proba')[:,-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we combine the predictions into one dataframe and merge some of the relevant predictor values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "rf_preds=pd.DataFrame(rf_hat, index=x_rf.index, columns=['RF XG'])\n",
    "logreg_preds = pd.DataFrame(logreg_hat, index=x_logreg.index, columns=['Logistic XG'])\n",
    "combined=pd.concat([rf_preds, logreg_preds], axis=1)\n",
    "\n",
    "final_df=pd.merge(combined, shots_model[['firstName', 'lastName', 'label', 'venue', 'date', 'eventSec', \n",
    "                                         'matchPeriod', 'Goal', 'Distance', 'Angle', 'X', 'Y', 'C',\n",
    "                                        'prev_cross', 'prev_touch', 'prev_pass', 'prev_smart_pass', 'free_kick',\n",
    "       'prev_duel', 'header', 'counter_attack', 'strong_foot', 'rebound']], left_index=True, \n",
    "         right_index=True,how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we'll add the difference between the Random Forest and Logistic Regression models as a column, as well as the minute of the match (to make it easier to find videos)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df['difference']=final_df['RF XG']-final_df['Logistic XG']\n",
    "\n",
    "final_df['Minute']=45*(final_df['matchPeriod']=='2H')+final_df['eventSec']/60"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We'll start by finding the 60 greatest differences in both directions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RF XG</th>\n",
       "      <th>Logistic XG</th>\n",
       "      <th>firstName</th>\n",
       "      <th>lastName</th>\n",
       "      <th>label</th>\n",
       "      <th>venue</th>\n",
       "      <th>date</th>\n",
       "      <th>eventSec</th>\n",
       "      <th>matchPeriod</th>\n",
       "      <th>Goal</th>\n",
       "      <th>Distance</th>\n",
       "      <th>Angle</th>\n",
       "      <th>X</th>\n",
       "      <th>Y</th>\n",
       "      <th>C</th>\n",
       "      <th>prev_cross</th>\n",
       "      <th>prev_touch</th>\n",
       "      <th>prev_pass</th>\n",
       "      <th>prev_smart_pass</th>\n",
       "      <th>free_kick</th>\n",
       "      <th>prev_duel</th>\n",
       "      <th>header</th>\n",
       "      <th>counter_attack</th>\n",
       "      <th>strong_foot</th>\n",
       "      <th>rebound</th>\n",
       "      <th>difference</th>\n",
       "      <th>Minute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>40145</th>\n",
       "      <td>0.217620</td>\n",
       "      <td>0.493537</td>\n",
       "      <td>Kristiano Samuel</td>\n",
       "      <td>Armenteros Nunez Mendoza Jansson</td>\n",
       "      <td>Udinese - Benevento, 2 - 0</td>\n",
       "      <td>Dacia Arena</td>\n",
       "      <td>December 10, 2017 at 3:00:00 PM GMT+1</td>\n",
       "      <td>2806.689474</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>10.580170</td>\n",
       "      <td>0.662397</td>\n",
       "      <td>10</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.275917</td>\n",
       "      <td>91.778158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21661</th>\n",
       "      <td>0.256614</td>\n",
       "      <td>0.527669</td>\n",
       "      <td>Admir</td>\n",
       "      <td>Mehmedi</td>\n",
       "      <td>Bayer Leverkusen - Hamburger SV, 3 - 0</td>\n",
       "      <td>BayArena</td>\n",
       "      <td>September 24, 2017 at 6:00:00 PM GMT+2</td>\n",
       "      <td>2145.388760</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.738863</td>\n",
       "      <td>0.584237</td>\n",
       "      <td>4</td>\n",
       "      <td>60</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.271055</td>\n",
       "      <td>35.756479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42685</th>\n",
       "      <td>0.354021</td>\n",
       "      <td>0.617410</td>\n",
       "      <td>Dries</td>\n",
       "      <td>Mertens</td>\n",
       "      <td>Napoli - Roma, 2 - 4</td>\n",
       "      <td>Stadio San Paolo</td>\n",
       "      <td>March 3, 2018 at 8:45:00 PM GMT+1</td>\n",
       "      <td>2289.957000</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.409453</td>\n",
       "      <td>0.837981</td>\n",
       "      <td>6</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.263389</td>\n",
       "      <td>83.165950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7759</th>\n",
       "      <td>0.395030</td>\n",
       "      <td>0.646921</td>\n",
       "      <td>Glenn</td>\n",
       "      <td>Murray</td>\n",
       "      <td>Brighton &amp; Hove Albion - AFC Bournemouth, 2 - 2</td>\n",
       "      <td>The American Express Community Stadium</td>\n",
       "      <td>January 1, 2018 at 1:30:00 PM GMT+1</td>\n",
       "      <td>140.051082</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>7.796313</td>\n",
       "      <td>0.848711</td>\n",
       "      <td>7</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.251892</td>\n",
       "      <td>47.334185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32066</th>\n",
       "      <td>0.229875</td>\n",
       "      <td>0.464514</td>\n",
       "      <td>Denis</td>\n",
       "      <td>Cheryshev</td>\n",
       "      <td>Villarreal - Levante, 2 - 1</td>\n",
       "      <td>Estadio de la Cer\\u00e1mica</td>\n",
       "      <td>January 20, 2018 at 6:30:00 PM GMT+1</td>\n",
       "      <td>289.494393</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.584262</td>\n",
       "      <td>0.251151</td>\n",
       "      <td>1</td>\n",
       "      <td>40</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.234639</td>\n",
       "      <td>49.824907</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33588</th>\n",
       "      <td>0.292053</td>\n",
       "      <td>0.525649</td>\n",
       "      <td>Franco Dami\\u00e1n</td>\n",
       "      <td>V\\u00e1zquez</td>\n",
       "      <td>Sevilla - Valencia, 0 - 2</td>\n",
       "      <td>Estadio Ram\\u00f3n S\\u00e1nchez Pizju\\u00e1n</td>\n",
       "      <td>March 10, 2018 at 4:15:00 PM GMT+1</td>\n",
       "      <td>2475.529138</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.333443</td>\n",
       "      <td>1.045706</td>\n",
       "      <td>6</td>\n",
       "      <td>49</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.233596</td>\n",
       "      <td>41.258819</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19827</th>\n",
       "      <td>0.436054</td>\n",
       "      <td>0.667333</td>\n",
       "      <td>Florent</td>\n",
       "      <td>Mollet</td>\n",
       "      <td>Metz - Angers SCO, 1 - 2</td>\n",
       "      <td>Stade Saint-Symphorien</td>\n",
       "      <td>May 6, 2018 at 5:00:00 PM GMT+2</td>\n",
       "      <td>1144.424911</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.858541</td>\n",
       "      <td>1.072150</td>\n",
       "      <td>5</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.231278</td>\n",
       "      <td>19.073749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44248</th>\n",
       "      <td>0.449920</td>\n",
       "      <td>0.674175</td>\n",
       "      <td>Stephan</td>\n",
       "      <td>El Shaarawy</td>\n",
       "      <td>SPAL - Roma, 0 - 3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>April 21, 2018 at 3:00:00 PM GMT+2</td>\n",
       "      <td>2765.606672</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.464081</td>\n",
       "      <td>0.904323</td>\n",
       "      <td>7</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.224256</td>\n",
       "      <td>46.093445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9411</th>\n",
       "      <td>0.299464</td>\n",
       "      <td>0.521954</td>\n",
       "      <td>Juan Manuel</td>\n",
       "      <td>Mata Garc\\u00eda</td>\n",
       "      <td>Crystal Palace - Manchester United, 2 - 3</td>\n",
       "      <td>Selhurst Park</td>\n",
       "      <td>March 5, 2018 at 9:00:00 PM GMT+1</td>\n",
       "      <td>2631.467600</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.879271</td>\n",
       "      <td>0.630723</td>\n",
       "      <td>8</td>\n",
       "      <td>42</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.222490</td>\n",
       "      <td>88.857793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35018</th>\n",
       "      <td>0.249914</td>\n",
       "      <td>0.466996</td>\n",
       "      <td>Sergio</td>\n",
       "      <td>Canales Madrazo</td>\n",
       "      <td>M\\u00e1laga - Real Sociedad, 2 - 0</td>\n",
       "      <td>Estadio La Rosaleda</td>\n",
       "      <td>April 22, 2018 at 4:15:00 PM GMT+2</td>\n",
       "      <td>1467.144243</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>10.488327</td>\n",
       "      <td>0.621395</td>\n",
       "      <td>9</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.217082</td>\n",
       "      <td>24.452404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7913</th>\n",
       "      <td>0.388570</td>\n",
       "      <td>0.604483</td>\n",
       "      <td>Ayoze</td>\n",
       "      <td>P\\u00e9rez Guti\\u00e9rrez</td>\n",
       "      <td>Stoke City - Newcastle United, 0 - 1</td>\n",
       "      <td>Bet365 Stadium</td>\n",
       "      <td>January 1, 2018 at 4:00:00 PM GMT+1</td>\n",
       "      <td>1632.239267</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>8.623369</td>\n",
       "      <td>0.789654</td>\n",
       "      <td>8</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.215913</td>\n",
       "      <td>72.203988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17832</th>\n",
       "      <td>0.286962</td>\n",
       "      <td>0.498004</td>\n",
       "      <td>Fr\\u00e9d\\u00e9ric</td>\n",
       "      <td>Guilbert</td>\n",
       "      <td>Caen - Strasbourg, 2 - 0</td>\n",
       "      <td>Stade Michel d'Ornano</td>\n",
       "      <td>March 4, 2018 at 3:00:00 PM GMT+1</td>\n",
       "      <td>2456.555416</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>10.520100</td>\n",
       "      <td>0.668685</td>\n",
       "      <td>10</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.211043</td>\n",
       "      <td>85.942590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45224</th>\n",
       "      <td>0.292737</td>\n",
       "      <td>0.501718</td>\n",
       "      <td>Matteo</td>\n",
       "      <td>Politano</td>\n",
       "      <td>Sassuolo - Roma, 0 - 1</td>\n",
       "      <td>MAPEI Stadium - Citt\\u00e0 del Tricolore</td>\n",
       "      <td>May 20, 2018 at 8:45:00 PM GMT+2</td>\n",
       "      <td>1659.189000</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>10.991474</td>\n",
       "      <td>0.621073</td>\n",
       "      <td>10</td>\n",
       "      <td>45</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.208981</td>\n",
       "      <td>27.653150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25708</th>\n",
       "      <td>0.465439</td>\n",
       "      <td>0.674125</td>\n",
       "      <td>Serge</td>\n",
       "      <td>Gnabry</td>\n",
       "      <td>Augsburg - Hoffenheim, 0 - 2</td>\n",
       "      <td>WWK Arena</td>\n",
       "      <td>March 3, 2018 at 3:30:00 PM GMT+1</td>\n",
       "      <td>284.781802</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.815424</td>\n",
       "      <td>0.948906</td>\n",
       "      <td>6</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.208686</td>\n",
       "      <td>49.746363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30682</th>\n",
       "      <td>0.306128</td>\n",
       "      <td>0.509448</td>\n",
       "      <td>Jonathan</td>\n",
       "      <td>Viera Ramos</td>\n",
       "      <td>Las Palmas - Real Betis, 1 - 0</td>\n",
       "      <td>Estadio de Gran Canaria</td>\n",
       "      <td>December 3, 2017 at 8:45:00 PM GMT+1</td>\n",
       "      <td>579.078469</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.649093</td>\n",
       "      <td>0.714753</td>\n",
       "      <td>9</td>\n",
       "      <td>47</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.203319</td>\n",
       "      <td>9.651308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18189</th>\n",
       "      <td>0.310311</td>\n",
       "      <td>0.513445</td>\n",
       "      <td>Benjamin</td>\n",
       "      <td>Bourigeaud</td>\n",
       "      <td>Rennes - Saint-\\u00c9tienne, 1 - 1</td>\n",
       "      <td>Roazhon Park</td>\n",
       "      <td>March 10, 2018 at 8:00:00 PM GMT+1</td>\n",
       "      <td>1318.302810</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.993248</td>\n",
       "      <td>0.674730</td>\n",
       "      <td>9</td>\n",
       "      <td>55</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.203135</td>\n",
       "      <td>21.971713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37410</th>\n",
       "      <td>0.663301</td>\n",
       "      <td>0.862720</td>\n",
       "      <td>Remo</td>\n",
       "      <td>Freuler</td>\n",
       "      <td>Fiorentina - Atalanta, 1 - 1</td>\n",
       "      <td>Stadio Artemio Franchi</td>\n",
       "      <td>September 24, 2017 at 8:45:00 PM GMT+2</td>\n",
       "      <td>1313.957863</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>3.704727</td>\n",
       "      <td>1.556512</td>\n",
       "      <td>3</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.199418</td>\n",
       "      <td>21.899298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29042</th>\n",
       "      <td>0.400861</td>\n",
       "      <td>0.595535</td>\n",
       "      <td>Xabier</td>\n",
       "      <td>Prieto Argarate</td>\n",
       "      <td>Real Sociedad - Real Betis, 4 - 4</td>\n",
       "      <td>Estadio Municipal de Anoeta</td>\n",
       "      <td>October 1, 2017 at 12:00:00 PM GMT+2</td>\n",
       "      <td>678.933508</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.731492</td>\n",
       "      <td>1.006617</td>\n",
       "      <td>4</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.194675</td>\n",
       "      <td>56.315558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38847</th>\n",
       "      <td>0.208262</td>\n",
       "      <td>0.402705</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Petagna</td>\n",
       "      <td>Atalanta - SPAL, 1 - 1</td>\n",
       "      <td>Stadio Atleti Azzurri d'Italia</td>\n",
       "      <td>November 5, 2017 at 6:00:00 PM GMT+1</td>\n",
       "      <td>221.841242</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.472328</td>\n",
       "      <td>0.736261</td>\n",
       "      <td>9</td>\n",
       "      <td>49</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.194443</td>\n",
       "      <td>48.697354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3005</th>\n",
       "      <td>0.580801</td>\n",
       "      <td>0.772780</td>\n",
       "      <td>Miguel \\u00c1ngel</td>\n",
       "      <td>Britos Cabrera</td>\n",
       "      <td>Watford - Liverpool, 3 - 3</td>\n",
       "      <td>Vicarage Road Stadium</td>\n",
       "      <td>August 12, 2017 at 1:30:00 PM GMT+2</td>\n",
       "      <td>2879.337836</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>1.671077</td>\n",
       "      <td>2.514362</td>\n",
       "      <td>1</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.191979</td>\n",
       "      <td>92.988964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26854</th>\n",
       "      <td>0.449867</td>\n",
       "      <td>0.639718</td>\n",
       "      <td>Breel-Donald</td>\n",
       "      <td>Embolo</td>\n",
       "      <td>K\\u00f6ln - Schalke 04, 2 - 2</td>\n",
       "      <td>RheinEnergieStadion</td>\n",
       "      <td>April 22, 2018 at 6:00:00 PM GMT+2</td>\n",
       "      <td>242.109040</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.300000</td>\n",
       "      <td>1.052592</td>\n",
       "      <td>6</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.189851</td>\n",
       "      <td>4.035151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19496</th>\n",
       "      <td>0.292967</td>\n",
       "      <td>0.472506</td>\n",
       "      <td>Lebo</td>\n",
       "      <td>Mothiba</td>\n",
       "      <td>Lille - Metz, 3 - 1</td>\n",
       "      <td>Stade Pierre-Mauroy</td>\n",
       "      <td>April 28, 2018 at 8:00:00 PM GMT+2</td>\n",
       "      <td>2255.028935</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>10.488327</td>\n",
       "      <td>0.621395</td>\n",
       "      <td>9</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.179539</td>\n",
       "      <td>37.583816</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42946</th>\n",
       "      <td>0.571013</td>\n",
       "      <td>0.750413</td>\n",
       "      <td>Alejandro Dar\\u00edo</td>\n",
       "      <td>G\\u00f3mez</td>\n",
       "      <td>Hellas Verona - Atalanta, 0 - 5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>March 18, 2018 at 3:00:00 PM GMT+1</td>\n",
       "      <td>1532.916041</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>2.214723</td>\n",
       "      <td>2.405892</td>\n",
       "      <td>1</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.179399</td>\n",
       "      <td>70.548601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>870</th>\n",
       "      <td>0.161720</td>\n",
       "      <td>0.339135</td>\n",
       "      <td>Arn\\u00f3r Ingvi</td>\n",
       "      <td>Traustason</td>\n",
       "      <td>Iceland - Austria, 2 - 1</td>\n",
       "      <td>Stade de France</td>\n",
       "      <td>June 22, 2016 at 6:00:00 PM GMT+2</td>\n",
       "      <td>2911.211839</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>9.018037</td>\n",
       "      <td>0.327237</td>\n",
       "      <td>3</td>\n",
       "      <td>37</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.177415</td>\n",
       "      <td>93.520197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45096</th>\n",
       "      <td>0.427599</td>\n",
       "      <td>0.603798</td>\n",
       "      <td>C\\u00e9sar Alejandro</td>\n",
       "      <td>Falletti dos Santos</td>\n",
       "      <td>Udinese - Bologna, 1 - 0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>May 20, 2018 at 6:00:00 PM GMT+2</td>\n",
       "      <td>2890.919496</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.858541</td>\n",
       "      <td>1.072150</td>\n",
       "      <td>5</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.176200</td>\n",
       "      <td>93.181992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1743</th>\n",
       "      <td>0.412174</td>\n",
       "      <td>0.581117</td>\n",
       "      <td>Ahmed</td>\n",
       "      <td>Musa</td>\n",
       "      <td>Nigeria - Iceland, 2 - 0</td>\n",
       "      <td>Volgograd Arena</td>\n",
       "      <td>June 22, 2018 at 5:00:00 PM GMT+2</td>\n",
       "      <td>1755.956972</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.594884</td>\n",
       "      <td>0.992571</td>\n",
       "      <td>6</td>\n",
       "      <td>47</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.168943</td>\n",
       "      <td>74.265950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12951</th>\n",
       "      <td>0.347985</td>\n",
       "      <td>0.515830</td>\n",
       "      <td>Nicolas</td>\n",
       "      <td>de Roussel de Pr\\u00e9ville</td>\n",
       "      <td>Bordeaux - Guingamp, 3 - 1</td>\n",
       "      <td>Stade Matmut-Atlantique</td>\n",
       "      <td>September 23, 2017 at 8:00:00 PM GMT+2</td>\n",
       "      <td>325.288796</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.553141</td>\n",
       "      <td>0.668401</td>\n",
       "      <td>8</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.167845</td>\n",
       "      <td>5.421480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5104</th>\n",
       "      <td>0.424602</td>\n",
       "      <td>0.591979</td>\n",
       "      <td>Bamidele</td>\n",
       "      <td>Alli</td>\n",
       "      <td>Manchester United - Tottenham Hotspur, 1 - 0</td>\n",
       "      <td>Old Trafford</td>\n",
       "      <td>October 28, 2017 at 1:30:00 PM GMT+2</td>\n",
       "      <td>1907.075300</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.333443</td>\n",
       "      <td>1.045706</td>\n",
       "      <td>6</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.167377</td>\n",
       "      <td>76.784588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22776</th>\n",
       "      <td>0.597319</td>\n",
       "      <td>0.764425</td>\n",
       "      <td>Kevin</td>\n",
       "      <td>Danso</td>\n",
       "      <td>Augsburg - Bayer Leverkusen, 1 - 1</td>\n",
       "      <td>WWK Arena</td>\n",
       "      <td>November 4, 2017 at 3:30:00 PM GMT+1</td>\n",
       "      <td>195.034537</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>2.214723</td>\n",
       "      <td>2.405892</td>\n",
       "      <td>1</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.167106</td>\n",
       "      <td>48.250576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11658</th>\n",
       "      <td>0.268614</td>\n",
       "      <td>0.435705</td>\n",
       "      <td>Clinton Mua</td>\n",
       "      <td>N'Jie</td>\n",
       "      <td>Olympique Marseille - Dijon, 3 - 0</td>\n",
       "      <td>Stade V\\u00e9lodrome</td>\n",
       "      <td>August 6, 2017 at 9:00:00 PM GMT+2</td>\n",
       "      <td>325.000126</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>7.378686</td>\n",
       "      <td>0.919045</td>\n",
       "      <td>7</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.167091</td>\n",
       "      <td>50.416669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39635</th>\n",
       "      <td>0.516280</td>\n",
       "      <td>0.683370</td>\n",
       "      <td>Kevin</td>\n",
       "      <td>Lasagna</td>\n",
       "      <td>Crotone - Udinese, 0 - 3</td>\n",
       "      <td>Stadio Ezio Scida</td>\n",
       "      <td>December 4, 2017 at 7:00:00 PM GMT+1</td>\n",
       "      <td>1241.991342</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.600446</td>\n",
       "      <td>1.133413</td>\n",
       "      <td>5</td>\n",
       "      <td>47</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.167090</td>\n",
       "      <td>65.699856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29580</th>\n",
       "      <td>0.282568</td>\n",
       "      <td>0.448639</td>\n",
       "      <td>Santiago</td>\n",
       "      <td>Mina Lorenzo</td>\n",
       "      <td>Valencia - Sevilla, 4 - 0</td>\n",
       "      <td>Estadio de Mestalla</td>\n",
       "      <td>October 21, 2017 at 6:30:00 PM GMT+2</td>\n",
       "      <td>2386.993086</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>11.200893</td>\n",
       "      <td>0.601170</td>\n",
       "      <td>10</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.166071</td>\n",
       "      <td>84.783218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33389</th>\n",
       "      <td>0.287895</td>\n",
       "      <td>0.452980</td>\n",
       "      <td>Anthony Rub\\u00e9n</td>\n",
       "      <td>Lozano Col\\u00f3n</td>\n",
       "      <td>Villarreal - Girona, 0 - 2</td>\n",
       "      <td>Estadio de la Cer\\u00e1mica</td>\n",
       "      <td>March 3, 2018 at 1:00:00 PM GMT+1</td>\n",
       "      <td>2155.375912</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.215505</td>\n",
       "      <td>0.547090</td>\n",
       "      <td>2</td>\n",
       "      <td>59</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.165085</td>\n",
       "      <td>80.922932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28650</th>\n",
       "      <td>0.230061</td>\n",
       "      <td>0.394132</td>\n",
       "      <td>Florin</td>\n",
       "      <td>Andone</td>\n",
       "      <td>Deportivo La Coru\\u00f1a - Deportivo Alav\\u00e9s, 1 - 0</td>\n",
       "      <td>Estadio Municipal de Riazor</td>\n",
       "      <td>September 20, 2017 at 9:00:00 PM GMT+2</td>\n",
       "      <td>2136.282459</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>8.292316</td>\n",
       "      <td>0.506886</td>\n",
       "      <td>4</td>\n",
