overwatch-elo-ratings-pymc3_Binomial.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_uuid": "a3e9c6da6470fe1ec0596ebba0011b76a25158cd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Playoffs.csv', 'playoff_win_loss.csv', 'regular_season.csv', 'regular_season', 'overwatch', 'Playoffs', 'regular_season_win_loss.csv']\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score# This Python 3 environment comes with many helpful analytics libraries installed\n",
    "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n",
    "# For example, here's several helpful packages to load in \n",
    "\n",
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "\n",
    "# Input data files are available in the \"../input/\" directory.\n",
    "# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n",
    "\n",
    "import os\n",
    "print(os.listdir(\"../input\"))\n",
    "from sklearn.metrics import roc_auc_score, accuracy_score\n",
    "# Any results you write to the current directory are saved as output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "_uuid": "5289aabd1ee78b49d3ba09806b6f7f52b85b82c7"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING (theano.tensor.blas): Using NumPy C-API based implementation for BLAS functions.\n"
     ]
    }
   ],
   "source": [
    "import theano\n",
    "\n",
    "def Logistic_Rating(rating):\n",
    "    return 10.0**(rating/400)\n",
    "\n",
    "def Expected(rating1, rating2):\n",
    "    \"\"\"\n",
    "    These should be Logistic Ratings\n",
    "    \"\"\"\n",
    "    log_rating1 = Logistic_Rating(rating1)\n",
    "    log_rating2 = Logistic_Rating(rating2)\n",
    "    return log_rating1/(log_rating1+log_rating2)\n",
    "\n",
    "def Update_Rating(rating1, K, S, E):\n",
    "    \"\"\"\n",
    "    rating1: rating should be the r\n",
    "    K : Scaling factor between 1,99\n",
    "    S : Outcome of game [0,1]\n",
    "    E : Expected Outcome of the game\n",
    "    \"\"\"\n",
    "    return rating1 + K*(S-E)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "_uuid": "8054d7d3ad9b418317f1112b1fe3a4e9a5cf62cb"
   },
   "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>Win</th>\n",
       "      <th>Lose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Florida Mayhem</td>\n",
       "      <td>San Francisco Shock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>Florida Mayhem</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>Florida Mayhem</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>Florida Mayhem</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Los Angeles Valiant</td>\n",
       "      <td>San Francisco Shock</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Win                 Lose\n",
       "0       Florida Mayhem  San Francisco Shock\n",
       "1  San Francisco Shock       Florida Mayhem\n",
       "2  San Francisco Shock       Florida Mayhem\n",
       "3  San Francisco Shock       Florida Mayhem\n",
       "4  Los Angeles Valiant  San Francisco Shock"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regular_season = pd.read_csv('../input/regular_season_win_loss.csv')\n",
    "regular_season.head()                             "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "_uuid": "bc10122e006b01e728553655dac429ef7d5b790e"
   },
   "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>index</th>\n",
       "      <th>team</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Florida Mayhem</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>San Francisco Shock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>Los Angeles Valiant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>Seoul Dynasty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>Los Angeles Gladiators</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>London Spitfire</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>Dallas Fuel</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>Houston Outlaws</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>New York Excelsior</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>Boston Uprising</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>Shanghai Dragons</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>11</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    index                    team\n",
       "0       0          Florida Mayhem\n",
       "1       1     San Francisco Shock\n",
       "2       2     Los Angeles Valiant\n",
       "3       3           Seoul Dynasty\n",