       "      <td>61</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.164071</td>\n",
       "      <td>80.604708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2996</th>\n",
       "      <td>0.395710</td>\n",
       "      <td>0.558862</td>\n",
       "      <td>Mohamed</td>\n",
       "      <td>Salah Ghaly</td>\n",
       "      <td>Watford - Liverpool, 3 - 3</td>\n",
       "      <td>Vicarage Road Stadium</td>\n",
       "      <td>August 12, 2017 at 1:30:00 PM GMT+2</td>\n",
       "      <td>712.521492</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.192132</td>\n",
       "      <td>0.889119</td>\n",
       "      <td>4</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.163153</td>\n",
       "      <td>56.875358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23997</th>\n",
       "      <td>0.422484</td>\n",
       "      <td>0.584955</td>\n",
       "      <td>Christian</td>\n",
       "      <td>Clemens</td>\n",
       "      <td>K\\u00f6ln - Wolfsburg, 1 - 0</td>\n",
       "      <td>RheinEnergieStadion</td>\n",
       "      <td>December 16, 2017 at 3:30:00 PM GMT+1</td>\n",
       "      <td>1278.345390</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.333443</td>\n",
       "      <td>1.045706</td>\n",
       "      <td>6</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.162471</td>\n",
       "      <td>66.305757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18132</th>\n",
       "      <td>0.181355</td>\n",
       "      <td>0.341649</td>\n",
       "      <td>Mariano</td>\n",
       "      <td>D\\u00edaz Mej\\u00eda</td>\n",
       "      <td>Olympique Lyonnais - Caen, 1 - 0</td>\n",
       "      <td>Groupama Stadium</td>\n",
       "      <td>March 11, 2018 at 5:00:00 PM GMT+1</td>\n",
       "      <td>2690.439304</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>11.550000</td>\n",
       "      <td>0.613746</td>\n",
       "      <td>11</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.160294</td>\n",
       "      <td>89.840655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9538</th>\n",
       "      <td>0.288041</td>\n",
       "      <td>0.447021</td>\n",
       "      <td>Harry</td>\n",
       "      <td>Kane</td>\n",
       "      <td>Tottenham Hotspur - Huddersfield Town, 2 - 0</td>\n",
       "      <td>Wembley Stadium</td>\n",
       "      <td>March 3, 2018 at 4:00:00 PM GMT+1</td>\n",
       "      <td>2580.676881</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>11.200893</td>\n",
       "      <td>0.601170</td>\n",
       "      <td>10</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.158980</td>\n",
       "      <td>88.011281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20441</th>\n",
       "      <td>0.578775</td>\n",
       "      <td>0.737085</td>\n",
       "      <td>Corentin</td>\n",
       "      <td>Tolisso</td>\n",
       "      <td>Bayern M\\u00fcnchen - Bayer Leverkusen, 3 - 1</td>\n",
       "      <td>Allianz Arena</td>\n",
       "      <td>August 18, 2017 at 8:30:00 PM GMT+2</td>\n",
       "      <td>1078.302328</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>2.865746</td>\n",
       "      <td>1.896003</td>\n",
       "      <td>2</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.158310</td>\n",
       "      <td>17.971705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25015</th>\n",
       "      <td>0.703848</td>\n",
       "      <td>0.861973</td>\n",
       "      <td>Michy</td>\n",
       "      <td>Batshuayi Tunga</td>\n",
       "      <td>Borussia Dortmund - Hamburger SV, 2 - 0</td>\n",
       "      <td>Signal-Iduna-Park</td>\n",
       "      <td>February 10, 2018 at 3:30:00 PM GMT+1</td>\n",
       "      <td>230.680415</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>2.198295</td>\n",
       "      <td>2.079042</td>\n",
       "      <td>2</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.158125</td>\n",
       "      <td>48.844674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12334</th>\n",
       "      <td>0.399635</td>\n",
       "      <td>0.557104</td>\n",
       "      <td>Malcom Filipe</td>\n",
       "      <td>Silva de Oliveira</td>\n",
       "      <td>Bordeaux - Troyes, 2 - 1</td>\n",
       "      <td>Stade Matmut-Atlantique</td>\n",
       "      <td>August 26, 2017 at 8:00:00 PM GMT+2</td>\n",
       "      <td>2182.296891</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.464081</td>\n",
       "      <td>0.904323</td>\n",
       "      <td>7</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.157469</td>\n",
       "      <td>81.371615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10391</th>\n",
       "      <td>0.421899</td>\n",
       "      <td>0.578096</td>\n",
       "      <td>Bernardo</td>\n",
       "      <td>Mota Veiga de Carvalho e Silva</td>\n",
       "      <td>Manchester City - Manchester United, 2 - 3</td>\n",
       "      <td>Etihad Stadium</td>\n",
       "      <td>April 7, 2018 at 6:30:00 PM GMT+2</td>\n",
       "      <td>1272.434663</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.815424</td>\n",
       "      <td>0.948906</td>\n",
       "      <td>6</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.156197</td>\n",
       "      <td>21.207244</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37887</th>\n",
       "      <td>0.370377</td>\n",
       "      <td>0.525835</td>\n",
       "      <td>Andrey</td>\n",
       "      <td>Galabinov</td>\n",
       "      <td>Cagliari - Genoa, 2 - 3</td>\n",
       "      <td>Sardegna Arena</td>\n",
       "      <td>October 15, 2017 at 3:00:00 PM GMT+2</td>\n",
       "      <td>460.443608</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>9.801148</td>\n",
       "      <td>0.696769</td>\n",
       "      <td>9</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.155458</td>\n",
       "      <td>7.674060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34840</th>\n",
       "      <td>0.609988</td>\n",
       "      <td>0.764942</td>\n",
       "      <td>Simone</td>\n",
       "      <td>Zaza</td>\n",
       "      <td>Valencia - Getafe, 1 - 2</td>\n",
       "      <td>Estadio de Mestalla</td>\n",
       "      <td>April 18, 2018 at 6:30:00 PM GMT+2</td>\n",
       "      <td>1643.208267</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>2.100000</td>\n",
       "      <td>2.099773</td>\n",
       "      <td>2</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.154953</td>\n",
       "      <td>72.386804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20143</th>\n",
       "      <td>0.328241</td>\n",
       "      <td>0.483178</td>\n",
       "      <td>Christian</td>\n",
       "      <td>Kouakou Yao</td>\n",
       "      <td>Nice - Caen, 4 - 1</td>\n",
       "      <td>Allianz Riviera</td>\n",
       "      <td>May 12, 2018 at 8:00:00 PM GMT+2</td>\n",
       "      <td>1837.855144</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.333443</td>\n",
       "      <td>1.045706</td>\n",
       "      <td>6</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.154937</td>\n",
       "      <td>75.630919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5836</th>\n",
       "      <td>0.343853</td>\n",
       "      <td>0.497420</td>\n",
       "      <td>Olivier</td>\n",
       "      <td>Giroud</td>\n",
       "      <td>Arsenal - Huddersfield Town, 5 - 0</td>\n",
       "      <td>Emirates Stadium</td>\n",
       "      <td>November 29, 2017 at 8:45:00 PM GMT+1</td>\n",
       "      <td>2496.069581</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.432729</td>\n",
       "      <td>1.025373</td>\n",
       "      <td>6</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.153567</td>\n",
       "      <td>86.601160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30103</th>\n",
       "      <td>0.273588</td>\n",
       "      <td>0.426953</td>\n",
       "      <td>Rodrigo</td>\n",
       "      <td>Moreno Machado</td>\n",
       "      <td>Valencia - Legan\\u00e9s, 3 - 0</td>\n",
       "      <td>Estadio de Mestalla</td>\n",
       "      <td>November 4, 2017 at 1:00:00 PM GMT+1</td>\n",
       "      <td>1512.953220</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.432729</td>\n",
       "      <td>1.025373</td>\n",
       "      <td>6</td>\n",
       "      <td>48</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.153365</td>\n",
       "      <td>70.215887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30013</th>\n",
       "      <td>0.412531</td>\n",
       "      <td>0.565264</td>\n",
       "      <td>Cristhian Ricardo</td>\n",
       "      <td>Stuani Curbelo</td>\n",
       "      <td>Levante - Girona, 1 - 2</td>\n",
       "      <td>Estadio Ciudad de Valencia</td>\n",
       "      <td>November 5, 2017 at 12:00:00 PM GMT+1</td>\n",
       "      <td>2280.793206</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>7.796313</td>\n",
       "      <td>0.848711</td>\n",
       "      <td>7</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.152733</td>\n",
       "      <td>83.013220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23484</th>\n",
       "      <td>0.503780</td>\n",
       "      <td>0.656154</td>\n",
       "      <td>Daniel</td>\n",
       "      <td>Didavi</td>\n",
       "      <td>Wolfsburg - Borussia M'gladbach, 3 - 0</td>\n",
       "      <td>VOLKSWAGEN ARENA</td>\n",
       "      <td>December 3, 2017 at 6:00:00 PM GMT+1</td>\n",
       "      <td>1482.606003</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.600446</td>\n",
       "      <td>1.133413</td>\n",
       "      <td>5</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.152375</td>\n",
       "      <td>24.710100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9776</th>\n",
       "      <td>0.495382</td>\n",
       "      <td>0.647725</td>\n",
       "      <td>Chris</td>\n",
       "      <td>Wood</td>\n",
       "      <td>West Ham United - Burnley, 0 - 3</td>\n",
       "      <td>London Stadium</td>\n",
       "      <td>March 10, 2018 at 4:00:00 PM GMT+1</td>\n",
       "      <td>1461.629332</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>6.174545</td>\n",
       "      <td>0.999010</td>\n",
       "      <td>5</td>\n",
       "      <td>45</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.152343</td>\n",
       "      <td>69.360489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6391</th>\n",
       "      <td>0.355992</td>\n",
       "      <td>0.508240</td>\n",
       "      <td>Bernardo</td>\n",
       "      <td>Mota Veiga de Carvalho e Silva</td>\n",
       "      <td>Manchester United - Manchester City, 1 - 2</td>\n",
       "      <td>Old Trafford</td>\n",
       "      <td>December 10, 2017 at 5:30:00 PM GMT+1</td>\n",
       "      <td>2886.312809</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.644170</td>\n",
       "      <td>0.643417</td>\n",
       "      <td>3</td>\n",
       "      <td>41</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.152248</td>\n",
       "      <td>93.105213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34234</th>\n",
       "      <td>0.307423</td>\n",
       "      <td>0.459380</td>\n",
       "      <td>Lucas</td>\n",
       "      <td>P\\u00e9rez Mart\\u00ednez</td>\n",
       "      <td>Deportivo La Coru\\u00f1a - M\\u00e1laga, 3 - 2</td>\n",
       "      <td>Estadio Municipal de Riazor</td>\n",
       "      <td>April 6, 2018 at 9:00:00 PM GMT+2</td>\n",
       "      <td>1778.116137</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>11.713454</td>\n",
       "      <td>0.599138</td>\n",
       "      <td>11</td>\n",
       "      <td>47</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151958</td>\n",
       "      <td>74.635269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22674</th>\n",
       "      <td>0.422031</td>\n",
       "      <td>0.573820</td>\n",
       "      <td>Dennis</td>\n",
       "      <td>Geiger</td>\n",
       "      <td>K\\u00f6ln - Hoffenheim, 0 - 3</td>\n",
       "      <td>RheinEnergieStadion</td>\n",
       "      <td>November 5, 2017 at 3:30:00 PM GMT+1</td>\n",
       "      <td>2051.205517</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.731492</td>\n",
       "      <td>1.006617</td>\n",
       "      <td>4</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151789</td>\n",
       "      <td>34.186759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15044</th>\n",
       "      <td>0.408934</td>\n",
       "      <td>0.560534</td>\n",
       "      <td>Marcelo Ant\\u00f4nio</td>\n",
       "      <td>Guedes Filho</td>\n",
       "      <td>Caen - Olympique Lyonnais, 1 - 2</td>\n",
       "      <td>Stade Michel d'Ornano</td>\n",
       "      <td>December 3, 2017 at 5:00:00 PM GMT+1</td>\n",
       "      <td>1988.054563</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.815424</td>\n",
       "      <td>0.948906</td>\n",
       "      <td>6</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151600</td>\n",
       "      <td>33.134243</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11553</th>\n",
       "      <td>0.413132</td>\n",
       "      <td>0.564663</td>\n",
       "      <td>Jamie</td>\n",
       "      <td>Vardy</td>\n",
       "      <td>Tottenham Hotspur - Leicester City, 5 - 4</td>\n",
       "      <td>Wembley Stadium</td>\n",
       "      <td>May 13, 2018 at 4:00:00 PM GMT+2</td>\n",
       "      <td>1667.785106</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>7.860344</td>\n",
       "      <td>0.671183</td>\n",
       "      <td>5</td>\n",
       "      <td>41</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151531</td>\n",
       "      <td>72.796418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38088</th>\n",
       "      <td>0.325629</td>\n",
       "      <td>0.477150</td>\n",
       "      <td>Remo</td>\n",
       "      <td>Freuler</td>\n",
       "      <td>Sampdoria - Atalanta, 3 - 1</td>\n",
       "      <td>Stadio Comunale Luigi Ferraris</td>\n",
       "      <td>October 15, 2017 at 3:00:00 PM GMT+2</td>\n",
       "      <td>1507.593420</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.684310</td>\n",
       "      <td>0.776686</td>\n",
       "      <td>4</td>\n",
       "      <td>58</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151522</td>\n",
       "      <td>70.126557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7451</th>\n",
       "      <td>0.258145</td>\n",
       "      <td>0.409197</td>\n",
       "      <td>Erik</td>\n",
       "      <td>Lamela</td>\n",
       "      <td>Tottenham Hotspur - Southampton, 5 - 2</td>\n",
       "      <td>Wembley Stadium</td>\n",
       "      <td>December 26, 2017 at 1:30:00 PM GMT+1</td>\n",
       "      <td>2251.052932</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>10.817116</td>\n",
       "      <td>0.638225</td>\n",
       "      <td>10</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.151052</td>\n",
       "      <td>82.517549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18283</th>\n",
       "      <td>0.322307</td>\n",
       "      <td>0.473133</td>\n",
       "      <td>Adama</td>\n",
       "      <td>Niane</td>\n",
       "      <td>Amiens SC - Troyes, 1 - 1</td>\n",
       "      <td>Stade de la Licorne</td>\n",
       "      <td>March 17, 2018 at 8:00:00 PM GMT+1</td>\n",
       "      <td>547.165515</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>6.432729</td>\n",
       "      <td>1.025373</td>\n",
       "      <td>6</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.150826</td>\n",
       "      <td>9.119425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1861</th>\n",
       "      <td>0.629682</td>\n",
       "      <td>0.780161</td>\n",
       "      <td>Neymar</td>\n",
       "      <td>da Silva Santos Junior</td>\n",
       "      <td>Brazil - Costa Rica, 2 - 0</td>\n",
       "      <td>Stadion Krestovskyi</td>\n",
       "      <td>June 22, 2018 at 2:00:00 PM GMT+2</td>\n",
       "      <td>3109.199922</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.250000</td>\n",
       "      <td>1.420148</td>\n",
       "      <td>4</td>\n",
       "      <td>51</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.150479</td>\n",
       "      <td>96.819999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20098</th>\n",
       "      <td>0.295477</td>\n",
       "      <td>0.445683</td>\n",
       "      <td>Rodrigue Casimir</td>\n",
       "      <td>Ninga</td>\n",
       "      <td>Montpellier - Troyes, 1 - 1</td>\n",
       "      <td>Stade de la Mosson</td>\n",
       "      <td>May 12, 2018 at 9:00:00 PM GMT+2</td>\n",
       "      <td>1661.498273</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.464081</td>\n",
       "      <td>0.904323</td>\n",
       "      <td>7</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.150205</td>\n",
       "      <td>27.691638</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          RF XG  Logistic XG             firstName                          lastName                                                    label                                         venue                                    date     eventSec matchPeriod  Goal   Distance     Angle   X   Y     C  prev_cross  prev_touch  prev_pass  prev_smart_pass  free_kick  prev_duel  header  counter_attack  strong_foot  rebound  difference     Minute\n",
       "40145  0.217620     0.493537      Kristiano Samuel  Armenteros Nunez Mendoza Jansson                               Udinese - Benevento, 2 - 0                                   Dacia Arena   December 10, 2017 at 3:00:00 PM GMT+1  2806.689474          2H     0  10.580170  0.662397  10  52   2.0           0           0          0                0          1          0       0               0            1        0   -0.275917  91.778158\n",
       "21661  0.256614     0.527669                 Admir                           Mehmedi                   Bayer Leverkusen - Hamburger SV, 3 - 0                                      BayArena  September 24, 2017 at 6:00:00 PM GMT+2  2145.388760          1H     0   7.738863  0.584237   4  60  10.0           0           0          0                1          0          0       0               1            1        0   -0.271055  35.756479\n",
       "42685  0.354021     0.617410                 Dries                           Mertens                                     Napoli - Roma, 2 - 4                              Stadio San Paolo       March 3, 2018 at 8:45:00 PM GMT+1  2289.957000          2H     0   7.409453  0.837981   6  56   6.0           0           0          0                0          1          0       0               0            1        0   -0.263389  83.165950\n",
       "7759   0.395030     0.646921                 Glenn                            Murray          Brighton & Hove Albion - AFC Bournemouth, 2 - 2        The American Express Community Stadium     January 1, 2018 at 1:30:00 PM GMT+1   140.051082          2H     1   7.796313  0.848711   7  54   4.0           0           0          0                1          0          0       0               1            1        0   -0.251892  47.334185\n",
       "32066  0.229875     0.464514                 Denis                         Cheryshev                              Villarreal - Levante, 2 - 1                   Estadio de la Cer\\u00e1mica    January 20, 2018 at 6:30:00 PM GMT+1   289.494393          2H     1   6.584262  0.251151   1  40  10.0           0           0          0                1          0          0       0               1            1        0   -0.234639  49.824907\n",
       "33588  0.292053     0.525649    Franco Dami\\u00e1n                      V\\u00e1zquez                                Sevilla - Valencia, 0 - 2  Estadio Ram\\u00f3n S\\u00e1nchez Pizju\\u00e1n      March 10, 2018 at 4:15:00 PM GMT+1  2475.529138          1H     0   6.333443  1.045706   6  49   1.0           0           0          0                0          0          0       1               1            0        1   -0.233596  41.258819\n",
       "19827  0.436054     0.667333               Florent                            Mollet                                 Metz - Angers SCO, 1 - 2                        Stade Saint-Symphorien         May 6, 2018 at 5:00:00 PM GMT+2  1144.424911          1H     1   5.858541  1.072150   5  46   4.0           0           0          0                0          0          1       0               1            1        0   -0.231278  19.073749\n",
       "44248  0.449920     0.674175               Stephan                       El Shaarawy                                       SPAL - Roma, 0 - 3                                           NaN      April 21, 2018 at 3:00:00 PM GMT+2  2765.606672          1H     0   7.464081  0.904323   7  52   2.0           0           0          0                1          0          0       0               1            1        0   -0.224256  46.093445\n",
       "9411   0.299464     0.521954           Juan Manuel                  Mata Garc\\u00eda                Crystal Palace - Manchester United, 2 - 3                                 Selhurst Park       March 5, 2018 at 9:00:00 PM GMT+1  2631.467600          2H     0   9.879271  0.630723   8  42   8.0           0           0          0                1          0          0       0               1            1        0   -0.222490  88.857793\n",
       "35018  0.249914     0.466996                Sergio                   Canales Madrazo                       M\\u00e1laga - Real Sociedad, 2 - 0                           Estadio La Rosaleda      April 22, 2018 at 4:15:00 PM GMT+2  1467.144243          1H     0  10.488327  0.621395   9  57   7.0           0           0          0                0          1          0       0               0            1        0   -0.217082  24.452404\n",
       "7913   0.388570     0.604483                 Ayoze         P\\u00e9rez Guti\\u00e9rrez                     Stoke City - Newcastle United, 0 - 1                                Bet365 Stadium     January 1, 2018 at 4:00:00 PM GMT+1  1632.239267          2H     1   8.623369  0.789654   8  53   3.0           0           0          0                1          0          0       0               1            1        0   -0.215913  72.203988\n",
       "17832  0.286962     0.498004    Fr\\u00e9d\\u00e9ric                          Guilbert                                 Caen - Strasbourg, 2 - 0                         Stade Michel d'Ornano       March 4, 2018 at 3:00:00 PM GMT+1  2456.555416          2H     1  10.520100  0.668685  10  51   1.0           0           0          0                1          0          0       0               1            1        0   -0.211043  85.942590\n",
       "45224  0.292737     0.501718                Matteo                          Politano                                   Sassuolo - Roma, 0 - 1      MAPEI Stadium - Citt\\u00e0 del Tricolore        May 20, 2018 at 8:45:00 PM GMT+2  1659.189000          1H     0  10.991474  0.621073  10  45   5.0           0           0          0                1          0          0       0               1            1        0   -0.208981  27.653150\n",
       "25708  0.465439     0.674125                Serge                             Gnabry                             Augsburg - Hoffenheim, 0 - 2                                     WWK Arena       March 3, 2018 at 3:30:00 PM GMT+1   284.781802          2H     1   6.815424  0.948906   6  46   4.0           0           0          0                1          0          0       0               1            1        0   -0.208686  49.746363\n",
       "30682  0.306128     0.509448              Jonathan                       Viera Ramos                           Las Palmas - Real Betis, 1 - 0                       Estadio de Gran Canaria    December 3, 2017 at 8:45:00 PM GMT+1   579.078469          1H     0   9.649093  0.714753   9  47   3.0           0           0          0                0          1          0       0               0            1        0   -0.203319   9.651308\n",
       "18189  0.310311     0.513445              Benjamin                        Bourigeaud                       Rennes - Saint-\\u00c9tienne, 1 - 1                                  Roazhon Park      March 10, 2018 at 8:00:00 PM GMT+1  1318.302810          1H     0   9.993248  0.674730   9  55   5.0           0           0          0                1          0          0       0               1            1        0   -0.203135  21.971713\n",
       "37410  0.663301     0.862720                  Remo                           Freuler                             Fiorentina - Atalanta, 1 - 1                        Stadio Artemio Franchi  September 24, 2017 at 8:45:00 PM GMT+2  1313.957863          1H     0   3.704727  1.556512   3  53   3.0           0           0          0                1          0          0       0               1            1        0   -0.199418  21.899298\n",
       "29042  0.400861     0.595535                Xabier                   Prieto Argarate                        Real Sociedad - Real Betis, 4 - 4                   Estadio Municipal de Anoeta    October 1, 2017 at 12:00:00 PM GMT+2   678.933508          2H     1   5.731492  1.006617   4  56   6.0           0           0          0                1          0          0       0               0            1        0   -0.194675  56.315558\n",
       "38847  0.208262     0.402705                Andrea                           Petagna                                   Atalanta - SPAL, 1 - 1                Stadio Atleti Azzurri d'Italia    November 5, 2017 at 6:00:00 PM GMT+1   221.841242          2H     0   9.472328  0.736261   9  49   1.0           0           0          0                1          0          0       1               1            0        0   -0.194443  48.697354\n",
       "3005   0.580801     0.772780     Miguel \\u00c1ngel                    Britos Cabrera                               Watford - Liverpool, 3 - 3                         Vicarage Road Stadium     August 12, 2017 at 1:30:00 PM GMT+2  2879.337836          2H     1   1.671077  2.514362   1  52   2.0           0           0          0                0          0          0       1               0            0        0   -0.191979  92.988964\n",
       "26854  0.449867     0.639718          Breel-Donald                            Embolo                            K\\u00f6ln - Schalke 04, 2 - 2                           RheinEnergieStadion      April 22, 2018 at 6:00:00 PM GMT+2   242.109040          1H     1   6.300000  1.052592   6  50   0.0           0           0          1                0          0          0       0               1            1        0   -0.189851   4.035151\n",
       "19496  0.292967     0.472506                  Lebo                           Mothiba                                      Lille - Metz, 3 - 1                           Stade Pierre-Mauroy      April 28, 2018 at 8:00:00 PM GMT+2  2255.028935          1H     0  10.488327  0.621395   9  57   7.0           0           0          0                1          0          0       0               1            1        0   -0.179539  37.583816\n",
       "42946  0.571013     0.750413  Alejandro Dar\\u00edo                        G\\u00f3mez                          Hellas Verona - Atalanta, 0 - 5                                           NaN      March 18, 2018 at 3:00:00 PM GMT+1  1532.916041          2H     1   2.214723  2.405892   1  53   3.0           0           0          0                0          0          0       1               0            0        0   -0.179399  70.548601\n",
       "870    0.161720     0.339135      Arn\\u00f3r Ingvi                        Traustason                                 Iceland - Austria, 2 - 1                               Stade de France       June 22, 2016 at 6:00:00 PM GMT+2  2911.211839          2H     1   9.018037  0.327237   3  37  13.0           0           0          0                1          0          0       0               1            0        0   -0.177415  93.520197\n",
       "45096  0.427599     0.603798  C\\u00e9sar Alejandro               Falletti dos Santos                                 Udinese - Bologna, 1 - 0                                           NaN        May 20, 2018 at 6:00:00 PM GMT+2  2890.919496          2H     0   5.858541  1.072150   5  54   4.0           0           0          0                1          0          0       0               0            1        0   -0.176200  93.181992\n",
       "1743   0.412174     0.581117                 Ahmed                              Musa                                 Nigeria - Iceland, 2 - 0                               Volgograd Arena       June 22, 2018 at 5:00:00 PM GMT+2  1755.956972          2H     1   6.594884  0.992571   6  47   3.0           0           0          0                0          0          1       0               1            1        0   -0.168943  74.265950\n",
       "12951  0.347985     0.515830               Nicolas       de Roussel de Pr\\u00e9ville                               Bordeaux - Guingamp, 3 - 1                       Stade Matmut-Atlantique  September 23, 2017 at 8:00:00 PM GMT+2   325.288796          1H     0   9.553141  0.668401   8  43   7.0           0           0          0                1          0          0       0               1            1        0   -0.167845   5.421480\n",
       "5104   0.424602     0.591979              Bamidele                              Alli             Manchester United - Tottenham Hotspur, 1 - 0                                  Old Trafford    October 28, 2017 at 1:30:00 PM GMT+2  1907.075300          2H     0   6.333443  1.045706   6  51   1.0           0           0          0                1          0          0       0               0            1        0   -0.167377  76.784588\n",
       "22776  0.597319     0.764425                 Kevin                             Danso                       Augsburg - Bayer Leverkusen, 1 - 1                                     WWK Arena    November 4, 2017 at 3:30:00 PM GMT+1   195.034537          2H     1   2.214723  2.405892   1  53   3.0           0           0          0                0          0          0       1               0            0        0   -0.167106  48.250576\n",
       "11658  0.268614     0.435705           Clinton Mua                             N'Jie                       Olympique Marseille - Dijon, 3 - 0                          Stade V\\u00e9lodrome      August 6, 2017 at 9:00:00 PM GMT+2   325.000126          2H     1   7.378686  0.919045   7  51   1.0           0           0          0                0          0          0       1               1            0        0   -0.167091  50.416669\n",
       "39635  0.516280     0.683370                 Kevin                           Lasagna                                 Crotone - Udinese, 0 - 3                             Stadio Ezio Scida    December 4, 2017 at 7:00:00 PM GMT+1  1241.991342          2H     1   5.600446  1.133413   5  47   3.0           0           0          1                0          0          0       0               1            0        0   -0.167090  65.699856\n",
       "29580  0.282568     0.448639              Santiago                      Mina Lorenzo                                Valencia - Sevilla, 4 - 0                           Estadio de Mestalla    October 21, 2017 at 6:30:00 PM GMT+2  2386.993086          2H     1  11.200893  0.601170  10  56   6.0           0           0          0                1          0          0       0               1            0        0   -0.166071  84.783218\n",
       "33389  0.287895     0.452980    Anthony Rub\\u00e9n                 Lozano Col\\u00f3n                               Villarreal - Girona, 0 - 2                   Estadio de la Cer\\u00e1mica       March 3, 2018 at 1:00:00 PM GMT+1  2155.375912          2H     0   6.215505  0.547090   2  59   9.0           0           0          0                0          0          1       0               1            1        0   -0.165085  80.922932\n",
       "28650  0.230061     0.394132                Florin                            Andone  Deportivo La Coru\\u00f1a - Deportivo Alav\\u00e9s, 1 - 0                   Estadio Municipal de Riazor  September 20, 2017 at 9:00:00 PM GMT+2  2136.282459          2H     0   8.292316  0.506886   4  61  11.0           0           0          1                0          0          0       0               1            1        0   -0.164071  80.604708\n",
       "2996   0.395710     0.558862              Mohamed                        Salah Ghaly                               Watford - Liverpool, 3 - 3                         Vicarage Road Stadium     August 12, 2017 at 1:30:00 PM GMT+2   712.521492          2H     1   6.192132  0.889119   4  57   7.0           0           0          0                1          0          0       0               0            1        0   -0.163153  56.875358\n",
       "23997  0.422484     0.584955             Christian                           Clemens                             K\\u00f6ln - Wolfsburg, 1 - 0                           RheinEnergieStadion   December 16, 2017 at 3:30:00 PM GMT+1  1278.345390          2H     1   6.333443  1.045706   6  51   1.0           0           0          0                1          0          0       0               0            1        0   -0.162471  66.305757\n",
       "18132  0.181355     0.341649               Mariano              D\\u00edaz Mej\\u00eda                         Olympique Lyonnais - Caen, 1 - 0                              Groupama Stadium      March 11, 2018 at 5:00:00 PM GMT+1  2690.439304          2H     0  11.550000  0.613746  11  50   0.0           0           0          0                0          0          0       0               1            0        0   -0.160294  89.840655\n",
       "9538   0.288041     0.447021                 Harry                              Kane             Tottenham Hotspur - Huddersfield Town, 2 - 0                               Wembley Stadium       March 3, 2018 at 4:00:00 PM GMT+1  2580.676881          2H     0  11.200893  0.601170  10  44   6.0           0           0          0                1          0          0       0               1            1        0   -0.158980  88.011281\n",
       "20441  0.578775     0.737085              Corentin                           Tolisso            Bayern M\\u00fcnchen - Bayer Leverkusen, 3 - 1                                 Allianz Arena     August 18, 2017 at 8:30:00 PM GMT+2  1078.302328          1H     1   2.865746  1.896003   2  53   3.0           0           0          0                1          0          0       1               0            0        0   -0.158310  17.971705\n",
       "25015  0.703848     0.861973                 Michy                   Batshuayi Tunga                  Borussia Dortmund - Hamburger SV, 2 - 0                             Signal-Iduna-Park   February 10, 2018 at 3:30:00 PM GMT+1   230.680415          2H     1   2.198295  2.079042   2  51   1.0           0           0          0                1          0          0       0               0            0        0   -0.158125  48.844674\n",
       "12334  0.399635     0.557104         Malcom Filipe                 Silva de Oliveira                                 Bordeaux - Troyes, 2 - 1                       Stade Matmut-Atlantique     August 26, 2017 at 8:00:00 PM GMT+2  2182.296891          2H     0   7.464081  0.904323   7  52   2.0           0           0          0                1          0          0       0               0            1        0   -0.157469  81.371615\n",
       "10391  0.421899     0.578096              Bernardo    Mota Veiga de Carvalho e Silva               Manchester City - Manchester United, 2 - 3                                Etihad Stadium       April 7, 2018 at 6:30:00 PM GMT+2  1272.434663          1H     0   6.815424  0.948906   6  54   4.0           0           0          0                1          0          0       0               0            1        0   -0.156197  21.207244\n",
       "37887  0.370377     0.525835                Andrey                         Galabinov                                  Cagliari - Genoa, 2 - 3                                Sardegna Arena    October 15, 2017 at 3:00:00 PM GMT+2   460.443608          1H     1   9.801148  0.696769   9  46   4.0           0           0          0                1          0          0       0               1            1        0   -0.155458   7.674060\n",
       "34840  0.609988     0.764942                Simone                              Zaza                                 Valencia - Getafe, 1 - 2                           Estadio de Mestalla      April 18, 2018 at 6:30:00 PM GMT+2  1643.208267          2H     0   2.100000  2.099773   2  50   0.0           1           0          0                0          0          0       1               1            0        0   -0.154953  72.386804\n",
       "20143  0.328241     0.483178             Christian                       Kouakou Yao                                       Nice - Caen, 4 - 1                               Allianz Riviera        May 12, 2018 at 8:00:00 PM GMT+2  1837.855144          2H     1   6.333443  1.045706   6  51   1.0           0           0          0                0          0          1       0               0            0        0   -0.154937  75.630919\n",
       "5836   0.343853     0.497420               Olivier                            Giroud                       Arsenal - Huddersfield Town, 5 - 0                              Emirates Stadium   November 29, 2017 at 8:45:00 PM GMT+1  2496.069581          2H     1   6.432729  1.025373   6  52   2.0           0           0          0                0          0          1       0               0            1        0   -0.153567  86.601160\n",
       "30103  0.273588     0.426953               Rodrigo                    Moreno Machado                           Valencia - Legan\\u00e9s, 3 - 0                           Estadio de Mestalla    November 4, 2017 at 1:00:00 PM GMT+1  1512.953220          2H     1   6.432729  1.025373   6  48   2.0           1           0          0                0          0          0       1               1            0        0   -0.153365  70.215887\n",
       "30013  0.412531     0.565264     Cristhian Ricardo                    Stuani Curbelo                                  Levante - Girona, 1 - 2                    Estadio Ciudad de Valencia   November 5, 2017 at 12:00:00 PM GMT+1  2280.793206          2H     1   7.796313  0.848711   7  54   4.0           0           0          1                0          0          0       0               1            1        0   -0.152733  83.013220\n",
       "23484  0.503780     0.656154                Daniel                            Didavi                   Wolfsburg - Borussia M'gladbach, 3 - 0                              VOLKSWAGEN ARENA    December 3, 2017 at 6:00:00 PM GMT+1  1482.606003          1H     1   5.600446  1.133413   5  53   3.0           0           0          0                1          0          0       0               0            1        0   -0.152375  24.710100\n",
       "9776   0.495382     0.647725                 Chris                              Wood                         West Ham United - Burnley, 0 - 3                                London Stadium      March 10, 2018 at 4:00:00 PM GMT+1  1461.629332          2H     1   6.174545  0.999010   5  45   5.0           0           0          1                0          0          0       0               1            1        0   -0.152343  69.360489\n",
       "6391   0.355992     0.508240              Bernardo    Mota Veiga de Carvalho e Silva               Manchester United - Manchester City, 1 - 2                                  Old Trafford   December 10, 2017 at 5:30:00 PM GMT+1  2886.312809          2H     0   6.644170  0.643417   3  41   9.0           0           0          0                0          0          1       0               1            1        0   -0.152248  93.105213\n",
       "34234  0.307423     0.459380                 Lucas          P\\u00e9rez Mart\\u00ednez            Deportivo La Coru\\u00f1a - M\\u00e1laga, 3 - 2                   Estadio Municipal de Riazor       April 6, 2018 at 9:00:00 PM GMT+2  1778.116137          2H     0  11.713454  0.599138  11  47   3.0           0           0          0                1          0          0       0               1            1        0   -0.151958  74.635269\n",
       "22674  0.422031     0.573820                Dennis                            Geiger                            K\\u00f6ln - Hoffenheim, 0 - 3                           RheinEnergieStadion    November 5, 2017 at 3:30:00 PM GMT+1  2051.205517          1H     0   5.731492  1.006617   4  56   6.0           0           0          0                1          0          0       0               0            1        0   -0.151789  34.186759\n",
       "15044  0.408934     0.560534  Marcelo Ant\\u00f4nio                      Guedes Filho                         Caen - Olympique Lyonnais, 1 - 2                         Stade Michel d'Ornano    December 3, 2017 at 5:00:00 PM GMT+1  1988.054563          1H     0   6.815424  0.948906   6  54   4.0           0           0          0                1          0          0       0               0            1        0   -0.151600  33.134243\n",
       "11553  0.413132     0.564663                 Jamie                             Vardy                Tottenham Hotspur - Leicester City, 5 - 4                               Wembley Stadium        May 13, 2018 at 4:00:00 PM GMT+2  1667.785106          2H     1   7.860344  0.671183   5  41   9.0           0           0          0                1          0          0       0               1            0        0   -0.151531  72.796418\n",
       "38088  0.325629     0.477150                  Remo                           Freuler                              Sampdoria - Atalanta, 3 - 1                Stadio Comunale Luigi Ferraris    October 15, 2017 at 3:00:00 PM GMT+2  1507.593420          2H     0   6.684310  0.776686   4  58   8.0           0           0          0                1          0          0       0               0            1        0   -0.151522  70.126557\n",
       "7451   0.258145     0.409197                  Erik                            Lamela                   Tottenham Hotspur - Southampton, 5 - 2                               Wembley Stadium   December 26, 2017 at 1:30:00 PM GMT+1  2251.052932          2H     0  10.817116  0.638225  10  46   4.0           0           0          1                0          0          0       0               1            1        0   -0.151052  82.517549\n",
       "18283  0.322307     0.473133                 Adama                             Niane                                Amiens SC - Troyes, 1 - 1                           Stade de la Licorne      March 17, 2018 at 8:00:00 PM GMT+1   547.165515          1H     0   6.432729  1.025373   6  52   2.0           0           0          0                0          0          1       0               0            0        0   -0.150826   9.119425\n",
       "1861   0.629682     0.780161                Neymar            da Silva Santos Junior                               Brazil - Costa Rica, 2 - 0                           Stadion Krestovskyi       June 22, 2018 at 2:00:00 PM GMT+2  3109.199922          2H     1   4.250000  1.420148   4  51   1.0           0           0          1                0          0          0       0               1            0        0   -0.150479  96.819999\n",
       "20098  0.295477     0.445683      Rodrigue Casimir                             Ninga                              Montpellier - Troyes, 1 - 1                            Stade de la Mosson        May 12, 2018 at 9:00:00 PM GMT+2  1661.498273          1H     0   7.464081  0.904323   7  52   2.0           0           0          0                0          0          0       1               1            0        0   -0.150205  27.691638"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.sort_values('difference').head(60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>RF XG</th>\n",
       "      <th>Logistic XG</th>\n",
       "      <th>firstName</th>\n",
       "      <th>lastName</th>\n",
       "      <th>label</th>\n",
       "      <th>venue</th>\n",
       "      <th>date</th>\n",
       "      <th>eventSec</th>\n",
       "      <th>matchPeriod</th>\n",
       "      <th>Goal</th>\n",
       "      <th>Distance</th>\n",
       "      <th>Angle</th>\n",
       "      <th>X</th>\n",
       "      <th>Y</th>\n",
       "      <th>C</th>\n",
       "      <th>prev_cross</th>\n",
       "      <th>prev_touch</th>\n",
       "      <th>prev_pass</th>\n",
       "      <th>prev_smart_pass</th>\n",
       "      <th>free_kick</th>\n",
       "      <th>prev_duel</th>\n",
       "      <th>header</th>\n",
       "      <th>counter_attack</th>\n",
       "      <th>strong_foot</th>\n",
       "      <th>rebound</th>\n",
       "      <th>difference</th>\n",
       "      <th>Minute</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>43227</th>\n",
       "      <td>0.410180</td>\n",
       "      <td>0.151999</td>\n",
       "      <td>Mauro Emanuel</td>\n",
       "      <td>Icardi Rivero</td>\n",
       "      <td>Internazionale - Hellas Verona, 3 - 0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>March 31, 2018 at 3:00:00 PM GMT+2</td>\n",
       "      <td>181.433786</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.550000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.258181</td>\n",
       "      <td>48.023896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39006</th>\n",
       "      <td>0.405738</td>\n",
       "      <td>0.218588</td>\n",
       "      <td>Joel</td>\n",
       "      <td>Chukwuma Obi</td>\n",
       "      <td>Internazionale - Torino, 1 - 1</td>\n",
       "      <td>Stadio Giuseppe Meazza</td>\n",
       "      <td>November 5, 2017 at 12:30:00 PM GMT+1</td>\n",
       "      <td>1285.061627</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.187150</td>\n",
       "      <td>66.417694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28244</th>\n",
       "      <td>0.389459</td>\n",
       "      <td>0.206158</td>\n",
       "      <td>Sergio</td>\n",
       "      <td>Enrich Ametller</td>\n",
       "      <td>Sevilla - Eibar, 3 - 0</td>\n",
       "      <td>Estadio Ram\\u00f3n S\\u00e1nchez Pizju\\u00e1n</td>\n",
       "      <td>September 9, 2017 at 6:30:00 PM GMT+2</td>\n",
       "      <td>1489.924615</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.183300</td>\n",
       "      <td>69.832077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38554</th>\n",
       "      <td>0.198272</td>\n",
       "      <td>0.020404</td>\n",
       "      <td>Aleksandar</td>\n",
       "      <td>Kolarov</td>\n",
       "      <td>Roma - Crotone, 1 - 0</td>\n",
       "      <td>Stadio Olimpico</td>\n",
       "      <td>October 25, 2017 at 8:45:00 PM GMT+2</td>\n",
       "      <td>1017.731451</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>30.159327</td>\n",
       "      <td>0.059962</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>45.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.177868</td>\n",
       "      <td>16.962191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14422</th>\n",
       "      <td>0.407958</td>\n",
       "      <td>0.242737</td>\n",
       "      <td>Hyun-Jun</td>\n",
       "      <td>Suk</td>\n",
       "      <td>Dijon - Troyes, 3 - 1</td>\n",
       "      <td>Stade Gaston-G\\u00e9rard</td>\n",
       "      <td>November 18, 2017 at 8:00:00 PM GMT+1</td>\n",
       "      <td>1659.729096</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.165221</td>\n",
       "      <td>27.662152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16699</th>\n",
       "      <td>0.407958</td>\n",
       "      <td>0.242737</td>\n",
       "      <td>Benjamin</td>\n",
       "      <td>Andr\\u00e9</td>\n",
       "      <td>Dijon - Rennes, 2 - 1</td>\n",
       "      <td>Stade Gaston-G\\u00e9rard</td>\n",
       "      <td>January 26, 2018 at 8:45:00 PM GMT+1</td>\n",
       "      <td>560.750499</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.165221</td>\n",
       "      <td>54.345842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29162</th>\n",
       "      <td>0.394737</td>\n",
       "      <td>0.242254</td>\n",
       "      <td>C\\u00e9dric</td>\n",
       "      <td>Bakambu</td>\n",
       "      <td>Villarreal - Eibar, 3 - 0</td>\n",