       "4       4  Los Angeles Gladiators\n",
       "5       5         London Spitfire\n",
       "6       6             Dallas Fuel\n",
       "7       7         Houston Outlaws\n",
       "8       8      New York Excelsior\n",
       "9       9         Boston Uprising\n",
       "10     10        Shanghai Dragons\n",
       "11     11     Philadelphia Fusion"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "teams = regular_season.Win.unique()\n",
    "n_teams = len(teams)\n",
    "teams = pd.DataFrame({\"index\":range(len(teams)), \"team\": teams})\n",
    "teams.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "_uuid": "a95fce148f5e61257e5f446a2400cf8743c430f3"
   },
   "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>Win</th>\n",
       "      <th>Success</th>\n",
       "      <th>win_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Boston Uprising</td>\n",
       "      <td>98.0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Dallas Fuel</td>\n",
       "      <td>65.0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Florida Mayhem</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Houston Outlaws</td>\n",
       "      <td>93.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>108.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Los Angeles Gladiators</td>\n",
       "      <td>99.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Los Angeles Valiant</td>\n",
       "      <td>108.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>New York Excelsior</td>\n",
       "      <td>138.0</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "      <td>94.0</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>77.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Seoul Dynasty</td>\n",
       "      <td>97.0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Shanghai Dragons</td>\n",
       "      <td>23.0</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Win  Success  win_index\n",
       "0          Boston Uprising     98.0          9\n",
       "1              Dallas Fuel     65.0          6\n",
       "2           Florida Mayhem     43.0          0\n",
       "3          Houston Outlaws     93.0          7\n",
       "4          London Spitfire    108.0          5\n",
       "5   Los Angeles Gladiators     99.0          4\n",
       "6      Los Angeles Valiant    108.0          2\n",
       "7       New York Excelsior    138.0          8\n",
       "8      Philadelphia Fusion     94.0         11\n",
       "9      San Francisco Shock     77.0          1\n",
       "10           Seoul Dynasty     97.0          3\n",
       "11        Shanghai Dragons     23.0         10"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regular_season['Success']=1.0\n",
    "wins=regular_season.groupby(\"Win\").agg({\"Success\":'sum'})\n",
    "wins = pd.merge(wins,\\\n",
    "                    teams.rename(index=str,\\\n",
    "                    columns={\"team\": \"Win\", \"index\": \"win_index\"}),\n",
    "                    on = ['Win'] )\n",
    "wins.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "_uuid": "2d26cb067495667d450b8e3eeee394624778e4b6"
   },
   "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>Success</th>\n",
       "      <th>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Success  win_index  lose_index\n",
       "0      1.0          0           1\n",
       "1      1.0          0           1\n",
       "2      1.0          0           1\n",
       "3      1.0          0           1\n",
       "4      1.0          2           1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regular_season = pd.merge(regular_season,\\\n",
    "                    teams.rename(index=str,\\\n",
    "                    columns={\"team\": \"Win\", \"index\": \"win_index\"}),\n",
    "                    on = ['Win'] )\n",
    "\n",
    "regular_season = pd.merge(regular_season,\\\n",
    "                    teams.rename(index=str,\\\n",
    "                    columns={\"team\": \"Lose\", \"index\": \"lose_index\"}),\n",
    "                    on = ['Lose'] )\n",
    "\n",
    "regular_season = regular_season.drop(['Win','Lose'], axis=1)\n",
    "\n",
    "regular_season.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "_uuid": "8c060db50907b4acf755d14f8ce90029183121f6"
   },
   "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>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "      <th>Success</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   win_index  lose_index  Success\n",