       "      <td>Estadio de la Cer\\u00e1mica</td>\n",
       "      <td>October 1, 2017 at 6:30:00 PM GMT+2</td>\n",
       "      <td>2090.491245</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.152483</td>\n",
       "      <td>79.841521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32597</th>\n",
       "      <td>0.394737</td>\n",
       "      <td>0.242254</td>\n",
       "      <td>Lorenzo Jes\\u00fas</td>\n",
       "      <td>Mor\\u00f3n Garc\\u00eda</td>\n",
       "      <td>Deportivo La Coru\\u00f1a - Real Betis, 0 - 1</td>\n",
       "      <td>Estadio Municipal de Riazor</td>\n",
       "      <td>February 12, 2018 at 9:00:00 PM GMT+1</td>\n",
       "      <td>515.552420</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.152483</td>\n",
       "      <td>53.592540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24959</th>\n",
       "      <td>0.438801</td>\n",
       "      <td>0.289733</td>\n",
       "      <td>Mario</td>\n",
       "      <td>G\\u00f3mez Garc\\u00eda</td>\n",
       "      <td>Wolfsburg - Stuttgart, 1 - 1</td>\n",
       "      <td>VOLKSWAGEN ARENA</td>\n",
       "      <td>February 3, 2018 at 3:30:00 PM GMT+1</td>\n",
       "      <td>548.042525</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.149068</td>\n",
       "      <td>9.134042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11716</th>\n",
       "      <td>0.397148</td>\n",
       "      <td>0.248335</td>\n",
       "      <td>K\\u00e9vin</td>\n",
       "      <td>B\\u00e9rigaud</td>\n",
       "      <td>Montpellier - Caen, 1 - 0</td>\n",
       "      <td>Stade de la Mosson</td>\n",
       "      <td>August 5, 2017 at 8:00:00 PM GMT+2</td>\n",
       "      <td>2041.293987</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.148813</td>\n",
       "      <td>34.021566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15024</th>\n",
       "      <td>0.397148</td>\n",
       "      <td>0.248335</td>\n",
       "      <td>Mario</td>\n",
       "      <td>Barwuah Balotelli</td>\n",
       "      <td>Toulouse - Nice, 1 - 2</td>\n",
       "      <td>Stadium Municipal</td>\n",
       "      <td>November 29, 2017 at 7:00:00 PM GMT+1</td>\n",
       "      <td>476.933201</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.148813</td>\n",
       "      <td>52.948887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13995</th>\n",
       "      <td>0.164758</td>\n",
       "      <td>0.019447</td>\n",
       "      <td>Anwar</td>\n",
       "      <td>El Ghazi</td>\n",
       "      <td>Lille - Olympique Marseille, 0 - 1</td>\n",
       "      <td>Stade Pierre-Mauroy</td>\n",
       "      <td>October 29, 2017 at 9:00:00 PM GMT+1</td>\n",
       "      <td>1242.861968</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>42.482849</td>\n",
       "      <td>0.132230</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.145311</td>\n",
       "      <td>20.714366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>388</th>\n",
       "      <td>0.385165</td>\n",
       "      <td>0.242668</td>\n",
       "      <td>Robert</td>\n",
       "      <td>Lewandowski</td>\n",
       "      <td>Ukraine - Poland, 0 - 1</td>\n",
       "      <td>Stade V\\u00e9lodrome</td>\n",
       "      <td>June 21, 2016 at 6:00:00 PM GMT+2</td>\n",
       "      <td>185.973598</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.142496</td>\n",
       "      <td>3.099560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5527</th>\n",
       "      <td>0.426379</td>\n",
       "      <td>0.285504</td>\n",
       "      <td>Anthony</td>\n",
       "      <td>Martial</td>\n",
       "      <td>Manchester United - Newcastle United, 4 - 1</td>\n",
       "      <td>Old Trafford</td>\n",
       "      <td>November 18, 2017 at 6:30:00 PM GMT+1</td>\n",
       "      <td>2202.593250</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.140875</td>\n",
       "      <td>36.709888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33785</th>\n",
       "      <td>0.381726</td>\n",
       "      <td>0.241950</td>\n",
       "      <td>Roger</td>\n",
       "      <td>Mart\\u00ed Salvador</td>\n",
       "      <td>Levante - Eibar, 2 - 1</td>\n",
       "      <td>Estadio Ciudad de Valencia</td>\n",
       "      <td>March 16, 2018 at 9:00:00 PM GMT+1</td>\n",
       "      <td>1490.338445</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.139776</td>\n",
       "      <td>24.838974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16228</th>\n",
       "      <td>0.436507</td>\n",
       "      <td>0.297612</td>\n",
       "      <td>Soualiho</td>\n",
       "      <td>Me\\u00eft\\u00e9</td>\n",
       "      <td>Troyes - Bordeaux, 0 - 1</td>\n",
       "      <td>Stade de l'Aube</td>\n",
       "      <td>January 13, 2018 at 8:00:00 PM GMT+1</td>\n",
       "      <td>1862.160402</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>55</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.138895</td>\n",
       "      <td>76.036007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38375</th>\n",
       "      <td>0.590021</td>\n",
       "      <td>0.451782</td>\n",
       "      <td>Marten</td>\n",
       "      <td>de Roon</td>\n",
       "      <td>Atalanta - Hellas Verona, 3 - 0</td>\n",
       "      <td>Stadio Atleti Azzurri d'Italia</td>\n",
       "      <td>October 25, 2017 at 6:30:00 PM GMT+2</td>\n",
       "      <td>2702.044567</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.138239</td>\n",
       "      <td>45.034076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38944</th>\n",
       "      <td>0.437796</td>\n",
       "      <td>0.300350</td>\n",
       "      <td>Konstantinos</td>\n",
       "      <td>Manolas</td>\n",
       "      <td>Fiorentina - Roma, 2 - 4</td>\n",
       "      <td>Stadio Artemio Franchi</td>\n",
       "      <td>November 5, 2017 at 3:00:00 PM GMT+1</td>\n",
       "      <td>246.580220</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.137446</td>\n",
       "      <td>49.109670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31887</th>\n",
       "      <td>0.529129</td>\n",
       "      <td>0.392009</td>\n",
       "      <td>Adri\\u00e1n</td>\n",
       "      <td>L\\u00f3pez \\u00c1lvarez</td>\n",
       "      <td>Real Madrid - Deportivo La Coru\\u00f1a, 7 - 1</td>\n",
       "      <td>Estadio Santiago Bernab\\u00e9u</td>\n",
       "      <td>January 21, 2018 at 4:15:00 PM GMT+1</td>\n",
       "      <td>1356.229466</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.137120</td>\n",
       "      <td>22.603824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39603</th>\n",
       "      <td>0.436872</td>\n",
       "      <td>0.300350</td>\n",
       "      <td>Leonardo</td>\n",
       "      <td>Pavoletti</td>\n",
       "      <td>Bologna - Cagliari, 1 - 1</td>\n",
       "      <td>Stadio Renato Dall'Ara</td>\n",
       "      <td>December 3, 2017 at 3:00:00 PM GMT+1</td>\n",
       "      <td>476.708858</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.136522</td>\n",
       "      <td>7.945148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38569</th>\n",
       "      <td>0.436872</td>\n",
       "      <td>0.300350</td>\n",
       "      <td>Maxime</td>\n",
       "      <td>Gonalons</td>\n",
       "      <td>Roma - Crotone, 1 - 0</td>\n",
       "      <td>Stadio Olimpico</td>\n",
       "      <td>October 25, 2017 at 8:45:00 PM GMT+2</td>\n",
       "      <td>2446.202217</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.136522</td>\n",
       "      <td>85.770037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39631</th>\n",
       "      <td>0.436872</td>\n",
       "      <td>0.300350</td>\n",
       "      <td>Gabriele</td>\n",
       "      <td>Angella</td>\n",
       "      <td>Crotone - Udinese, 0 - 3</td>\n",
       "      <td>Stadio Ezio Scida</td>\n",
       "      <td>December 4, 2017 at 7:00:00 PM GMT+1</td>\n",
       "      <td>798.307165</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.136522</td>\n",
       "      <td>58.305119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33959</th>\n",
       "      <td>0.436872</td>\n",
       "      <td>0.300350</td>\n",
       "      <td>Cristhian Ricardo</td>\n",
       "      <td>Stuani Curbelo</td>\n",
       "      <td>Real Madrid - Girona, 6 - 3</td>\n",
       "      <td>Estadio Santiago Bernab\\u00e9u</td>\n",
       "      <td>March 18, 2018 at 8:45:00 PM GMT+1</td>\n",
       "      <td>1313.810278</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.136522</td>\n",
       "      <td>66.896838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42068</th>\n",
       "      <td>0.522738</td>\n",
       "      <td>0.387040</td>\n",
       "      <td>Giovanni</td>\n",
       "      <td>Crociata</td>\n",
       "      <td>Benevento - Crotone, 3 - 2</td>\n",
       "      <td>Stadio Ciro Vigorito</td>\n",
       "      <td>February 18, 2018 at 3:00:00 PM GMT+1</td>\n",
       "      <td>639.416196</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.135698</td>\n",
       "      <td>10.656937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5531</th>\n",
       "      <td>0.338781</td>\n",
       "      <td>0.204223</td>\n",
       "      <td>Chris</td>\n",
       "      <td>Smalling</td>\n",
       "      <td>Manchester United - Newcastle United, 4 - 1</td>\n",
       "      <td>Old Trafford</td>\n",
       "      <td>November 18, 2017 at 6:30:00 PM GMT+1</td>\n",
       "      <td>2754.483406</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.134557</td>\n",
       "      <td>45.908057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8999</th>\n",
       "      <td>0.560988</td>\n",
       "      <td>0.426878</td>\n",
       "      <td>Wesley</td>\n",
       "      <td>Hoedt</td>\n",
       "      <td>Southampton - Liverpool, 0 - 2</td>\n",
       "      <td>St. Mary's Stadium</td>\n",
       "      <td>February 11, 2018 at 5:30:00 PM GMT+1</td>\n",
       "      <td>809.491575</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.134111</td>\n",
       "      <td>58.491526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21045</th>\n",
       "      <td>0.153877</td>\n",
       "      <td>0.021078</td>\n",
       "      <td>Levin Mete</td>\n",
       "      <td>\\u00d6ztunali</td>\n",
       "      <td>Mainz 05 - Bayer Leverkusen, 3 - 1</td>\n",
       "      <td>OPEL ARENA</td>\n",
       "      <td>September 9, 2017 at 3:30:00 PM GMT+2</td>\n",
       "      <td>503.758742</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>38.608030</td>\n",
       "      <td>0.134465</td>\n",
       "      <td>26</td>\n",
       "      <td>8</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.132799</td>\n",
       "      <td>53.395979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25569</th>\n",
       "      <td>0.450637</td>\n",
       "      <td>0.317861</td>\n",
       "      <td>Erik</td>\n",
       "      <td>Thommy</td>\n",
       "      <td>Stuttgart - Eintracht Frankfurt, 1 - 0</td>\n",
       "      <td>Mercedes-Benz-Arena</td>\n",
       "      <td>February 24, 2018 at 3:30:00 PM GMT+1</td>\n",
       "      <td>731.374619</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>1.300000</td>\n",
       "      <td>-0.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>52</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.132776</td>\n",
       "      <td>12.189577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37091</th>\n",
       "      <td>0.583800</td>\n",
       "      <td>0.451782</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Belotti</td>\n",
       "      <td>Torino - Sampdoria, 2 - 2</td>\n",
       "      <td>Stadio Olimpico Grande Torino</td>\n",
       "      <td>September 17, 2017 at 3:00:00 PM GMT+2</td>\n",
       "      <td>1870.262658</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.132018</td>\n",
       "      <td>76.171044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40441</th>\n",
       "      <td>0.142784</td>\n",
       "      <td>0.010784</td>\n",
       "      <td>Simone</td>\n",
       "      <td>Verdi</td>\n",
       "      <td>Chievo - Bologna, 2 - 3</td>\n",
       "      <td>Stadio Marc'Antonio Bentegodi</td>\n",
       "      <td>December 22, 2017 at 6:00:00 PM GMT+1</td>\n",
       "      <td>2763.753113</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>27.942486</td>\n",
       "      <td>0.079956</td>\n",
       "      <td>8</td>\n",
       "      <td>91</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.132000</td>\n",
       "      <td>46.062552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39367</th>\n",
       "      <td>0.299620</td>\n",
       "      <td>0.167642</td>\n",
       "      <td>Mat\\u00edas Agust\\u00edn</td>\n",
       "      <td>Silvestre</td>\n",
       "      <td>Bologna - Sampdoria, 3 - 0</td>\n",
       "      <td>Stadio Renato Dall'Ara</td>\n",
       "      <td>November 25, 2017 at 3:00:00 PM GMT+1</td>\n",
       "      <td>1384.404262</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.304951</td>\n",
       "      <td>0.480459</td>\n",
       "      <td>1</td>\n",
       "      <td>42</td>\n",
       "      <td>8.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.131977</td>\n",
       "      <td>68.073404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6273</th>\n",
       "      <td>0.402479</td>\n",
       "      <td>0.270788</td>\n",
       "      <td>Jeff</td>\n",
       "      <td>Hendrick</td>\n",
       "      <td>Burnley - Watford, 1 - 0</td>\n",
       "      <td>Turf Moor</td>\n",
       "      <td>December 9, 2017 at 4:00:00 PM GMT+1</td>\n",
       "      <td>1055.812711</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>7.201562</td>\n",
       "      <td>0.674273</td>\n",
       "      <td>4</td>\n",
       "      <td>41</td>\n",
       "      <td>9.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.131691</td>\n",
       "      <td>17.596879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9535</th>\n",
       "      <td>0.378322</td>\n",
       "      <td>0.248322</td>\n",
       "      <td>Harry</td>\n",
       "      <td>Kane</td>\n",
       "      <td>Tottenham Hotspur - Huddersfield Town, 2 - 0</td>\n",
       "      <td>Wembley Stadium</td>\n",
       "      <td>March 3, 2018 at 4:00:00 PM GMT+1</td>\n",
       "      <td>2011.749290</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>9.450000</td>\n",
       "      <td>0.739024</td>\n",
       "      <td>9</td>\n",
       "      <td>50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.130000</td>\n",
       "      <td>78.529155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5721</th>\n",
       "      <td>0.433509</td>\n",
       "      <td>0.303562</td>\n",
       "      <td>Charlie</td>\n",
       "      <td>Austin</td>\n",
       "      <td>Southampton - Everton, 4 - 1</td>\n",
       "      <td>St. Mary's Stadium</td>\n",
       "      <td>November 26, 2017 at 2:30:00 PM GMT+1</td>\n",
       "      <td>411.968931</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>45</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129946</td>\n",
       "      <td>51.866149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4457</th>\n",
       "      <td>0.433509</td>\n",
       "      <td>0.303562</td>\n",
       "      <td>Ahmed</td>\n",
       "      <td>Hegazy</td>\n",
       "      <td>West Bromwich Albion - Watford, 2 - 2</td>\n",
       "      <td>The Hawthorns</td>\n",
       "      <td>September 30, 2017 at 4:00:00 PM GMT+2</td>\n",
       "      <td>1892.973924</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>45</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129946</td>\n",
       "      <td>31.549565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43798</th>\n",
       "      <td>0.410843</td>\n",
       "      <td>0.280911</td>\n",
       "      <td>Fabio</td>\n",
       "      <td>Quagliarella</td>\n",
       "      <td>Juventus - Sampdoria, 3 - 0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>April 15, 2018 at 6:00:00 PM GMT+2</td>\n",
       "      <td>1256.855706</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.429447</td>\n",
       "      <td>1.186057</td>\n",
       "      <td>2</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129932</td>\n",
       "      <td>20.947595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16512</th>\n",
       "      <td>0.547058</td>\n",
       "      <td>0.417874</td>\n",
       "      <td>Rafael</td>\n",
       "      <td>Pereira da Silva</td>\n",
       "      <td>Olympique Lyonnais - PSG, 2 - 1</td>\n",
       "      <td>Groupama Stadium</td>\n",
       "      <td>January 21, 2018 at 9:00:00 PM GMT+1</td>\n",
       "      <td>1963.196299</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>3.342155</td>\n",
       "      <td>1.714580</td>\n",
       "      <td>2</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129185</td>\n",
       "      <td>77.719938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16907</th>\n",
       "      <td>0.547058</td>\n",
       "      <td>0.417874</td>\n",
       "      <td>Damien</td>\n",
       "      <td>Da Silva</td>\n",
       "      <td>Caen - Nantes, 3 - 2</td>\n",
       "      <td>Stade Michel d'Ornano</td>\n",
       "      <td>February 4, 2018 at 5:00:00 PM GMT+1</td>\n",
       "      <td>2060.083861</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>3.342155</td>\n",
       "      <td>1.714580</td>\n",
       "      <td>2</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129185</td>\n",
       "      <td>79.334731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44225</th>\n",
       "      <td>0.413372</td>\n",
       "      <td>0.284200</td>\n",
       "      <td>Benedikt</td>\n",
       "      <td>H\\u00f6wedes</td>\n",
       "      <td>Juventus - Napoli, 0 - 1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>April 22, 2018 at 8:45:00 PM GMT+2</td>\n",
       "      <td>769.871978</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129172</td>\n",
       "      <td>12.831200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43250</th>\n",
       "      <td>0.413372</td>\n",
       "      <td>0.284200</td>\n",
       "      <td>Leonardo</td>\n",
       "      <td>Bonucci</td>\n",
       "      <td>Juventus - Milan, 3 - 1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>March 31, 2018 at 8:45:00 PM GMT+2</td>\n",
       "      <td>1656.863687</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.129172</td>\n",
       "      <td>27.614395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38814</th>\n",
       "      <td>0.408213</td>\n",
       "      <td>0.279629</td>\n",
       "      <td>Andrea</td>\n",
       "      <td>Belotti</td>\n",
       "      <td>Torino - Cagliari, 2 - 1</td>\n",
       "      <td>Stadio Olimpico Grande Torino</td>\n",
       "      <td>October 29, 2017 at 8:45:00 PM GMT+1</td>\n",
       "      <td>1147.947197</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.128584</td>\n",
       "      <td>64.132453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11598</th>\n",
       "      <td>0.148144</td>\n",
       "      <td>0.019777</td>\n",
       "      <td>Thomas</td>\n",
       "      <td>Mangani</td>\n",
       "      <td>Angers SCO - Bordeaux, 2 - 2</td>\n",
       "      <td>Stade Raymond Kopa</td>\n",
       "      <td>August 6, 2017 at 5:00:00 PM GMT+2</td>\n",
       "      <td>1177.624908</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>29.920770</td>\n",
       "      <td>0.052247</td>\n",
       "      <td>6</td>\n",
       "      <td>95</td>\n",
       "      <td>45.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.128367</td>\n",
       "      <td>64.627082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31558</th>\n",
       "      <td>0.146985</td>\n",
       "      <td>0.019787</td>\n",
       "      <td>Johan Andr\\u00e9s</td>\n",
       "      <td>Mojica Palacio</td>\n",
       "      <td>Valencia - Girona, 2 - 1</td>\n",
       "      <td>Estadio de Mestalla</td>\n",
       "      <td>January 6, 2018 at 4:15:00 PM GMT+1</td>\n",
       "      <td>1103.834403</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>28.900260</td>\n",
       "      <td>0.065373</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.127199</td>\n",
       "      <td>63.397240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10480</th>\n",
       "      <td>0.420941</td>\n",
       "      <td>0.294824</td>\n",
       "      <td>Kevin</td>\n",
       "      <td>Long</td>\n",
       "      <td>Burnley - Leicester City, 2 - 1</td>\n",
       "      <td>Turf Moor</td>\n",
       "      <td>April 14, 2018 at 4:00:00 PM GMT+2</td>\n",
       "      <td>490.414371</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>46</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.126118</td>\n",
       "      <td>8.173573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9979</th>\n",
       "      <td>0.420584</td>\n",
       "      <td>0.294824</td>\n",
       "      <td>Dwight</td>\n",
       "      <td>Gayle</td>\n",
       "      <td>Tottenham Hotspur - Newcastle United, 1 - 0</td>\n",
       "      <td>Wembley Stadium</td>\n",
       "      <td>May 9, 2018 at 9:00:00 PM GMT+2</td>\n",
       "      <td>1691.893432</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.939636</td>\n",
       "      <td>1.227069</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.125760</td>\n",
       "      <td>28.198224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16789</th>\n",
       "      <td>0.183819</td>\n",
       "      <td>0.059416</td>\n",
       "      <td>Mario</td>\n",
       "      <td>Barwuah Balotelli</td>\n",
       "      <td>Metz - Nice, 2 - 1</td>\n",
       "      <td>Stade Saint-Symphorien</td>\n",
       "      <td>January 27, 2018 at 8:00:00 PM GMT+1</td>\n",
       "      <td>2797.378458</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>65.353959</td>\n",
       "      <td>0.097173</td>\n",
       "      <td>54</td>\n",
       "      <td>0</td>\n",
       "      <td>50.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.124403</td>\n",
       "      <td>91.622974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18493</th>\n",
       "      <td>0.384471</td>\n",
       "      <td>0.261086</td>\n",
       "      <td>Max-Alain</td>\n",
       "      <td>Gradel</td>\n",
       "      <td>Toulouse - Strasbourg, 2 - 2</td>\n",
       "      <td>Stadium Municipal</td>\n",
       "      <td>March 17, 2018 at 8:00:00 PM GMT+1</td>\n",
       "      <td>1449.737972</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.123385</td>\n",
       "      <td>69.162300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39284</th>\n",
       "      <td>0.330167</td>\n",
       "      <td>0.206798</td>\n",
       "      <td>Perparim</td>\n",
       "      <td>Hetemaj</td>\n",
       "      <td>Torino - Chievo, 1 - 1</td>\n",
       "      <td>Stadio Olimpico Grande Torino</td>\n",
       "      <td>November 19, 2017 at 3:00:00 PM GMT+1</td>\n",
       "      <td>801.579435</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.123370</td>\n",
       "      <td>13.359657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25061</th>\n",
       "      <td>0.441783</td>\n",
       "      <td>0.318973</td>\n",
       "      <td>\\u00c1d\\u00e1m</td>\n",
       "      <td>Szalai</td>\n",
       "      <td>Hoffenheim - Mainz 05, 4 - 2</td>\n",
       "      <td>WIRSOL Rhein-Neckar-Arena</td>\n",
       "      <td>February 10, 2018 at 3:30:00 PM GMT+1</td>\n",
       "      <td>2524.988635</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.630605</td>\n",
       "      <td>1.314800</td>\n",
       "      <td>4</td>\n",
       "      <td>53</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122809</td>\n",
       "      <td>42.083144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20871</th>\n",
       "      <td>0.438264</td>\n",
       "      <td>0.315581</td>\n",
       "      <td>Yoshinori</td>\n",
       "      <td>Muto</td>\n",
       "      <td>Stuttgart - Mainz 05, 1 - 0</td>\n",
       "      <td>Mercedes-Benz-Arena</td>\n",
       "      <td>August 26, 2017 at 3:30:00 PM GMT+2</td>\n",
       "      <td>2822.661547</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>55</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122684</td>\n",
       "      <td>92.044359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21163</th>\n",
       "      <td>0.438264</td>\n",
       "      <td>0.315581</td>\n",
       "      <td>Timo</td>\n",
       "      <td>Werner</td>\n",
       "      <td>RB Leipzig - Borussia M'gladbach, 2 - 2</td>\n",
       "      <td>Red Bull Arena</td>\n",
       "      <td>September 16, 2017 at 6:30:00 PM GMT+2</td>\n",
       "      <td>2098.759279</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>55</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122684</td>\n",
       "      <td>79.979321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21907</th>\n",
       "      <td>0.402299</td>\n",
       "      <td>0.279687</td>\n",
       "      <td>Marc</td>\n",
       "      <td>Bartra Aregall</td>\n",
       "      <td>Augsburg - Borussia Dortmund, 1 - 2</td>\n",
       "      <td>WWK Arena</td>\n",
       "      <td>September 30, 2017 at 3:30:00 PM GMT+2</td>\n",
       "      <td>1697.117480</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122613</td>\n",
       "      <td>28.285291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25636</th>\n",
       "      <td>0.563004</td>\n",
       "      <td>0.440416</td>\n",
       "      <td>Thomas</td>\n",
       "      <td>M\\u00fcller</td>\n",
       "      <td>Freiburg - Bayern M\\u00fcnchen, 0 - 4</td>\n",
       "      <td>Schwarzwald-Stadion</td>\n",
       "      <td>March 4, 2018 at 6:00:00 PM GMT+1</td>\n",
       "      <td>1399.201051</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>5.013232</td>\n",
       "      <td>1.099962</td>\n",
       "      <td>3</td>\n",
       "      <td>56</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122588</td>\n",
       "      <td>68.320018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40071</th>\n",
       "      <td>0.424671</td>\n",
       "      <td>0.302244</td>\n",
       "      <td>Edoardo</td>\n",
       "      <td>Goldaniga</td>\n",
       "      <td>Sassuolo - Crotone, 2 - 1</td>\n",
       "      <td>MAPEI Stadium - Citt\\u00e0 del Tricolore</td>\n",
       "      <td>December 10, 2017 at 6:00:00 PM GMT+1</td>\n",
       "      <td>200.830730</td>\n",
       "      <td>2H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.526036</td>\n",
       "      <td>1.272510</td>\n",
       "      <td>3</td>\n",
       "      <td>45</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.122427</td>\n",
       "      <td>48.347179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36461</th>\n",
       "      <td>0.611426</td>\n",
       "      <td>0.489785</td>\n",
       "      <td>Patrick</td>\n",
       "      <td>Cutrone</td>\n",
       "      <td>Milan - Cagliari, 2 - 1</td>\n",
       "      <td>Stadio Giuseppe Meazza</td>\n",
       "      <td>August 27, 2017 at 8:45:00 PM GMT+2</td>\n",
       "      <td>585.115060</td>\n",
       "      <td>1H</td>\n",
       "      <td>1</td>\n",
       "      <td>4.429447</td>\n",
       "      <td>1.186057</td>\n",
       "      <td>2</td>\n",
       "      <td>44</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.121642</td>\n",
       "      <td>9.751918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43249</th>\n",
       "      <td>0.328164</td>\n",
       "      <td>0.206798</td>\n",
       "      <td>Andr\\u00e9 Miguel</td>\n",
       "      <td>Valente Silva</td>\n",
       "      <td>Juventus - Milan, 3 - 1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>March 31, 2018 at 8:45:00 PM GMT+2</td>\n",
       "      <td>908.084895</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.011237</td>\n",
       "      <td>0.919518</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.121366</td>\n",
       "      <td>15.134748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37604</th>\n",
       "      <td>0.331155</td>\n",
       "      <td>0.210887</td>\n",
       "      <td>Jasmin</td>\n",
       "      <td>Kurti\\u010d</td>\n",
       "      <td>Atalanta - Juventus, 2 - 2</td>\n",
       "      <td>Stadio Atleti Azzurri d'Italia</td>\n",
       "      <td>October 1, 2017 at 8:45:00 PM GMT+2</td>\n",
       "      <td>1995.816106</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.533986</td>\n",
       "      <td>0.929077</td>\n",
       "      <td>3</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.120268</td>\n",
       "      <td>33.263602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34967</th>\n",
       "      <td>0.331155</td>\n",
       "      <td>0.210887</td>\n",
       "      <td>Cristhian Ricardo</td>\n",
       "      <td>Stuani Curbelo</td>\n",
       "      <td>Girona - Espanyol, 0 - 2</td>\n",
       "      <td>Estadi Municipal de Montilivi</td>\n",
       "      <td>April 22, 2018 at 12:00:00 PM GMT+2</td>\n",
       "      <td>1364.922783</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.533986</td>\n",
       "      <td>0.929077</td>\n",
       "      <td>3</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.120268</td>\n",
       "      <td>67.748713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39416</th>\n",
       "      <td>0.331155</td>\n",
       "      <td>0.210887</td>\n",
       "      <td>Alberto</td>\n",
       "      <td>Grassi</td>\n",
       "      <td>Chievo - SPAL, 2 - 1</td>\n",
       "      <td>Stadio Marc'Antonio Bentegodi</td>\n",
       "      <td>November 25, 2017 at 6:00:00 PM GMT+1</td>\n",
       "      <td>714.819160</td>\n",
       "      <td>2H</td>\n",
       "      <td>0</td>\n",
       "      <td>5.533986</td>\n",
       "      <td>0.929077</td>\n",
       "      <td>3</td>\n",
       "      <td>43</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.120268</td>\n",
       "      <td>56.913653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14133</th>\n",
       "      <td>0.127432</td>\n",
       "      <td>0.008118</td>\n",
       "      <td>Morgan</td>\n",
       "      <td>Sanson</td>\n",
       "      <td>Olympique Marseille - Caen, 5 - 0</td>\n",
       "      <td>Stade V\\u00e9lodrome</td>\n",
       "      <td>November 5, 2017 at 5:00:00 PM GMT+1</td>\n",
       "      <td>1917.190108</td>\n",
       "      <td>1H</td>\n",
       "      <td>0</td>\n",
       "      <td>33.220852</td>\n",
       "      <td>0.105357</td>\n",
       "      <td>15</td>\n",
       "      <td>5</td>\n",
       "      <td>45.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.119314</td>\n",
       "      <td>31.953168</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          RF XG  Logistic XG                 firstName                 lastName                                          label                                         venue                                    date     eventSec matchPeriod  Goal   Distance     Angle   X   Y     C  prev_cross  prev_touch  prev_pass  prev_smart_pass  free_kick  prev_duel  header  counter_attack  strong_foot  rebound  difference     Minute\n",
       "43227  0.410180     0.151999             Mauro Emanuel           Icardi Rivero           Internazionale - Hellas Verona, 3 - 0                                           NaN      March 31, 2018 at 3:00:00 PM GMT+2   181.433786          2H     1   4.550000  0.000000   0  57   7.0           1           0          0                0          0          0       0               0            1        0    0.258181  48.023896\n",
       "39006  0.405738     0.218588                      Joel             Chukwuma Obi                 Internazionale - Torino, 1 - 1                        Stadio Giuseppe Meazza   November 5, 2017 at 12:30:00 PM GMT+1  1285.061627          2H     0   5.011237  0.919518   2  57   7.0           0           0          0                0          0          1       1               0            0        0    0.187150  66.417694\n",
       "28244  0.389459     0.206158                    Sergio          Enrich Ametller                         Sevilla - Eibar, 3 - 0  Estadio Ram\\u00f3n S\\u00e1nchez Pizju\\u00e1n   September 9, 2017 at 6:30:00 PM GMT+2  1489.924615          2H     0   5.011237  0.919518   2  43   7.0           0           0          0                0          0          1       1               0            0        0    0.183300  69.832077\n",
       "38554  0.198272     0.020404                Aleksandar                  Kolarov                          Roma - Crotone, 1 - 0                               Stadio Olimpico    October 25, 2017 at 8:45:00 PM GMT+2  1017.731451          1H     0  30.159327  0.059962   7   5  45.0           0           0          0                0          1          0       0               0            1        0    0.177868  16.962191\n",
       "14422  0.407958     0.242737                  Hyun-Jun                      Suk                          Dijon - Troyes, 3 - 1                      Stade Gaston-G\\u00e9rard   November 18, 2017 at 8:00:00 PM GMT+1  1659.729096          1H     0   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.165221  27.662152\n",
       "16699  0.407958     0.242737                  Benjamin               Andr\\u00e9                          Dijon - Rennes, 2 - 1                      Stade Gaston-G\\u00e9rard    January 26, 2018 at 8:45:00 PM GMT+1   560.750499          2H     0   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.165221  54.345842\n",
       "29162  0.394737     0.242254               C\\u00e9dric                  Bakambu                      Villarreal - Eibar, 3 - 0                   Estadio de la Cer\\u00e1mica     October 1, 2017 at 6:30:00 PM GMT+2  2090.491245          2H     0   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.152483  79.841521\n",
       "32597  0.394737     0.242254        Lorenzo Jes\\u00fas   Mor\\u00f3n Garc\\u00eda   Deportivo La Coru\\u00f1a - Real Betis, 0 - 1                   Estadio Municipal de Riazor   February 12, 2018 at 9:00:00 PM GMT+1   515.552420          2H     1   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.152483  53.592540\n",
       "24959  0.438801     0.289733                     Mario   G\\u00f3mez Garc\\u00eda                   Wolfsburg - Stuttgart, 1 - 1                              VOLKSWAGEN ARENA    February 3, 2018 at 3:30:00 PM GMT+1   548.042525          1H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.149068   9.134042\n",
       "11716  0.397148     0.248335                K\\u00e9vin            B\\u00e9rigaud                      Montpellier - Caen, 1 - 0                            Stade de la Mosson      August 5, 2017 at 8:00:00 PM GMT+2  2041.293987          1H     0   5.013232  1.099962   3  44   6.0           0           0          0                0          0          1       1               0            0        0    0.148813  34.021566\n",
       "15024  0.397148     0.248335                    Mario         Barwuah Balotelli                         Toulouse - Nice, 1 - 2                             Stadium Municipal   November 29, 2017 at 7:00:00 PM GMT+1   476.933201          2H     0   5.013232  1.099962   3  44   6.0           0           0          0                0          0          1       1               0            0        0    0.148813  52.948887\n",
       "13995  0.164758     0.019447                     Anwar                 El Ghazi             Lille - Olympique Marseille, 0 - 1                           Stade Pierre-Mauroy    October 29, 2017 at 9:00:00 PM GMT+1  1242.861968          1H     0  42.482849  0.132230  31   8  42.0           0           0          0                0          1          0       0               0            1        0    0.145311  20.714366\n",
       "388    0.385165     0.242668                    Robert              Lewandowski                        Ukraine - Poland, 0 - 1                          Stade V\\u00e9lodrome       June 21, 2016 at 6:00:00 PM GMT+2   185.973598          1H     0   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.142496   3.099560\n",
       "5527   0.426379     0.285504                   Anthony                  Martial    Manchester United - Newcastle United, 4 - 1                                  Old Trafford   November 18, 2017 at 6:30:00 PM GMT+1  2202.593250          1H     1   4.939636  1.227069   4  46   4.0           0           0          0                0          0          1       1               0            0        0    0.140875  36.709888\n",
       "33785  0.381726     0.241950                     Roger      Mart\\u00ed Salvador                         Levante - Eibar, 2 - 1                    Estadio Ciudad de Valencia      March 16, 2018 at 9:00:00 PM GMT+1  1490.338445          1H     1   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.139776  24.838974\n",
       "16228  0.436507     0.297612                  Soualiho          Me\\u00eft\\u00e9                       Troyes - Bordeaux, 0 - 1                               Stade de l'Aube    January 13, 2018 at 8:00:00 PM GMT+1  1862.160402          2H     0   4.526036  1.272510   3  55   5.0           0           0          0                0          0          1       1               0            0        0    0.138895  76.036007\n",
       "38375  0.590021     0.451782                    Marten                  de Roon                Atalanta - Hellas Verona, 3 - 0                Stadio Atleti Azzurri d'Italia    October 25, 2017 at 6:30:00 PM GMT+2  2702.044567          1H     0   5.013232  1.099962   3  44   6.0           1           0          0                0          0          0       0               0            1        0    0.138239  45.034076\n",
       "38944  0.437796     0.300350              Konstantinos                  Manolas                       Fiorentina - Roma, 2 - 4                        Stadio Artemio Franchi    November 5, 2017 at 3:00:00 PM GMT+1   246.580220          2H     1   4.939636  1.227069   4  46   4.0           0           0          0                0          0          1       1               0            0        0    0.137446  49.109670\n",
       "31887  0.529129     0.392009               Adri\\u00e1n  L\\u00f3pez \\u00c1lvarez  Real Madrid - Deportivo La Coru\\u00f1a, 7 - 1                Estadio Santiago Bernab\\u00e9u    January 21, 2018 at 4:15:00 PM GMT+1  1356.229466          1H     1   5.011237  0.919518   2  57   7.0           1           0          0                0          0          0       0               0            1        0    0.137120  22.603824\n",
       "39603  0.436872     0.300350                  Leonardo                Pavoletti                      Bologna - Cagliari, 1 - 1                        Stadio Renato Dall'Ara    December 3, 2017 at 3:00:00 PM GMT+1   476.708858          1H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.136522   7.945148\n",
       "38569  0.436872     0.300350                    Maxime                 Gonalons                          Roma - Crotone, 1 - 0                               Stadio Olimpico    October 25, 2017 at 8:45:00 PM GMT+2  2446.202217          2H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.136522  85.770037\n",
       "39631  0.436872     0.300350                  Gabriele                  Angella                       Crotone - Udinese, 0 - 3                             Stadio Ezio Scida    December 4, 2017 at 7:00:00 PM GMT+1   798.307165          2H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.136522  58.305119\n",
       "33959  0.436872     0.300350         Cristhian Ricardo           Stuani Curbelo                    Real Madrid - Girona, 6 - 3                Estadio Santiago Bernab\\u00e9u      March 18, 2018 at 8:45:00 PM GMT+1  1313.810278          2H     1   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.136522  66.896838\n",
       "42068  0.522738     0.387040                  Giovanni                 Crociata                     Benevento - Crotone, 3 - 2                          Stadio Ciro Vigorito   February 18, 2018 at 3:00:00 PM GMT+1   639.416196          1H     1   5.011237  0.919518   2  43   7.0           1           0          0                0          0          0       0               0            1        0    0.135698  10.656937\n",
       "5531   0.338781     0.204223                     Chris                 Smalling    Manchester United - Newcastle United, 4 - 1                                  Old Trafford   November 18, 2017 at 6:30:00 PM GMT+1  2754.483406          1H     1   5.011237  0.919518   2  57   7.0           0           0          0                0          0          1       1               0            0        0    0.134557  45.908057\n",
       "8999   0.560988     0.426878                    Wesley                    Hoedt                 Southampton - Liverpool, 0 - 2                            St. Mary's Stadium   February 11, 2018 at 5:30:00 PM GMT+1   809.491575          2H     0   5.011237  0.919518   2  43   7.0           1           0          0                0          0          0       0               0            1        0    0.134111  58.491526\n",
       "21045  0.153877     0.021078                Levin Mete            \\u00d6ztunali             Mainz 05 - Bayer Leverkusen, 3 - 1                                    OPEL ARENA   September 9, 2017 at 3:30:00 PM GMT+2   503.758742          2H     0  38.608030  0.134465  26   8  42.0           0           0          0                0          1          0       0               0            1        0    0.132799  53.395979\n",
       "25569  0.450637     0.317861                      Erik                   Thommy         Stuttgart - Eintracht Frankfurt, 1 - 0                           Mercedes-Benz-Arena   February 24, 2018 at 3:30:00 PM GMT+1   731.374619          1H     1   1.300000 -0.000000   0  52   2.0           0           0          0                0          0          0       0               0            1        0    0.132776  12.189577\n",
       "37091  0.583800     0.451782                    Andrea                  Belotti                      Torino - Sampdoria, 2 - 2                 Stadio Olimpico Grande Torino  September 17, 2017 at 3:00:00 PM GMT+2  1870.262658          2H     0   5.013232  1.099962   3  56   6.0           1           0          0                0          0          0       0               0            1        0    0.132018  76.171044\n",
       "40441  0.142784     0.010784                    Simone                    Verdi                        Chievo - Bologna, 2 - 3                 Stadio Marc'Antonio Bentegodi   December 22, 2017 at 6:00:00 PM GMT+1  2763.753113          1H     0  27.942486  0.079956   8  91  41.0           0           0          0                0          0          0       0               0            1        0    0.132000  46.062552\n",
       "39367  0.299620     0.167642  Mat\\u00edas Agust\\u00edn                Silvestre                     Bologna - Sampdoria, 3 - 0                        Stadio Renato Dall'Ara   November 25, 2017 at 3:00:00 PM GMT+1  1384.404262          2H     0   5.304951  0.480459   1  42   8.0           1           0          0                0          0          0       1               0            0        0    0.131977  68.073404\n",
       "6273   0.402479     0.270788                      Jeff                 Hendrick                       Burnley - Watford, 1 - 0                                     Turf Moor    December 9, 2017 at 4:00:00 PM GMT+1  1055.812711          1H     0   7.201562  0.674273   4  41   9.0           1           0          0                0          0          0       0               0            0        0    0.131691  17.596879\n",
       "9535   0.378322     0.248322                     Harry                     Kane   Tottenham Hotspur - Huddersfield Town, 2 - 0                               Wembley Stadium       March 3, 2018 at 4:00:00 PM GMT+1  2011.749290          2H     0   9.450000  0.739024   9  50   0.0           1           0          0                0          0          0       0               0            1        0    0.130000  78.529155\n",
       "5721   0.433509     0.303562                   Charlie                   Austin                   Southampton - Everton, 4 - 1                            St. Mary's Stadium   November 26, 2017 at 2:30:00 PM GMT+1   411.968931          2H     1   4.526036  1.272510   3  45   5.0           0           0          0                0          0          1       1               0            0        0    0.129946  51.866149\n",
       "4457   0.433509     0.303562                     Ahmed                   Hegazy          West Bromwich Albion - Watford, 2 - 2                                 The Hawthorns  September 30, 2017 at 4:00:00 PM GMT+2  1892.973924          1H     0   4.526036  1.272510   3  45   5.0           0           0          0                0          0          1       1               0            0        0    0.129946  31.549565\n",
       "43798  0.410843     0.280911                     Fabio             Quagliarella                    Juventus - Sampdoria, 3 - 0                                           NaN      April 15, 2018 at 6:00:00 PM GMT+2  1256.855706          1H     0   4.429447  1.186057   2  44   6.0           0           0          0                0          0          1       1               0            0        0    0.129932  20.947595\n",
       "16512  0.547058     0.417874                    Rafael         Pereira da Silva                Olympique Lyonnais - PSG, 2 - 1                              Groupama Stadium    January 21, 2018 at 9:00:00 PM GMT+1  1963.196299          2H     0   3.342155  1.714580   2  46   4.0           0           0          0                0          0          1       1               0            0        0    0.129185  77.719938\n",
       "16907  0.547058     0.417874                    Damien                 Da Silva                           Caen - Nantes, 3 - 2                         Stade Michel d'Ornano    February 4, 2018 at 5:00:00 PM GMT+1  2060.083861          2H     1   3.342155  1.714580   2  46   4.0           0           0          0                0          0          1       1               0            0        0    0.129185  79.334731\n",
       "44225  0.413372     0.284200                  Benedikt             H\\u00f6wedes                       Juventus - Napoli, 0 - 1                                           NaN      April 22, 2018 at 8:45:00 PM GMT+2   769.871978          1H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.129172  12.831200\n",
       "43250  0.413372     0.284200                  Leonardo                  Bonucci                        Juventus - Milan, 3 - 1                                           NaN      March 31, 2018 at 8:45:00 PM GMT+2  1656.863687          1H     1   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.129172  27.614395\n",
       "38814  0.408213     0.279629                    Andrea                  Belotti                       Torino - Cagliari, 2 - 1                 Stadio Olimpico Grande Torino    October 29, 2017 at 8:45:00 PM GMT+1  1147.947197          2H     0   5.011237  0.919518   2  57   7.0           1           0          0                0          0          0       1               0            0        0    0.128584  64.132453\n",
       "11598  0.148144     0.019777                    Thomas                  Mangani                   Angers SCO - Bordeaux, 2 - 2                            Stade Raymond Kopa      August 6, 2017 at 5:00:00 PM GMT+2  1177.624908          2H     0  29.920770  0.052247   6  95  45.0           0           0          0                0          1          0       0               0            1        0    0.128367  64.627082\n",
       "31558  0.146985     0.019787         Johan Andr\\u00e9s           Mojica Palacio                       Valencia - Girona, 2 - 1                           Estadio de Mestalla     January 6, 2018 at 4:15:00 PM GMT+1  1103.834403          2H     0  28.900260  0.065373   7   7  43.0           0           0          0                0          0          1       0               1            1        0    0.127199  63.397240\n",
       "10480  0.420941     0.294824                     Kevin                     Long                Burnley - Leicester City, 2 - 1                                     Turf Moor      April 14, 2018 at 4:00:00 PM GMT+2   490.414371          1H     1   4.939636  1.227069   4  46   4.0           0           0          0                0          0          1       1               0            0        0    0.126118   8.173573\n",
       "9979   0.420584     0.294824                    Dwight                    Gayle    Tottenham Hotspur - Newcastle United, 1 - 0                               Wembley Stadium         May 9, 2018 at 9:00:00 PM GMT+2  1691.893432          1H     0   4.939636  1.227069   4  54   4.0           0           0          0                0          0          1       1               0            0        0    0.125760  28.198224\n",
       "16789  0.183819     0.059416                    Mario         Barwuah Balotelli                             Metz - Nice, 2 - 1                        Stade Saint-Symphorien    January 27, 2018 at 8:00:00 PM GMT+1  2797.378458          2H     0  65.353959  0.097173  54   0  50.0           0           0          0                0          1          0       0               0            1        0    0.124403  91.622974\n",
       "18493  0.384471     0.261086                 Max-Alain                   Gradel                   Toulouse - Strasbourg, 2 - 2                             Stadium Municipal      March 17, 2018 at 8:00:00 PM GMT+1  1449.737972          2H     0   5.013232  1.099962   3  44   6.0           0           0          0                0          0          1       1               0            0        0    0.123385  69.162300\n",
       "39284  0.330167     0.206798                  Perparim                  Hetemaj                         Torino - Chievo, 1 - 1                 Stadio Olimpico Grande Torino   November 19, 2017 at 3:00:00 PM GMT+1   801.579435          1H     1   5.011237  0.919518   2  43   7.0           0           0          0                0          0          1       1               0            0        0    0.123370  13.359657\n",
       "25061  0.441783     0.318973            \\u00c1d\\u00e1m                   Szalai                   Hoffenheim - Mainz 05, 4 - 2                     WIRSOL Rhein-Neckar-Arena   February 10, 2018 at 3:30:00 PM GMT+1  2524.988635          1H     0   4.630605  1.314800   4  53   3.0           0           0          0                0          0          1       1               0            0        0    0.122809  42.083144\n",
       "20871  0.438264     0.315581                 Yoshinori                     Muto                    Stuttgart - Mainz 05, 1 - 0                           Mercedes-Benz-Arena     August 26, 2017 at 3:30:00 PM GMT+2  2822.661547          2H     0   4.526036  1.272510   3  55   5.0           0           0          0                0          0          1       1               0            0        0    0.122684  92.044359\n",
       "21163  0.438264     0.315581                      Timo                   Werner        RB Leipzig - Borussia M'gladbach, 2 - 2                                Red Bull Arena  September 16, 2017 at 6:30:00 PM GMT+2  2098.759279          2H     0   4.526036  1.272510   3  55   5.0           0           0          0                0          0          1       1               0            0        0    0.122684  79.979321\n",
       "21907  0.402299     0.279687                      Marc           Bartra Aregall            Augsburg - Borussia Dortmund, 1 - 2                                     WWK Arena  September 30, 2017 at 3:30:00 PM GMT+2  1697.117480          1H     0   5.011237  0.919518   2  43   7.0           1           0          0                0          0          0       1               0            0        0    0.122613  28.285291\n",
       "25636  0.563004     0.440416                    Thomas              M\\u00fcller          Freiburg - Bayern M\\u00fcnchen, 0 - 4                           Schwarzwald-Stadion       March 4, 2018 at 6:00:00 PM GMT+1  1399.201051          2H     1   5.013232  1.099962   3  56   6.0           1           0          0                0          0          0       0               0            1        0    0.122588  68.320018\n",
       "40071  0.424671     0.302244                   Edoardo                Goldaniga                      Sassuolo - Crotone, 2 - 1      MAPEI Stadium - Citt\\u00e0 del Tricolore   December 10, 2017 at 6:00:00 PM GMT+1   200.830730          2H     1   4.526036  1.272510   3  45   5.0           0           0          0                0          0          1       1               0            0        0    0.122427  48.347179\n",
       "36461  0.611426     0.489785                   Patrick                  Cutrone                        Milan - Cagliari, 2 - 1                        Stadio Giuseppe Meazza     August 27, 2017 at 8:45:00 PM GMT+2   585.115060          1H     1   4.429447  1.186057   2  44   6.0           1           0          0                0          0          0       0               0            1        0    0.121642   9.751918\n",
       "43249  0.328164     0.206798         Andr\\u00e9 Miguel            Valente Silva                        Juventus - Milan, 3 - 1                                           NaN      March 31, 2018 at 8:45:00 PM GMT+2   908.084895          1H     0   5.011237  0.919518   2  57   7.0           0           0          0                0          0          1       1               0            0        0    0.121366  15.134748\n",
       "37604  0.331155     0.210887                    Jasmin              Kurti\\u010d                     Atalanta - Juventus, 2 - 2                Stadio Atleti Azzurri d'Italia     October 1, 2017 at 8:45:00 PM GMT+2  1995.816106          1H     0   5.533986  0.929077   3  43   7.0           0           0          0                0          0          1       1               0            0        0    0.120268  33.263602\n",
       "34967  0.331155     0.210887         Cristhian Ricardo           Stuani Curbelo                       Girona - Espanyol, 0 - 2                 Estadi Municipal de Montilivi     April 22, 2018 at 12:00:00 PM GMT+2  1364.922783          2H     0   5.533986  0.929077   3  43   7.0           0           0          0                0          0          1       1               0            0        0    0.120268  67.748713\n",
       "39416  0.331155     0.210887                   Alberto                   Grassi                           Chievo - SPAL, 2 - 1                 Stadio Marc'Antonio Bentegodi   November 25, 2017 at 6:00:00 PM GMT+1   714.819160          2H     0   5.533986  0.929077   3  43   7.0           0           0          0                0          0          1       1               0            0        0    0.120268  56.913653\n",
       "14133  0.127432     0.008118                    Morgan                   Sanson              Olympique Marseille - Caen, 5 - 0                          Stade V\\u00e9lodrome    November 5, 2017 at 5:00:00 PM GMT+1  1917.190108          1H     0  33.220852  0.105357  15   5  45.0           0           0          1                0          0          0       0               0            1        0    0.119314  31.953168"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.sort_values('difference', ascending=False).head(60)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Twitter Poll Review:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Last week, I posted a series of polls on Twitter (https://twitter.com/andrew_puopolo/status/1262739966548942848) with 8 shots from the 17/18 Premier League season, and asked people to vote on their xG. Those 8 shots were the 8 shots in the above two dataframes that were a) In the Premier League (thus higher probability people on my Twitter would remember) and b) I could find a good video for. They are sorted by chronological order"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mohamed Salah vs Watford"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video: </b> https://twitter.com/andrew_puopolo/status/1262739966548942848\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.4: 12% <br>\n",
    "0.4: 6.6% <br>\n",
    "0.55%: 19.7% <br>\n",
    "Greater: 61.6% <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                          0.39571\n",
       "Logistic XG                                   0.558862\n",
       "firstName                                     Mohamed \n",
       "lastName                                   Salah Ghaly\n",
       "label                       Watford - Liverpool, 3 - 3\n",
       "venue                            Vicarage Road Stadium\n",
       "date               August 12, 2017 at 1:30:00 PM GMT+2\n",
       "eventSec                                       712.521\n",
       "matchPeriod                                         2H\n",
       "Goal                                                 1\n",
       "Distance                                       6.19213\n",
       "Angle                                         0.889119\n",
       "X                                                    4\n",
       "Y                                                   57\n",
       "C                                                    7\n",
       "prev_cross                                           0\n",
       "prev_touch                                           0\n",
       "prev_pass                                            0\n",
       "prev_smart_pass                                      1\n",
       "free_kick                                            0\n",
       "prev_duel                                            0\n",
       "header                                               0\n",
       "counter_attack                                       0\n",
       "strong_foot                                          1\n",
       "rebound                                              0\n",
       "difference                                   -0.163153\n",
       "Minute                                         56.8754\n",
       "Name: 2996, dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[2996]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the first video, we have Mo Salah finishing into basically an empty net. Most of the people voted for an xG that was higher than the two models' output. This is likely because Salah was shooting into an empty net. However, it is important to note that this model does not pick this up, and only sees some bump from the previous event being a \"smart pass\", which we will talk about more later. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another interesting notef rom this shot was that while the actual shot was high expected value, it was incredibly difficult for Salah to latch onto that ball, and thus it was incredibly low probability for any shot to be taken from that position. This points directly at one of the assumptions of expected goals models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dele vs Manchester United"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video </b>: https://twitter.com/andrew_puopolo/status/1262740396007845890\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.42:  27.6% <br>\n",
    "0.42: 33% <br>\n",
    "0.6: 24% <br>\n",
    "Greater than 0.6: 15.1% <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                                  0.424602\n",
       "Logistic XG                                            0.591979\n",
       "firstName                                              Bamidele\n",
       "lastName                                                   Alli\n",
       "label              Manchester United - Tottenham Hotspur, 1 - 0\n",
       "venue                                              Old Trafford\n",
       "date                       October 28, 2017 at 1:30:00 PM GMT+2\n",
       "eventSec                                                1907.08\n",
       "matchPeriod                                                  2H\n",
       "Goal                                                          0\n",
       "Distance                                                6.33344\n",
       "Angle                                                   1.04571\n",
       "X                                                             6\n",
       "Y                                                            51\n",
       "C                                                             1\n",
       "prev_cross                                                    0\n",
       "prev_touch                                                    0\n",
       "prev_pass                                                     0\n",
       "prev_smart_pass                                               1\n",
       "free_kick                                                     0\n",
       "prev_duel                                                     0\n",
       "header                                                        0\n",
       "counter_attack                                                0\n",
       "strong_foot                                                   1\n",
       "rebound                                                       0\n",
       "difference                                            -0.167377\n",
       "Minute                                                  76.7846\n",
       "Name: 5104, dtype: object"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[5104]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we see the Twitter poll is incredibly divided between the four options. If we look at the feature values, they are remarkably similar to the Salah chance. Both shots are good angles with the strong foot coming off a \"smart pass.\" So what makes this shot different in the eyes of the viewers? Well, in this  shot Dele has to twist his body to shoot directly into De Gea, who has already cut off the angle. Unfortunately, our model does not account for this, as that data is not available tou s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Olivier Giroud vs Huddersfield"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video: </b> https://twitter.com/andrew_puopolo/status/1262740688938008578\n",
    "\n",
    "Vote: \n",
    "\n",
    "Less than 0.35: 8% <br> \n",
    "0.35%: 11.4% <br>\n",
    "0.5: 25.1% <br>\n",
    "Greater than  0.5: 55.4% <br>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                           0.343853\n",
       "Logistic XG                                      0.49742\n",
       "firstName                                        Olivier\n",
       "lastName                                          Giroud\n",
       "label                 Arsenal - Huddersfield Town, 5 - 0\n",
       "venue                                   Emirates Stadium\n",
       "date               November 29, 2017 at 8:45:00 PM GMT+1\n",
       "eventSec                                         2496.07\n",
       "matchPeriod                                           2H\n",
       "Goal                                                   1\n",
       "Distance                                         6.43273\n",
       "Angle                                            1.02537\n",
       "X                                                      6\n",
       "Y                                                     52\n",
       "C                                                      2\n",
       "prev_cross                                             0\n",
       "prev_touch                                             0\n",
       "prev_pass                                              0\n",
       "prev_smart_pass                                        0\n",
       "free_kick                                              0\n",
       "prev_duel                                              1\n",
       "header                                                 0\n",
       "counter_attack                                         0\n",
       "strong_foot                                            1\n",
       "rebound                                                0\n",
       "difference                                     -0.153567\n",
       "Minute                                           86.6012\n",
       "Name: 5836, dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[5836]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On this shot, everyone took the \"over\". The location once again is very similar, but this goal is tagged with a \"duel\" as the previous event, as there was a bit of a skirmish in the box. This is a weird facet of the data. In addition,  there  are likely very few datapoints (strong foot shots after a duel from right in front of the net) that are similar to this, which can lead to the models doing weird things as it is much harder to learn"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Glenn Murray vs AFC Bournemouth"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video: </b> https://twitter.com/andrew_puopolo/status/1262740974544973827\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.4: 3% <br>\n",
    "0.4: 13.3% <br>\n",
    "0.65: 41% <br>\n",
    "Greater than 0.65: 42.8% "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                                      0.39503\n",
       "Logistic XG                                               0.646921\n",
       "firstName                                                    Glenn\n",
       "lastName                                                    Murray\n",
       "label              Brighton & Hove Albion - AFC Bournemouth, 2 - 2\n",
       "venue                       The American Express Community Stadium\n",
       "date                           January 1, 2018 at 1:30:00 PM GMT+1\n",
       "eventSec                                                   140.051\n",
       "matchPeriod                                                     2H\n",
       "Goal                                                             1\n",
       "Distance                                                   7.79631\n",
       "Angle                                                     0.848711\n",
       "X                                                                7\n",
       "Y                                                               54\n",
       "C                                                                4\n",
       "prev_cross                                                       0\n",
       "prev_touch                                                       0\n",
       "prev_pass                                                        0\n",
       "prev_smart_pass                                                  1\n",
       "free_kick                                                        0\n",
       "prev_duel                                                        0\n",
       "header                                                           0\n",
       "counter_attack                                                   1\n",
       "strong_foot                                                      1\n",
       "rebound                                                          0\n",
       "difference                                               -0.251892\n",
       "Minute                                                     47.3342\n",
       "Name: 7759, dtype: object"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[7759]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we have almost the same features as above. A strong foot shot from the 6 yard box, but the main difference here is that we are on a counter attack, which boosts the expected goals models. Again, the higher values were among the higher voted"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ayoze Perez vs Stoke City"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video </b>: https://twitter.com/andrew_puopolo/status/1262741307014787075\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.4: 9.6% <br>\n",
    "0.4: 30.1% <br>\n",
    "0.6: 37.2% <br>\n",
    "Greater than 0.6: 23.1% "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                           0.38857\n",
       "Logistic XG                                    0.604483\n",
       "firstName                                         Ayoze\n",
       "lastName                      P\\u00e9rez Guti\\u00e9rrez\n",
       "label              Stoke City - Newcastle United, 0 - 1\n",
       "venue                                    Bet365 Stadium\n",
       "date                January 1, 2018 at 4:00:00 PM GMT+1\n",
       "eventSec                                        1632.24\n",
       "matchPeriod                                          2H\n",
       "Goal                                                  1\n",
       "Distance                                        8.62337\n",
       "Angle                                          0.789654\n",
       "X                                                     8\n",
       "Y                                                    53\n",
       "C                                                     3\n",
       "prev_cross                                            0\n",
       "prev_touch                                            0\n",
       "prev_pass                                             0\n",
       "prev_smart_pass                                       1\n",
       "free_kick                                             0\n",
       "prev_duel                                             0\n",
       "header                                                0\n",
       "counter_attack                                        1\n",
       "strong_foot                                           1\n",
       "rebound                                               0\n",
       "difference                                    -0.215913\n",
       "Minute                                           72.204\n",
       "Name: 7913, dtype: object"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[7913]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we see similar results to the Murray goal. It appears as though the Random Forest doesn't boost the xG prediction on counter attacks and previous smart passes as much as the Logistic Regression does, so that's something to keep in mind"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Wesley Hoedt vs Liverpool"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video: </b> https://twitter.com/andrew_puopolo/status/1262741636037046277\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.45: 71.4% <br>\n",
    "0.45: 15.6% <br>\n",
    "0.55: 8.2% <br>\n",
    "Greater than 0.55: 4.8% "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                           0.560988\n",
       "Logistic XG                                     0.426878\n",
       "firstName                                         Wesley\n",
       "lastName                                           Hoedt\n",
       "label                     Southampton - Liverpool, 0 - 2\n",
       "venue                                 St. Mary's Stadium\n",
       "date               February 11, 2018 at 5:30:00 PM GMT+1\n",
       "eventSec                                         809.492\n",
       "matchPeriod                                           2H\n",
       "Goal                                                   0\n",
       "Distance                                         5.01124\n",
       "Angle                                           0.919518\n",
       "X                                                      2\n",
       "Y                                                     43\n",
       "C                                                      7\n",
       "prev_cross                                             1\n",
       "prev_touch                                             0\n",
       "prev_pass                                              0\n",
       "prev_smart_pass                                        0\n",
       "free_kick                                              0\n",
       "prev_duel                                              0\n",
       "header                                                 0\n",
       "counter_attack                                         0\n",
       "strong_foot                                            1\n",
       "rebound                                                0\n",
       "difference                                      0.134111\n",
       "Minute                                           58.4915\n",
       "Name: 8999, dtype: object"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[8999]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we have our first example of a Random Forest spitting out a higher prediction than the Logistic. Here, the people voted that the expected goals was lower than either model's prediction, but why? There's a few reasons. The first is that the shot did not result in a goal, and thus far the votes have been skewed towards the actual outcome. The second is that the keeper meets Hoedt at the ball. The third potential reason is the voters are biasedtowards the shooter. In the above examples, the shooting player is either a striker or an attacking midfielder. Hoedt is a center back, and as such his shooting form is a bit awkward, which likely biases the viewers towards voting for lower xG."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dwight Gayle vs Spurs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video </b>: https://twitter.com/andrew_puopolo/status/1262741872138649600\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.3: 53.5% <br>\n",
    "0.3: 27.5% <br>\n",
    "0.4: 12% <br>\n",
    "Greater than 0.4: 7%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                                 0.420584\n",
       "Logistic XG                                           0.294824\n",
       "firstName                                               Dwight\n",
       "lastName                                                 Gayle\n",
       "label              Tottenham Hotspur - Newcastle United, 1 - 0\n",
       "venue                                          Wembley Stadium\n",
       "date                           May 9, 2018 at 9:00:00 PM GMT+2\n",
       "eventSec                                               1691.89\n",
       "matchPeriod                                                 1H\n",
       "Goal                                                         0\n",
       "Distance                                               4.93964\n",
       "Angle                                                  1.22707\n",
       "X                                                            4\n",
       "Y                                                           54\n",
       "C                                                            4\n",
       "prev_cross                                                   0\n",
       "prev_touch                                                   0\n",
       "prev_pass                                                    0\n",
       "prev_smart_pass                                              0\n",
       "free_kick                                                    0\n",
       "prev_duel                                                    1\n",
       "header                                                       1\n",
       "counter_attack                                               0\n",
       "strong_foot                                                  0\n",
       "rebound                                                      0\n",
       "difference                                             0.12576\n",
       "Minute                                                 28.1982\n",
       "Name: 9979, dtype: object"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[9979]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we see similar trends to above. Gayle is at great stretch to reach the ball.  We also notice a potential trend of Random Forest's failing to discount the previous event being a cross in its prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Jamie Vardy vs Spurs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Video: </b> https://twitter.com/andrew_puopolo/status/1262742261541924865\n",
    "\n",
    "<b> Vote: </b>\n",
    "\n",
    "Less than 0.4: 73.7% <br>\n",
    "0.4: 21.1% <br>\n",
    "0.55: 3.9% <br>\n",
    "Greater than 0.55: 1.3%"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RF XG                                               0.413132\n",
       "Logistic XG                                         0.564663\n",
       "firstName                                              Jamie\n",
       "lastName                                               Vardy\n",
       "label              Tottenham Hotspur - Leicester City, 5 - 4\n",
       "venue                                        Wembley Stadium\n",
       "date                        May 13, 2018 at 4:00:00 PM GMT+2\n",
       "eventSec                                             1667.79\n",
       "matchPeriod                                               2H\n",
       "Goal                                                       1\n",
       "Distance                                             7.86034\n",
       "Angle                                               0.671183\n",
       "X                                                          5\n",
       "Y                                                         41\n",
       "C                                                          9\n",
       "prev_cross                                                 0\n",
       "prev_touch                                                 0\n",
       "prev_pass                                                  0\n",
       "prev_smart_pass                                            1\n",
       "free_kick                                                  0\n",
       "prev_duel                                                  0\n",
       "header                                                     0\n",
       "counter_attack                                             1\n",
       "strong_foot                                                0\n",
       "rebound                                                    0\n",
       "difference                                         -0.151531\n",
       "Minute                                               72.7964\n",
       "Name: 11553, dtype: object"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.loc[11553]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, we get to Jamie Vardy's shot against Spurs. The voters strayed away from results bias, and voted for low values. However, the trend of the logistic boosting smart pass/counter attack more continues. i"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###  Takeaways  from video exercise"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Point 1: </b> People are reactant to the final result. On 7 out of 8 shots, the votes were skewed towards the actual result of the shot. This can sometimes be fine (especially when there are other factors at play, see point #2) but this is something to consider when giving your expected goals model the \"eye test\"."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Point 2: </b> Expected goals models are only as good as your data. Our expected goals models only take into account information from the sequence of events, and thus fail to take into account other information, such as the location of other players or the goalkeeper, or the player's body pose when attempting the shot. Other datasets (like StatsBomb) include this information, so this could be part of the discrepency between different xG models that exist on the internet."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b> Point 3: </b> The Random Forest gives higher values to shots where the previous event was a cross, whereas the logistic regression gives higher values to counter attacks and goals where the previous event was a smart pass. We will take a look at this in the next section."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting Our Expected Goals Surfaces"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the previous section, we developed some theories about how our models were performing. The best way to test those out would be to plot the models against each other on a playing surface."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To do this, we leverage code from David Sumpter and FC Python, but adapt it to compare two models to each other."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On Friday, I posted some teasers of these plots on Twitter (https://twitter.com/andrew_puopolo/status/1263833631107227648). In this notebook, we will look at different plots"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, we start by fitting our model on <b> all </b> the data, since our plots will technically be predicting on new data. By fitting on all the data, we are (in theory) maximizing our accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "logistic_regression = LogisticRegression(random_state=42, max_iter=1000).fit(x_logreg,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "random_forest_xg_model = RandomForestClassifier(random_state=42, n_estimators=500, criterion='gini', max_depth=8,\n",
    "                                                max_features=8, min_samples_leaf=5, min_samples_split=250\n",
    "                                               ).fit(x_rf, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we create a grid for every X/Y location on the pitch, and convert this into a dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xs=np.linspace(0, 49, 50)\n",
    "Ys=np.linspace(0, 99, 100)\n",
    "datapoints=[]\n",
    "for x in Xs:\n",
    "    for y in Ys:\n",
    "        datapoints.append([x,y])\n",
    "initial_data = pd.DataFrame(datapoints, columns=['X', 'Y'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here, we construct the values of the features that can be directly derived from the X and Y coordinates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "initial_data['side_of_field']= 1*(initial_data['Y'] < 50)\n",
    "initial_data['C']=abs(initial_data['Y']-50)\n",
    "initial_data['Distance']=np.sqrt(initial_data['X']**2 + initial_data['C']**2)\n",
    "initial_data['Angle']=np.arctan(7.32 *initial_data['X'] /(initial_data['X']**2 + initial_data['C']**2 - (7.32/2)**2))\n",
    "initial_data['Angle']=initial_data['Angle']+(initial_data['Angle'] < 0)*np.pi\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we create a function that automatically generates the design  matrix  for our model, given user inputted features values for each feature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def initialize_data_constants(dataset, right_foot=1, left_foot=0, header=0, counter_attack=0, \n",
    "                              strong_foot=1, rebound=0, prev_cross=0, prev_pass=0, prev_touch=0, prev_smart_pass=0,\n",
    "                              free_kick=0, prev_duel=0, random_forest=False):\n",
    "    dataset=dataset.copy()\n",
    "    dataset['counter_attack'] = counter_attack\n",
    "    dataset['free_kick'] = free_kick\n",
    "    dataset['header'] = header\n",
    "    dataset['prev_cross'] = prev_cross\n",
    "    dataset['prev_duel'] = prev_duel\n",
    "    dataset['prev_smart_pass'] = prev_smart_pass\n",
    "    dataset['strong_foot'] = strong_foot\n",
    "    dataset['right_foot'] = right_foot\n",
    "    \n",
    "    if random_forest:\n",
    "        dataset['left_foot'] = left_foot\n",
    "        dataset['rebound'] = rebound\n",
    "        dataset['prev_pass'] = prev_pass\n",
    "        dataset['prev_touch'] = prev_touch\n",
    "        dataset['in_swinging'] = 1*(dataset['side_of_field'] == 0)*(dataset['left_foot'] == 1) \\\n",
    "                            + 1*(dataset['side_of_field'] == 1)*(dataset['right_foot'] == 1)\n",
    "        dataset['out_swinging'] = 1*(dataset['side_of_field'] == 0)*(dataset['right_foot'] == 1) \\\n",
    "                           + 1*(dataset['side_of_field'] ==1)*(dataset['left_foot'] == 1)\n",
    "        dataset = dataset.reindex(x_rf.columns, axis=1)\n",
    "\n",
    "    else:\n",
    "        dataset=dataset.drop(columns=['C', 'X', 'Y', 'side_of_field'])\n",
    "        dataset['distance_angle'] = dataset['Distance']*dataset['Angle']\n",
    "        dataset['distance_squared'] = dataset['Distance']*dataset['Distance']\n",
    "        dataset['duel_header'] = dataset['prev_duel']*dataset['header']\n",
    "        dataset['free_kick_angle'] = dataset['free_kick']*dataset['Angle']\n",
    "        dataset['strong_angle'] = dataset['strong_foot']*dataset['Angle']\n",
    "        dataset['strong_distance'] = dataset['strong_foot']*dataset['Distance']\n",
    "        dataset['strong_distance_squared'] = dataset['strong_foot']*dataset['Distance']*dataset['Distance']\n",
    "        dataset = dataset.reindex(x_logreg.columns, axis=1)\n",
    "\n",
    "    return dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this function, we calculate the expected goals model for each value on the pitch. However, we add some flexibility:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we want to plot the expected goals surface of one model, we only use the arguments ``model_1`` and ``params_1``. ``model_1`` is the model object we wish to use (either the Random Forest or the Logistic Regression), while ``params`` is the feature values we want to plug into the above function."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we want to plot the difference between two models, we would use both the ``model_2`` and ``params_2`` arguments. Here, we have the flexibility to test two different things. 1.) We have the flexibility to compare the Random Forest and Logistic Regression model with the same features. 2.) We have the flexibility to see how the surfaces compare after changing the features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_difference_plot(model_1, params_1, plot_title, model_2=None, params_2=None):\n",
    "    \n",
    "    (fig,ax) = FCPython.createGoalMouth()\n",
    "\n",
    "    if model_1 == random_forest_xg_model:\n",
    "        model_1_data = initialize_data_constants(initial_data, random_forest=True, **params_1)\n",
    "        model_1_data['prediction']=model_1.predict_proba(model_1_data)[:,-1]\n",
    "        \n",
    "    else:\n",
    "        model_1_data = initialize_data_constants(initial_data, **params_1)\n",
    "        model_1_data['prediction']=model_1.predict_proba(model_1_data)[:,-1]\n",
    "        model_1_data=pd.merge(model_1_data, initial_data[['Distance', \"Angle\", \"X\", \"Y\"]], \n",
    "                             on=[\"Distance\", \"Angle\"], how='left')\n",
    "        model_1_data=model_1_data.drop_duplicates()\n",
    "    model_1_pred=np.zeros((50,100))\n",
    "    \n",
    "    for ind,row in model_1_data.iterrows():\n",
    "        x=row['X']\n",
    "        y=row['Y']\n",
    "        model_1_pred[int(x)][int(y)]=row['prediction']\n",
    "    \n",
    "    if model_2 is not None:\n",
    "        if model_2 == random_forest_xg_model:\n",
    "            model_2_data = initialize_data_constants(initial_data, random_forest=True, **params_2)\n",
    "            model_2_data['prediction']=model_2.predict_proba(model_2_data)[:,-1]\n",
    "            \n",
    "        else:\n",
    "            model_2_data = initialize_data_constants(initial_data, **params_2)\n",
    "            model_2_data['prediction']=model_2.predict_proba(model_2_data)[:,-1]\n",
    "            model_2_data=pd.merge(model_2_data, initial_data[['Distance', \"Angle\", \"X\", \"Y\"]], \n",
    "                                 on=[\"Distance\", \"Angle\"], how='left')\n",
    "            model_2_data=model_2_data.drop_duplicates()\n",
    "        model_2_pred=np.zeros((50,100))\n",
    "        for ind,row in model_2_data.iterrows():\n",
    "            x=row['X']\n",
    "            y=row['Y']\n",
    "            model_2_pred[int(x)][int(y)]=row['prediction']\n",
    "        min_value = np.min(model_1_pred-model_2_pred)\n",
    "        max_value = np.max(model_1_pred-model_2_pred)\n",
    "        flip=1\n",
    "        cmap='bwr'\n",
    "        if min_value > 0:\n",
    "            scale_min = 0\n",
    "            scale_max = max_value\n",
    "            cmap='Reds'\n",
    "        elif max_value < 0:\n",
    "            scale_min = 0\n",
    "            scale_max = -1*min_value\n",
    "            flip=-1\n",
    "            cmap='Blues'\n",
    "\n",
    "        else:\n",
    "            scale_max = max(abs(min_value), max_value)\n",
    "            scale_min = -1*scale_max\n",
    "\n",
    "        pos=ax.imshow(flip*(model_1_pred-model_2_pred), extent=[-1,66,51.5,-1], aspect='auto',cmap=cmap,\n",
    "                      vmin=scale_min, vmax=scale_max)\n",
    "        fig.colorbar(pos, ax=ax)\n",
    "        ax.set_title(plot_title)\n",
    "        plt.ylim((-3,35))\n",
    "        plt.tight_layout()\n",
    "        plt.gca().set_aspect('equal', adjustable='box')\n",
    "        plt.show()     \n",
    "    \n",
    "    else:\n",
    "        pos=ax.imshow(model_1_pred, extent=[-1,66,51.5,-1], aspect='auto',cmap=plt.cm.Reds, vmin=0, \n",
    "                      vmax=np.max(model_1_pred))\n",
    "        fig.colorbar(pos, ax=ax)\n",
    "        ax.set_title(plot_title)\n",
    "        plt.ylim((-3,35))\n",
    "        plt.tight_layout()\n",
    "        plt.gca().set_aspect('equal', adjustable='box')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's look at a simple surface for a logistic regression model, with default parameters. This assumes that the player took a shot with his strong foot, with no additional taggers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=logistic_regression, params_1={}, plot_title='Strong Foot Logistic Regression')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we'll do the same thing for the Random Forest model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, params_1={}, \n",
    "                         plot_title='Strong Foot Random Forest')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find that this xG surface is a lot less smooth than the one in the Logistic Regression plot. This matches up with the analysis from Jesse Davis and Pieter Robberechts (https://dtai.cs.kuleuven.be/sports/blog/illustrating-the-interplay-between-features-and-models-in-xg)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This conclusion makes sense, as the Random Forest uses Decision Trees to get the xG surface, thus grouping many points  together and creating clearly defined boundaries. Logistic Regression models, on the other hand, fits a smooth surface as the model is predicting based on a logit-linear combination of the Distance and Angle"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's take a look at the difference between the two models by subtracting their surfaces from one another"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note: In the following plots, the red regions represent area of the pitch where the Random Forest has higher xG, while the blug regions represent regions where the Logistic Regression has higher values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={},\n",
    "                         plot_title='Diff Between Random Forest and Logistic Regression')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well, this is really interesting! In front of the net, the models have very similar outputs, with the Logistic Regression being slightly favored with the best angles. However, we see that the Random Forest has a <b> much </b> greater expected goals value when shots are attempted from the corner. At first glance, that looks horribly wrong (and it likely is), but it's important for us to consider <b> why </b> this has happened, as it gives us great insight into what our model is telling us, and why our intuitive definition for expected goals models does not match up with what this specific model is actually doing."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The reason why the Random Forest is doing this is that the model is effectively calculating the empirical probability of shots taken from similar locations on the pitch. When considering locations in front of the goal, there are many shots to build a model off of, and we are able to obtain realistic estimates. However, there are likely very few shots that were attempted from these locations, and thus our model is overfitting. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The (obvious) wrong takeaway from this would be to go to a coach and advocate shooting from these locations. It is likely that the shots that were attempted were done so under exceptional circumstances, which is why the empirical probability is so high. Our model is describing what happened in the past, rather than what will happen in the future. More on this in the  conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the shots in the previous section, we developed a theory that the two models were handling smart passes differently. So, here's the plot!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={'prev_smart_pass': 1},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={'prev_smart_pass': 1},\n",
    "                         plot_title='Difference Between Models After A Smart Pass')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This plot confirms our theory, as the shots that occur after smart passes have a much high probability right around the 6 yard box (where many of the above shots were taken)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This result is likely the result of an error in our feature selection in Notebook 2. Since we did not interact smart pass with any other variables in the Logistic regression model, it adds the same value on the logit scale to every shot with that tag. Thus, these high probability shots (where the value of the linear predictor is around 0, signifying an xG probability of around .5) see big boosts in their probability due to the shape of the logistic curve. In retrospect, we should have interacted smart pass with both distance  and angle, in order to effecitvely fit a  different surface for these events"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In fact, let's compare our logistic regression models with and without the smart pass tagger"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEICAYAAABbOlNNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO2deZgdVZn/P2/fXtJZSEKCIAmEJSCEVQQUxEFFMThqVGBYdMQZRkRl5ueoCI4Og4wLOArjwow4ggrKoow6cYyiyAwoCBIQZaIgYZMEWbKns/X2/v6oul2nqm/drtt9b/e9yffzPP30qapT55xabr3nnO95zzF3RwghhCjTNtEFEEII0VzIMAghhEghwyCEECKFDIMQQogUMgxCCCFStE90AYQQYntjD2v3rRQf8bmKwVvcfWEDi1QTMgxCCFFntuKczJTC8a9i4+wGFqdmZBiEEKLOGK3dTy/DIIQQDaDNrHjkJvMzlmEQQog6oxaDEEKIYbTV0GBQi0EIIXYA1GIQQggxhGG1aQxNhgyDEEI0ALUYhBBCDGFAe+s2GGQYhBCi7hiYupKEEEKU0XBVIYQQw6hpuGqTIcMghBANQC0GIYQQQxg1TonRZMgwCCFEA1CLQQghxBBRi2GiSzF6ZBiEEKIBtHKLYVzKbmZfNrN/DLbfY2bPmlmPmc0ys5eb2SPx9pvHo0w7Amb2dTP7RMG4T5jZaxpQhk+Y2Soze6beaQvRzLRhhf+ajTEbhviDssXMNprZOjO7y8zONbOhtN39XHf/5zh+B3A5cKK7T3X31cAlwJfi7e9XyaPHzNaa2Q/NbI+C5Xulma0Y63U2EjN7p5m5mV2R2b8o3v/1CSpaIcxsbzMbNLN/z+zfE/ggsMDdd4uv8xd1zrvTzD5nZivi9+MJM/vXeuYxQv4jGt/4GW6Ky7fSzC43s9I4lW+Gmf27mT1jZpvN7EEz+6sxpvlCM7vazP4U/+4fMrOPm1nxJctGl+/FZvbNRuZRL8pdSUX/mo16tRje6O7TgHnApcAFwNU5cXcFJgHLgn3zMtt5eUwFXgg8C3xxTCVuPh4F/sLMwu69s4A/TFB5auEdwFrgNDPrCvbvCax29+fqkUnm3pT5CHAkcDQwDXglcH898itQnlo+7ofF7+8JwJnAuxpTqgQz6wRuJfp9HQNMB84HLjWzD4wyzZ2BXwLdwDHx7/61wAxg33qUu1HkvD8No62Gv2ajrmVy9/Xuvhg4DTjLzA6GpFZlZvsDD8fR15nZbWb2KLAP8IO4RtVVOfWhPLYCNwMLyvvMrMvMPmtmf4y7qL5sZt1xDeZHwO5x2j1mtnvc+pgdn/tRM+s3s53i7X8u1zjz0g3yfYOZPRC0lA4Njj1hZh8ys9+a2Xozu8nMJlW5tGeAB4HXxefvDBwLLA4jmdmbzGxZnOf/mtmBwbEXm9n9cS3uJiIDHJ6bW95MvKPNbKmZbYiv+/K8QpuZERmGjwF9wBvj/a8Bfhrc+5uALwPHxNvrRrrH5daemV1gUVfU1yoU4Sjge+7+tEc84e7XBuV7wszOj5/Dprimu6uZ/Si+T7ea2cwg/nfi2vV6M7vDzA4Kjn09rn0vMbNNwNnA24APx9f0g7z7VMbdHwJ+DhxsZvvGv4HVFnW3fcvMZgT5XWBRC2OjmT1sZifU+Hz+ksg4n+ruj7t7n7v/GPg74JLgna/lXf0AsBF4u7s/EV/TU+7+/9z9t3F6x5rZvXFa95rZsZnn8Zpge6gVYGZ7WdS6Oit+H1aZ2UfjYwuBfyCqfPSY2W/i/dMtab2stOg7U4qPvdPM7jSzK8xsNXDxSM+nXlgNrYXtucWQwt1/BawAXpHZ/weg/EOb4e6vdvd9gT8StwjcfVu1tM1sMpHhuTvYfSmwP3A4MB+YA1zk7puAk4Cn47SnuvvTwL3A8fG5xwNPAi8Ptm+vlm5cjhcD1wDvBmYBVwGLLW3Y/gJYCOwNHAq8s9q1AdcSfWQBTgf+Cxi6HxYZ1huA9wO7AEuIDGqnRbXD7wPXATsD3wFODs4tUt4ynwc+7+47EdUCv12lzMcBc4Eb43hnAbj7raTv/WnAucAv4+3yBzD3HsfsFl/PPOCcCvnfDXzAzN5rZofEhirLyUS12v2JDNePiD4yuxD9Bv4uiPsjYD/gBUQtj29l0joT+CRR6+Ta+Phn4mt6Y95NKmNmC4h+F78m6nH4NLA7cCCwB/HHy8xeBJwHHBXXyl8HPBEnU/T5vBb4Ufw7CPlPokrDMcG+ou/qa4DvuvtgzvXtDPwQ+ALRe3Y58EMzm5WTXiWOA15E1Lq6yMwOjA3ap4Cb4nt9WBz360A/0bvzYuBE4G+CtF4KPEbUU/HJGsowZnZojaEKTxP9oOvF9+Na5nqiF/5fYKjGeg7w9+6+xt03Er1Ap1dJ63bgeIualocSvcTHx7Wko4A7CqR7DnCVu9/j7gPu/g2ij/jLgny+ENdk1wA/IPr4VeN7wCvNbDqRgbg2c/w04Ifu/lN37wM+S9SkPzbOtwP417hmeDORASxTpLxl+oD5Zjbb3Xvc/e4KccqcRfTxWQtcDyw0sxeMcJ1A4Wc3CPyTu29z9y0Vkvk0cBlRzX0psNLMzsrE+aK7P+vuK4lq6/e4+6/j1uf3iD4oALj7Ne6+Ma6gXAwcFj+PMv/l7ne6+2B8flHuN7O1RO/BV4Gvufvy+Fluc/fniT6i5QrLANAFLDCzjrgl9Gh8rOjzmQ38KbvT3fuBVfHxMkXf1VmV0gz4c+ARd7/O3fvd/QbgIeKWZEE+7u5b3P03wG+AwypFMrNdgdcD73f3TXGX5RWk35+n3f2LcVkqvT8NQy2GyswB1tQxvTfHtcxJRDWp281sN6Ja32TgPou6SNYBP47353E7UV/0EUTdNz8l+kG+DFgeC+IjpTsP+GD5WHx8D6LaX5lwJM5mYGq1C4xf3B8SdcvMcvc7M1F2J2rdlOMPAk8R3evdgZXuHi4S+GQQLlLeMmcT1a4firsC3lCpvHGXz6nEtWp3/yVR6+/MatcZUOTZPV/tAxwbuSvd/eVE/dyfBK6xoIuNSJMqs6XC9tT4ekpmdqmZPWpmG0hq6OEH9KmC15blCHef6e77uvvH3H0w7tK6Me4C2QB8s5yXuy8nahleDDwXxys/q0LPh+jj/8LszrhCNDs+Xqbou7q6UpoBqXc05kmid7QoRcsyj6gy9Kfg/bmKqLVXZrTPa0yUJ9GTxhBgZkcRvQh1HYECQx+C7xLVqI4jerm3AAe5+4z4b3os9EHl1VTvImqqvgW43d1/R9QX+3qSbqSR0n0K+GRwbIa7T45rSGPhWqKRPJVGXzxN9GMAhmrcewAriWpxczJdKXsG4cLldfdH3P0Moh/YZcDNVnnEyVuAnYB/s6hf/hmi556tsQ8lndke6R5XOieXuJZ5JZEQvmCk+BU4E1hE1F0yHdgr3h/e02x5xrJa76fi8w+Ju4XeHubl7te7+3FEz9yJnkUtz+dW4KQKx04mai1WawnmcSvwFgtGHWZIvaMxexK9owCbiCoDZXarIe/svX6K6DpmB+/PTu5+UJVzxg21GGLMbKe49nIj8E13f7Ce6cd5mJktAmYCv49rzf8BXFHuwjCzOWb2uviUZ4FZYXeAu28G7gPeR2II7iLqA789jjNSuv8BnGtmL43LNMXM/tzMpo3xEm8n6iqrNOrq28Cfm9kJFg37/SDRD+MuopEi/cDfmVmHmb2VaKROmcLlNbO3m9ku8T1YF++u1Kd8FpFucQhR18PhRFrNYWZ2SIX4zwJzYz2kyD0eETN7v0UidbeZtcfdSNOI+vBrZRrR/VxN9PH6VIFzniUaPDEapgE9wHozm0M0YgiINAYze3WsAW0lMqCD8bGiz+c6Iq3vOxYJux3xvf0CcLG7rx9FmS8nqgx8w8zmxeWZY9EQ3EOJdK/9zezM+HmcRmSk/zs+/wHg9LgsRwKn1JD3s8BeZaPk7n8CfgJ8Lv72tFkk6B9fNZVxwGrQF7ZnjeEHZraRyIJ/lOjlGdNY6Zw8eoANRN0FZ7l7eYjrBcBy4O64SX4rUYugPArkBuCxuLlZbo7fTtQM/VWwPQ24I8izWrpLiYYcfomohrqckcXlEfGIn8V9vdljDxPVKr9IVNt+I5Fo3+vuvcBb4zKsIdIjvhucW0t5FwLL4vv9eeD0bP9s/CE7gUjTeCb4u4+oO6hSq+E2omHJz5hZuRsj9x4XZDPwOaLuh1VExv5kd3+shjTKXEvU7bES+B3FatRXE+kA68xsmA/OCHycqDtzPVEX4neDY11Ewvwqomt7AdHQXCjwfABineQ1RL/Le4h+O5cDH3X3f6mxrOU01xBpWn3APfHv/mfxNZS7Yd9AVGlZDXwYeIO7l5/3PxIJ5mvj67++huy/E/9fbWblIcnvADqJntdaohGL1bq6xo1WbjFYuktaCCHEWNmr1OEXTZkxcsSYszeuus/dj2xgkWqiGXUPIYRoaRrh+WxmCy3yZ1luZhdWON5lkQ/KcjO7x8z2ivd3mNk3LPJ6/72ZfSR7bhYZBiGEqDNm0G5W+G/k9KwEXEnkG7QAOMMin5iQs4G17j6faNjuZfH+U4Eudz8EeAnw7rLRyEOGQQghGkCdh6seTaThPBbriTcSjaALWQR8Iw7fDJwQj1J0YEo8TLkb6CXSm6qWXQghRJ2xGv6A2RZNc1L+y3r5zyHtk7GC4b4hQ3FiJ8b1RA6JNxMNE/4TkZ/RZysNbgmpOqmUr16RKNNmvOpN0ciy/1l8cxCrSjOoyNJ2rbj8Xa2CfeH4BeLlpTXWPKqdXyjpMQ5iKFL+Mb8rRTtza82/cCdxjemOIp/R3KNW+w0WeVeqxHnVm6KZYtLfMbCd59TtRkQaQ03JrWqg+Hw0kd/X7kTD/H9uZrdWG7lXfbbB0fwQtldjMJGMyhjUeE7V39rYfoijilfruYXeu6L55EXMyb9qusHBsPxWYL+oTN69y4szLJ4F/xs3KrPOT3IlkTNrmbkkToPZOCvibqPpREOGzwR+7NE0Os+Z2Z1EMxLnGgZ1JQkhRAOosStpJO4F9rNo7ZNOovmgFmfiLCbxHzoFuC2eIuePwKsBYi/4lxHNX5WLlvYUQogGUM8Wg7v3m9l5wC1ACbjG3ZeZ2SXAUo+WO7gauM7MlhM5uZYnE7wS+JqZLYuL9TWPp0jPQ4ZBCCEagNW5W9DdlxBNORLuuygIbyUampo9r6fS/mrIMAghRJ2poYuoKanBMASXWtQSjtcokppplmlAqo0EqlVwHk1ajSpLwXTrSUVxsUqcau9mrvZsleNUTTenXKMSogtc47jTyHKMYaRa9t2sFM+soa9nKwu4ajEIIUQDaOUBZjIMQgjRAKxpWnW1I8MghBB1ZgfSGIQQQhRlOzYModjWKMF5Im5fSjlsTBbjss5FQcE4N1oYZxTi86im12jQfbGcZ5orTo4mj/D8gt7ReeVq1Hs/Ll7U4/WbrdEDPXVqNc/nKvHqSDMuwFMUtRiEEKLumDQGIYQQCdIYhBBCpDENVxVCCJGhhe3CaKfdrjFOFLFgvO2dsXoYFxR8i3g1e574XFTUHoUXdl5atVJUXMwTia3KwIoiXtRFvKOjHTn55AwOqDal9HgL2U3LKETpCai+t7Xwc1GLQQgh6ow0BiGEEMPQcFUhhBApNFxVCCHEENGazxNditFT3DBUE+vSEUdfmh2Nwh7CRcTjYQdHzicvj2FCdq1lGcVa1LUyLJkaReZUeTPnps6plmel/DLbnpNArUI0aG3oEal2Tyo8vAbfw1Z+QmoxCCFEA5BhEEIIkUIagxBCiBSt3NsnwyCEEHXG2J6X9ixk8lrYLAJ1nYJ7PDyBcwXnKoJxEZG5msBcq7dzYSE65/xcRjP1eyg4FxCoq52Tl261cuVO1V1kavAC6xZn0yqyrvSo2J5+59trjvVDLQYhhGgA1sJ9STIMQgjRAFrXLMgwCCFE3dFcSUIIIdKY7YhdSWO84EbdsHFZZ3mMjEaMLSI4F/VWLiI4Dw4WLEtBz+lUsUaxTnSZKu9N6keY6z1cREgmLRK3BWNL8opYdPxCISE674QMreYF3ZS/eaNha76zo0yJIYQQojDWwpZBhkEIIeqM0RoNuTxkGIQQot5ozWchhBBZdhDxeRQXOd43ptp6ueNOjV7M2XPG7GGcF28wZ39OnGFpexAtL62seJ2TVrV4I2HpCQc8710L4wVxLCUqV5l2ezBnf+r30FZ5NxQTrHO9laukVasX9EQwHr//ol7j+QnUrSjDUm5du6AWgxBCNIIdpMUghBCiCBKfhRBCpDEoabiqEEKIhO3a87nAhbXwxRem8NrMY4hT9fwi6VabKruASJzydk6nlSsyDw6MnEfWi7pB4nPqPWyrLDiH57iXKschK0wH5Q33p2bbH8zZT75XdK6nNrVTxAu6WpxW86IuwgRfhzH8FW0l1GIQQoh6YxKfhRBCZGhhuyDDIIQQjUAtBiGEECla2C6M0jC0whVXE9gmksKCMZXjFZ3quoiwnBKVB4NgRhTOE5kHBkaOU7VceVNwB3HyFLzsUMAcD+fU/rZAcC7l5xEK0ykhOk9jTk3TPaygQT6pTHL2V5nDu4iQPdE0U1lyicto1rDvggFtLXEvKqMWgxBC1BtNoieEECJLK2sMLTzSVgghmhez4n/F0rOFZvawmS03swsrHO8ys5vi4/eY2V7BsUPN7JdmtszMHjSzSdXykmEQQog6U54rqV6GwcxKwJXAScAC4AwzW5CJdjaw1t3nA1cAl8XntgPfBM5194OAVwJ91fIr3pVkxoRP49v01FPIKjC9djUv4hxhOVdkHswRkiFfZA7DRYRoMms+FxGiQ0LBOSMYWyHBOQgPlirHgZQwnfKQDuJZap3mHIEaMlWvlGJNRYqu+VyIJpqCu1lpVHePWb2X9jwaWO7uj0XJ243AIuB3QZxFwMVx+GbgSxb9ME4EfuvuvwFw99UjZaYWgxBCNIAaWwyzzWxp8HdOJrk5wFPB9op4X8U47t4PrAdmAfsDbma3mNn9Zvbhkcou8VkIIRpAjcNVV7n7kQ0qSjtwHHAUsBn4mZnd5+4/yztBLQYhhKgz9dYYgJXAHsH23HhfxTixrjAdWE3UurjD3Ve5+2ZgCXBEtcxkGIQQogGYWeG/AtwL7Gdme5tZJ3A6sDgTZzFwVhw+BbjNI1HvFuAQM5scG4zjSWsTw6jelTRWYWas5zeLt/JoyF2PuZq3cpG0CqzZPCyfAoLzQH8QzorPfZWPBfu9iBCdKUvV6bkrkTedNuAFBGcbCPd3VIwTJRYeq/xQUrJuWL2q6nEfhnPcqPOm487uqOIgnVuWVmMs5Z/ob0edHdzcvd/MziP6yJeAa9x9mZldAix198XA1cB1ZrYcWENkPHD3tWZ2OZFxcWCJu/+wWn7SGIQQogHU28HN3ZcQdQOF+y4KwluBU3PO/SbRkNVCyDAIIUQDaOUGmwyDEELUmUh8bl3LIMMghBD1xnb0pT0baRWLrF+7XVGjYJ3j0QykxNyU4ByKzKEw3B8KzBlv+f7kHA+P9YeCdY5AnRWf87yix+j5nBKcc8RnDwXn9qQcNhjsz5bLM8eG9ifl9eBnZFkP49yPQyg45zzfYa9/69ZAc2mg5/HEYlipdS2DWgxCCNEIJtw4jR4ZBiGEqDcmjUEIIUSW+k6iN67IMAghRN1p7SXctt81nxvJWMTv3HWdq8UrIERnPYe9wJTafduScCgw9/em08oTpvuCeKGoHcavNoV3GC5yT8P3LuutXMoRn9tDL+awXJ1J1qWsQJ6cY+05Yn9HV+VyZX8bKQfnMF6B5zvMizpno9bfY/Ze72i/5/LlNnLNZ6Pe026PK2oxCCFEI2hhgyvDIIQQDUAtBiFGwfpNW/jj82t4atVanlu7gd7+fvoGBjCMjvYSU7o62X3WDPbcZWfmzJpOV2fnyIkK0SyoxSBEPu7O8qef465lj3DXske49w9P8Pgzqxh0Z89dZjJ39kxeOHM6XR3ttLe14Tj9A4Ns3LKNFavXsmLVWp5es56dp05h/zm7cswB+3Dswftx7IH7svO0KRN9eUIMx2w7HpWUtXjNcp2pqYkLikejOacQOWnlTrdcLaki3s7FPJ89NY12jjCcJzj3ZcTnUKROhYN4/WlReltfHz++73fcdMdS/nfZI3S2t3Psi/bm2P3n8a7jj2TfXWcxo7szGeudJ6THU20PDA7yzPoelq14lrsefoIv/Oct/OWjf2TuzjM46YgDOeMVR3L43nOi9NqD17ojaGWEQnR4T0IhOVOW3FWTc+Y7yD5pKwVlcascTgnJVaZSz62B5gxosJw86vlDHk2tuJlq0g0si/wYhAAGBwe54/8e4Ybbfsn37n6Ag/fcnTOOO4JL//JN7LnLzChStekyqlBqa2POztOZs/N0Tjx0f2gr0T8wwG+efJrv/epBTv3sNXR3dnDGK17CGa88mr13nV3nqxOiRrbbFoMQBejt6+f6W+/kc//5E0ptxtuOP4r7Lv8Ie+wyMz2fUp1pL5V4yT578JJ99uCfz3wDdz38ODf+/H6OOf+zvPzAfTj/ra/lZQe/qGH5C5FLeW3PFkWGQYyavv5+rv3pnXzy+h+w/5wX8K/nns6rDzsA6+8b+eQ6Y2a8/IB9ePkB+3DpX72Fr936S9722a+x39xd+fjbF/HSA/YZ9zKJHZsde3ZVsUNyx29+z3s+91V2nz2Tb33kXI7Zf95EF2mIKZO6OO8Nr+Sc1x3H12+/l9M+fRXHLtiXK977dnadOX2iiyd2FNRiEMAoRO3sms9FvJ0rr5nsWc/nvCm1A8E4JTj3bq0YJ9pOjq1bvZYLr7mZH937IF941yksetlh0YGejZXPz3Yl5U3bXeuaz+2ZVzc1pXZ0rBM45xVH8PajD+aSm3/C4e/+GJ9+xyLOevXLsK7uJH52yvIcgb+QfFtlLWoLq5BhnrkDI7Lic40CcmoK79b9SLUkZi3tx9DCjR0x3nz/F0s59D0X0V5q47df/nhiFJqcyV2dXPq2N7Dkn97Hvy+5gxMv+iKP/en5iS6W2N4xK/7XZMgwiBEZGBjkw1fdwPlfvp4bLnw3X3rf25k+ZfJEF6tmXrzPHtz5mQ9x0ksO4rgPfoqf3L9soosktmfarPhfk6GuJFGVtRt7OPOfLmdgcJC7r/w4sya1tvdxe6nEB958AkceuB9vu+wq/v6tJ/L3p7+xpceci+bDWnw9BrUYRC6/e+IpjjnnIxw4b3eWXHo+s6ZPm+gi1Y0/O2R/7rz8H7jhf+7hnZdexZZtvSOfJEQtqMUwgTTMo7lBpLTFbHmLeDvnCNFVp7cuMG12KD73buPBx59i4QWf4ZN/fSrvfNVR0Bt7PG/dkoo3RLh/W76Q7X15a0NXFtVTzzcUnzPTbltHID6H3s5dk4L8Qg/wfvac2sXtl5zHO6+8kZM/9jm++7H3MqmzI73+dEokDtdpDssVhjPTgYfnhOX3vPc2xwt6WDyrGGxaWrj2PDqaUzsoiloMYhi/e3IlJ134GS5/z9t458I/m+jiNJTJXZ1cf8HfsNPkbk791JfZ1jf+PhhiO8TASm2F/5qN5iuRmFAeenIlCy/4DJ855wxOe9XLJro440J7qcR1HzqbSZ0dnH7pf9Db1zhvbbHjYGaF/5oNGQYxxB+fXcXrPvgJPvHXp3DmCcdOdHHGlY72Et86/28wg7Mu+zKDRfwqhKhGC2sMMgwCgM1bt3HKxz7L3558Eu848RUTXZwJobOjnRsueBdPPb+GT12/eKKLI1qZWnwYmrDFUAfxeawX1ZyC8RF/9hqeX7Wa+fvsPULMIms4503NXTRejdNxQ0aYDrx3c/Zv7NnEM2vWs+Su+1jyi3vTaeWtGT1YOa3ccLUy17rm87Ap4XNE6kLhjGA86Hzxu7fws6UPYqm1pCufb3kCM+QL1qnyV7mudGIVg7lxcncPj7P8scfZZdYs7v/5rVXybxaa70NaiVb2fG79UUkN4vlVq+nZtHmiizFuTJvczbTJ3SNH3AHobG/noHlzJroY48qO9K6PG03YEiiKDEMO5ZbC//z3dyvUxseyoA6V92e3c2vmyf7UYjz9VRbXCYaVejgsdduWkcMAWzYF4c0Vw54arhoOYw3yA+grMFdShYV6hoWzcyWFw1UnBUNUu5JFeGxSYPi6J1cOA3QHq8KFcyrlhK0zyK8zs+hPuAhQezKMNrWAT1teq6TK0NdUOC9OgRZKfOxVbzwZUUeMptQOiiLDIIQQDaAZRxsVRYZBCCHqTnOONipKDYahSlN0TFRLq0ZhuiW8oKsJxnllLrDO80DG8zlczzk11XW4tnNOt1LYdQSwKdje3JOkuzncH3Yx5XQrZbY98BcY7Bt5mc+2jqRrxToyr27QZZQKdyddPqHXtaW8rgu+K+EPPRClPQhbf6b7py0oZylvje6cwYHVPONrnYJ7IhhzjbnR19Xg9NViEEIIMYSW9hRCCDEMGQYhhBAJNtynpYWQYRBCiEaw/bYYJvrCwvybVUzOY4zlzfGJ8HB/ytchM/Fbas3n4Fg4DXYYDv0QtmScnULBOVzbuSfZHwrU3pOEBzan/SsGtyTbg9sSMdhD8Tln2m0LxOe2rsBvAWjrTnwESpMDf4GpldefDtfIHvaWhyKz5fhOhGtMp8IZET18LoPJ+R74KFie/8uofn4tIErn0mrlrYI0BiGEEMOQYRBCCJEgjUEIIUQWtRiEEEIMIY2hBWlWD+lC6zwH4YGc6bDJeDvnrfO8LWfN5k2BqAx4uL0xEJ83bEiyWJ/E6V+fpDXQk55EbzAQo7dtS8o8MBBODZ7Et5SzcdI07+pKexi3BYJzaWoyqV17IHCXQhE+FPQzP+DUHDfhRHbhWtKlymtM+0D6J2WB4Jx6XqUanzU0jzbbwh+8caWF71PrdoIJIUTTEmsMRf+KpGi20MweNrPlZnZhheNdZnZTfPweM9src3xPM+sxsw+NlJcMgxBC1BujrobBzErAlcBJwALgDDNbkIl2NrDW3ecDVwCXZY5fDvyoSPFlGIQQohHUd2nPo4Hl7v6Yu/cCNwKLMnEWAd+IwzcDJ1jcL2pmbwYeB5YVyeSJxdcAABFMSURBVEyGQQgh6oxhWFtb4b8CzAGeCrZXxPsqxnH3fmA9MMvMpgIXAB8vWv4WEp9b2Qs6j6LXkTftdugFnUkrFDr7c8TnnNXcUquxQdrDORCc+9dsDMJJnL61iefzpk1pj+zNW5Lt3kB87h+o7N0dCsHtpSTcmRGfJ3cn1zhlU7hiXTj9eJJu6sXP/DA98HC2zkBwDqfzDoXo8P4Gq7QB6eeSt6rfaN6D7YbWFWhHpDbxebaZLQ22v+LuX6lTSS4GrnD3nqKLB7WQYRBCiBah9uGqq9z9yCrHVwJ7BNtz432V4qwws3ZgOrAaeClwipl9BpgBDJrZVnf/Ul5mMgyiaRhw57mBAbqAmaXSiPGFaGrqO1z1XmA/M9ubyACcDpyZibMYOAv4JXAKcJtHze9XJEWyi4GeakYBZBhEE+DufHPjBq7euJ6t7gy4M7+jgwtm7MxhXZNGTkCIpqO+U2K4e7+ZnQfcApSAa9x9mZldAix198XA1cB1ZrYcWENkPEaFDIOYcC5ft5YbN25ka9AH//u+Ps5d9Rxf3WVXDursqnK2EE1KnR3c3H0JsCSz76IgvBU4dYQ0Li6S1/ZrGLIPpZCHc4OEsDDrquXwyuFcz9hAwBzMTvccbA8EAnAolOZ6PmfWfA7E59DDuW91Ij5vXZ2cs2F9InBv7AnyAzYHYvCWQWft4ADf2rCBdKw4TXc+tWY1l0+fDUB78Ey7t6Sf1eTNyau8bVvilbxTIDhPCkXtQMgutWd+Bh2BV3NKfO4OwqFneTi1dpXnkDdwwHOeezWxOXWOVQyOnVEk1sLevnVFU2IIMXru7t1GCaMv5yP4cH8fW3yQbtPIatFiyDAIMTp63RmsUjM2oN/Zrkc1iu0RTbstxKg5tKOz6vFZbSWmtnDNS+zAtPB726KGYXt0ditIqp86JwwZjSEI5yzt6b3hMp/pGVHDpTr71yXLfvauSfavXZvoFevXJ+F1A+k+903BLKpbBweZCuxX6uDhgT4yi5PSCby5s5vn4llRO4Mf2qRMbWxKsFRnX7BM6EDgYDYrnDS1M3n1s8uEWneiJYT3xfKWRa0yy23h57VD0bofzMK0uMbQum0dsd3wgSk7cUCpgw6gC5hEZBTe0jmZYzs1XFW0IvWfXXU8adEWg9iemGxtnD9lOs8MDLB8oI8OjIPbOyQ4i9amhVsMMgyiaditVGK3UmnYtE9CtCQyDEIIIYZocY1hdIZhrBfcTMtpjgeWszHW2+CVZyQFYDDH+W0wR4hOObulxeeBnuRY/9pEfN6wIXHyCgXn1cESmuv602Lsuv6kXBsDIbo3KH/YYmgLblcoPk8rpbuZZrQn2wPhvVifBDuCOO1dyXWUpqR1jPZp4ayzwX0J71fePR1Mi8qpmWLr+d5bo5zaWoyxfIvMGvgt0nBVIYQQWXa4FoMQQojqyDAIIYQYwoAWHlUnwyCEEHXHoIXXFKluGBrVFMpLd1RCUEEv6DDPMJ+8/WNmnL2zs560ebOwDuR5ROfMugoMbE62t25KBNgNG5Nw6OG8NhCc/9SbFp/DY5uDmU/7w1VKg/vVFtzH9uCWbiilr3fzYPIjDO92KXi+nUF5J09OXv32zenrbU8JzuEsqjn3Lm/W1OyxcWGcf7OFvxFjLNd4f4uaPe0GoxaDEELUG9OoJCGEEFnUYhBCCJFC4rMQQogUajHUiVEtxzkOVHvAnrtRJOERtutE6Eoc3tPQYzfwVvbe9EKbg1sS0XbL1iTepm1JeEMgKj8bCM7P96XF557A27mv0PNN4mwLovd6+l6F4nVYTwu9pacE5Q2vY3JwfZC+fuvPWbYzdR8b6D1bdXuUaWXf52b9gDVruYogjUEIIcQwWtiwyTAIIUQjkMYghBBiCLP0DJAthgyDEEI0ArUYGkTNXsmN8Da24mnV1Ys6RywMX7a8/dljtdZcMmX3YKrs3t4kvCUQXcMptNcG8Tf0pz1/B4J7acE1hiJxWNzBlOCcbGSF6zCfMK0Z7cn+GUFi4XV4fxWv8SK0VRN2Czyv1Dl1rGWOqo+7nrXcGtNq4T75irTw9TS3YRBCiJbE1GIQQggRYEhjEM3Nke+9mOfX97Dv7i+A/sBHIQz39lYOA4Nbkni9gS/A1r6kC2ZbMFnc1qDLJu0RAZ7qlgu6jzwVqSKZtdFSW2G3VEfgejBpc7K/a0tSg5u0KQl3PvdsKq22ZY8nG52dFcOPrlrPLjtNZeml769cWCHUlSSamefXb6Rn67aRI4rC9GzrhQ09E10M0cyoK2kcqKewW2QK7rGSJyh6XpyC+ecImJY51wOvy33n7ArAbf9yIaxbk0Ra/VwS//kgvPLpVFpblifHVj65bij85LotQ+E/BN7DT4be0ZlppztSwnBSxqnBGs5hnFBkDr2m12UE4zDelOC+zOtKXvH9u5Ma/7wZ3UPhOfNmpNLqnv+CobDN2T0J75Lsf/WXbo4CbaX0ByDj7WqFBOccqr4feYMTxkG8Ho88Wh0NVxVCCDEMtRiEEEKkaOHWkAyDEELUHQ1XFUIIEaLhqs1K9qHUcWrkuq5ZnZNurmgZCp2lyuFh25bsaw8eeUcwFDNYuNw60q+FdSR5tuW87L3BENVwuKplnsPsjiSfUBie0Z7sbw+uvT+4p+uCqb2fDIbNAjwTTPUd5t+bMyV2eB3h9UXbwfWHC7qH96tcxvb24s8hDOc90/EQkhvaxdG6H8O6k30XWojt2DAIIcREofUYhBBChBgSn4UQQmSQ+CyEECLB1GIYd8bDCzodKbNdJF7lOYGwnP3DTqksFlrQbxl6N6dEUoBSx/DzOzrTAmpXVxKeNCkJdydewQClqZOCQ8kr0xHUiELBOPRJnpwRq/edlJw/b1KS/8zupLwdnUm6fcH02GuDOZvaM7cunHZ7cyA4h+UKyxteR3h98cEkHN6X8H7l3dPwvkP6uQTPK3yOhYXhVLzUgbwT8tOqNZ68nWtHGoMQQoghpDEIIYRIIwc3IYQQWdRiEEIIkUIthgmksBBd63rQlqQ/GoE7t7aQMwV39ljqpQrk3DDdUuihmxY9LfRwLotgHZ3QFYqpichqk6ckxZqShAHaZybbO83YOBSeuWrrUHh2b1LeR5PdTC6l78PegeB8wD7Th8KdOyd5tAVe0IOBt/ML12xK4jy2PpXuw+FiQsHzmh14Nc/sTO7JTjOS+xBeHwDB9Yf3JbxfQ/e0a1JKfE7dd0g/l/B55U6h3VZ5/7B4DZp2u+bzC8Zv4drzqGjAtNtmthD4PFACvurul2aOdwHXAi8BVgOnufsTZvZa4FKgE+gFznf326rl1bomTQghmhlrK/43UlJmJeBK4CRgAXCGmS3IRDsbWOvu84ErgMvi/auAN7r7IcBZwHUj5SfDIIQQjcCs+N/IHA0sd/fH3L0XuBFYlImzCPhGHL4ZOMHMzN1/7e7l1beWAd1x6yIXGQYhhKg7VtcWAzAHeCrYXhHvqxjH3fuB9cCsTJyTgfvdvepav62vMQghRBOSXW53BGab2dJg+yvu/pU6l+cgou6lE0eKO3bDMBpRaazeymMmR4gufC15HqiVky3mBU3aUzKcLjrc7znTbrdnPG4rThHdmRZQuycn4WmJEExvujLR1pus5zwpEHnnBuEtTyYi8cpgCuzs2syHHvHCJPvD5weZBNc1kJwfeg53DyZpHTpjeSrd/771kaRcwXTeodg9d04iJE/aI6lIte2SqVTNnJmEw/sS3q9yLa+rO7qvZcL7Dunnkjftdp4XdNZztlZv51xReqyCcRMJyc0qahu1jkpa5e5HVjm+Etgj2J4b76sUZ4WZtQPTiURozGwu8D3gHe7+6EiFUVeSEELUnbp3Jd0L7Gdme5tZJ3A6sDgTZzGRuAxwCnCbu7uZzQB+CFzo7ncWyUyGQQghGkGbFf8bgVgzOA+4Bfg98G13X2Zml5jZm+JoVwOzzGw58AHgwnj/ecB84CIzeyD+e0G1/KQxCCFEI6izg5u7LwGWZPZdFIS3AqdWOO8TwCdqycu8Sn+/b1ydOrjHAYfQ07OJww89OEyilvzKKY/inPHlgd/+HwCHH3IwYy5v7j2ukm7eOeH+VDjdl89g0k//wENR//th+++T7r8fCJbH7A/CfYmmEG0nWoJvTY4NbEnCm7f2syJOb1qgg/RnrmPezETjaJsa6B15joo5+wd7tqTSfXJtsr0+dgicZSWmtydlmRzM7FrqDpzSJmV0gY5AFwg1g8B57TePR927h82fl14WtdrSnrUu4Vm1/3w05+TzwIPLgPL7Xj2L5qL2Qj7w2weZOnUKTz30YDqlabPqdsVHLtjPf3XtFwrHLx31+vtG0BjGleItBjN2mV0W6cZ6/wopthPK1JT37xjXj851uq6WbgFP7WpiZPARmjo5Ek3N2lLadSrdlJHJCNk5ubQHTeCp7W30bowMw65dyfm9vWmDtW5jImz3rk9cpLcF+fcFRekIMuwKPnqdmeb3LsE6zesGIkO2a3cH7Z3Jh7ktKBed4cc/c72hYBxOmx2Ep06OjVpbCctbvxkyBoACjMYYFEm3ep5TM97uzUf9rNTUqVOib1mj177eUabEuP/O2xtVjoh6jlYqnFaBmnnVY3kf15w4g4OV90N6JFJQ40+1BsJa/mAS9v6kVg/Ati0Vw741mVaCzRuS8MZgiol1a1NJ+dpVycbq1Un2zzyXZLFiDa+/5yEAbtgzGV79zLObU2k905uU8+kgvHJbcr2r+pLw7GB6jDldSXj3zvTHfLdg+283RuX9zvw9mf7CaUP7u+buPBRu2y3oYp2VHpVkM2cnGzNyRihN3imJPyln2ozMtqVGKIXTmYRTZeSMOoNMX3Te6KVRjEqqucUyAdNgNOvoo2pkn18LIY1BCCHqTQPmShpPZBiEEKIR7ChdSUIIIQrSit1fMVUNQz1VetFYJvpBdQK/MPtfgLlrHnplef/cCSjL3eVyPJCUQ4jxZQcSn4UQQhRke20xCCGEGAW1z5XUVMgwiHoyH5hqcVfOBHI40DPBZRA7NDZ8EsQWQoZB1JPnJ7oAMT00T1nEDkqN0243FTIMom64+4snugxCNA3qShJCCDGEIfFZCCFEiIarCiGEyKIWgxBCiBQalSSEEGIIM7UYhBBCZJDGIIQQIoVaDEIIIdLIMAghhBhCGoMQQogsMgxCCCHSyDAIIYQoY2jNZyGEEFlkGIQQQgwh8VkIIUQWGQYhhBBpZBiEEEKEqMUghBAijQyDEEKIMppdVQghxDBkGIQQQqSRYRBCCBFgajEIIYRIIcMghBAiwVBXkhBCiDRqMQghhBjCkGEQQgiRRYZBCCFEiFoMQgghUrSuXZBhEEKI+qNRSUIIIbKoK0kIIcQQGpUkhBBiGNY20SUYNa1bciGEaFosmXq7yF+RFM0WmtnDZrbczC6scLzLzG6Kj99jZnsFxz4S73/YzF43Ul4yDEII0RCshr8RUjIrAVcCJwELgDPMbEEm2tnAWnefD1wBXBafuwA4HTgIWAj8W5xeLjIMQgjRCOrbYjgaWO7uj7l7L3AjsCgTZxHwjTh8M3CCRVO8LgJudPdt7v44sDxOLxdpDEIIUWfu+/UDt9iUGbNrOGWSmS0Ntr/i7l8JtucATwXbK4CXZtIYiuPu/Wa2HpgV7787c+6caoWRYRBCiDrj7gsnugxjQV1JQgjR/KwE9gi258b7KsYxs3ZgOrC64LkpZBiEEKL5uRfYz8z2NrNOIjF5cSbOYuCsOHwKcJu7e7z/9HjU0t7AfsCvqmWmriQhhGhyYs3gPOAWoARc4+7LzOwSYKm7LwauBq4zs+XAGiLjQRzv28DvgH7gfe4+UC0/iwyKEEIIEaGuJCGEEClkGIQQQqSQYRBCCJFChkEIIUQKGQYhhBApZBiEEEKkkGEQQgiR4v8DDP0URa7g8dQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=logistic_regression, \n",
    "                         params_1={'prev_smart_pass': 1},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={},\n",
    "                         plot_title='Diff Between Models After Smart Pass On Counter')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find these shots are boosted by close to 10% near the goal, but have little impact elsewhere. This makes sense. However, it is important to note what this is doing to the model. If we remember the Dele Alli and Mo Salah shots from above, they both had this tagger. What the tagger is actually doing is providing us with a proxy for information that we do not have access to (goalkeeper/defender positions, body poses etc.). The reason why this tagger boosts the xG model is that when these events occur, it <b> usually </b> stretches the defense and provides the shooter with more space to shoot. However, this isn't always the case (like with the Dele shot). Thus, this tagger is simply a proxy for stuff that tracking data will tell us"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's also look at the difference between the two models with both a smart pass and a counter attack (like the Ayoze Perez goal)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={'prev_smart_pass': 1, 'counter_attack': 1},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={'prev_smart_pass': 1, 'counter_attack': 1},\n",
    "                         plot_title='Diff Between Models After Smart Pass On Counter')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see an even more pronounced version of the effect above"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "On the Wesley Hoedt and Dwight Gayle chances, we developed a theory that the Random Forest placed a higher weighting on shots that came after a cross, so let's investigate that, first with shots after a cross with the strong foot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={'prev_cross': 1},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={'prev_cross': 1},\n",
    "                         plot_title='Diff Between Models After Cross')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And now with headers after a cross"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={'prev_cross': 1, 'strong_foot': 0, 'header': 1},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={'prev_cross': 1, 'strong_foot': 0, 'header': 1},\n",
    "                         plot_title='Diff Between Models With Headed Shot After Cross')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find that these plots support our hypothesis."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, I wanted to take a look at how our two models handle stronger vs weaker footed shots, as that will inform one of our takeaways in the  final section"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=random_forest_xg_model, \n",
    "                         params_1={},\n",
    "                         model_2=random_forest_xg_model,\n",
    "                         params_2={'strong_foot': 0},\n",
    "                         plot_title='Dif Between Strong/Weak Foot Random Forest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate_difference_plot(model_1=logistic_regression, \n",
    "                         params_1={},\n",
    "                         model_2=logistic_regression,\n",
    "                         params_2={'strong_foot': 0},\n",
    "                         plot_title='Dif Between Strong/Weak Foot Logistic')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see a much smoother plot (as expected) for the Logistic than for the Random Forest. We also see a couple weird effects, notably that weaker foot shots have a higher xG on the RF model close to goal from weird angles, and that there's some blue far away from goal on the sidelines for the Logistic model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Final Takeaways"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are a few takeaways to take from this series, and I wanted to highlight them here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The way most  people think about expected  goals models is slightly flawed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Most people think about expected goals models in a similar way to how it's defined by fbref:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "fbref definition: \"xG (or expected goals) is the probability that a shot will result in a goal based on the characteristics of that shot and the events leading up to it\" https://fbref.com/en/expected-goals-model-explained/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, as we explored above, this is not exactly what exepected goals models are representing. Thus, I'd like to propose an alternative definition for this expected goals model:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "New definition: The approximate historical frequency of attempted shots leading to goals from similar locations, with similar characteristics and events leading up to it"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ideally, we would like our model to represent the first definition, as that is where it would have the most impact for analyzing to coaches and analyzing games that use out of sample predictions for the xG (as most do). However, our model is accounting the for historical frequency of events with incomplete information, which leads us to be hesitant  of using this model to draw insights"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A perfect example of why the incomplete information is a problem is in the comparison between strong and weak footed shots depicted above."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our model doesn't take into account position of the player, and intuitively it would make sense that strikers had higher xG chances than defenders (like Hoedt in our example above). It is also intuitive that higher percentage of shots with the weak foot are taken by strikers, compared to the percentage of strong foot shots taken by strikers. Thus, our weak footed shots will have artifically higher xG because the historical frequency of shots (and thus percentage of successful shots) is skewed towards strikers. Thus, we see a much smaller difference in xG between strong and weak footed shots as a result."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus, we need some form of prior knowledge to avoid overfitting our xG models in regions where we have little data to build a model. Fortunately, David Sumpter has some good suggestions for this, most notably adding \"fake data in places where it's unlikely to score.\" Doing this would help get more reliable estimates in those weird spots where the data is overfit."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Our expected goals models are only as good as our data and our additional features are only proxies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As shown by the videos above, our expected goals models suffer from not having complete information about the situation on the field at the time the shot was taken. The following are examples of features that we would like to have, but do not in the dataset used to build  this expected goals model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1.) Position of goalkeeper <br>\n",
    "2.) Position of other player <br> \n",
    "3.) Body pose of the shooting player <br>\n",
    "4.) Ability/position of shooting player <br>\n",
    "and countless others"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We try to account for some of this missing information through the use of <b> proxies </b>. On counter attacks, there is often more space for players to exploit, and thus our expected goals model reflects this by boosting the chances on counter attacks. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This leaves the opportunity for these models to be misinterpreted by taking the output too literally. Thus, we need  to understand our model's outputs and know where we are incorporating information properly"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "N.B. I know the StatsBomb data contains some of this information. However, the public StatsBomb data is heavily biased by Messi shots as most of the data is from matches he played, so I decided to go with the dataset that"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Open sourcing is the way to go"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This was my first time building an expected goals model, either publicly or privately. As a result, I made many mistakes during the process. However, by open sourcing code, it is easy for others to see and critique your work, and many people in the community have reached out to offer feedback as I've rolled out both this series and my pitch control series. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thanks to the Friends of Tracking initiative, the football analytics community has progressed immeasurably in the last 2-3 months, both in size, quality and quantity. A large part of the reason for this is due to open sourcing code. We all have different strengths and weaknesses, and by open sourcing, we help account for other's weaknesses and make it easier for others to publish strong research. On the flip side, you will also benefit from other people's code, and we as a result we all build off of each other instead of everyone reinventing the wheel."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I hope you all enjoyed reading this, and if you have any feedback/suggestions for improvement, please reach out to me on Twitter @andrew_puopolo."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}