       "0          0           1      1.0"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def build_training_df(df,add_losses=False):\n",
    "    columns = ['win_index','lose_index','Success']\n",
    "    df['Success'] = 1.0\n",
    "    if add_losses:\n",
    "            inverse_season = df.copy()\n",
    "            inverse_season['tmp_lose_index'] = inverse_season['win_index']\n",
    "            inverse_season['tmp_win_index'] = inverse_season['lose_index']\n",
    "\n",
    "            inverse_season = inverse_season.drop(['win_index','lose_index'], axis=1)\n",
    "            inverse_season = inverse_season.rename(index=str,\\\n",
    "                                columns={\"tmp_win_index\": \"win_index\", \"tmp_lose_index\": \"lose_index\"})\n",
    "            inverse_season['Success'] = 0.0\n",
    "            df=pd.concat([df[columns],inverse_season[columns]],axis=0)\n",
    "    return df[columns]\n",
    "    \n",
    "X = build_training_df(regular_season)\n",
    "X.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "_uuid": "c14ffba49ab5bafd0a914476d193a08bfed11ec6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MAYBE FOR TRAINING, ONLY HAVE TRUTH VALUES\n"
     ]
    }
   ],
   "source": [
    "import theano.tensor as T\n",
    "print (\"MAYBE FOR TRAINING, ONLY HAVE TRUTH VALUES\")\n",
    "winner_index = theano.shared(np.array(X['win_index']))\n",
    "loser_index = theano.shared(np.array(X['lose_index']))\n",
    "model_output = theano.shared(np.array(X['Success']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "_uuid": "75be142baf468e99200ab4abcd89b28b92b6bb8c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on PyMC3 v3.6\n"
     ]
    }
   ],
   "source": [
    "import pymc3 as pm\n",
    "\n",
    "print('Running on PyMC3 v{}'.format(pm.__version__))\n",
    "\n",
    "elo = pm.Model()\n",
    "\n",
    "with elo:\n",
    "    sigma = pm.Gamma('noise', alpha=10.0, beta=1)\n",
    "    #sigma = pm.Gamma('noise_worse', alpha=10.0, beta=2)\n",
    "    #sigma = pm.HalfCauchy('noise', beta=3.0)\n",
    "    elo_team = pm.Normal(\"elo_team\", mu = 2000.0, sd=sigma, shape=n_teams)\n",
    "\n",
    "    log_rating2 = 10.0**(elo_team[loser_index]/400.0)\n",
    "    log_rating1 = 10.0**(elo_team[winner_index]/400.0)\n",
    "    E = log_rating1/(log_rating1+log_rating2) # Expected value\n",
    "    \n",
    "    out = pm.Bernoulli('out', \n",
    "                   E,\n",
    "                   observed=model_output)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "_uuid": "d80be5b686b12b96c82c7bfdf1d3443009421ba0"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Auto-assigning NUTS sampler...\n",
      "Initializing NUTS using jitter+adapt_diag...\n",
      "Multiprocess sampling (2 chains in 2 jobs)\n",
      "NUTS: [elo_team, noise]\n",
      "Sampling 2 chains: 100%|██████████| 17000/17000 [00:47<00:00, 359.84draws/s]\n"
     ]
    }
   ],
   "source": [
    "use_advi = False\n",
    "\n",
    "with elo:\n",
    "    if use_advi:\n",
    "        inference = pm.ADVI()\n",
    "        advi_approx = pm.fit(n=100000, method=inference)\n",
    "    else:\n",
    "        trace = pm.sample(8000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "_uuid": "1bc2b2b65ca824173140c03c3b835623c825eba1"
   },
   "outputs": [],
   "source": [
    "#Now we sample from our approximation in order to get a similar trace\n",
    "if use_advi:\n",
    "    trace = advi_approx.sample(10000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "_uuid": "eecd5fd8d811c6d95f9e3d3fbb3d580442dd41b5"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pm.traceplot(trace);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "_uuid": "2c5eb969958e7751b74453878966bd728720c730"
   },
   "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>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>mc_error</th>\n",
       "      <th>hpd_2.5</th>\n",
       "      <th>hpd_97.5</th>\n",
       "      <th>n_eff</th>\n",
       "      <th>Rhat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>elo_team__0</th>\n",
       "      <td>1886.902479</td>\n",
       "      <td>26.804245</td>\n",
       "      <td>0.184216</td>\n",
       "      <td>1831.736686</td>\n",
       "      <td>1937.092529</td>\n",
       "      <td>20662.672671</td>\n",
       "      <td>1.000014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__1</th>\n",
       "      <td>1984.699994</td>\n",
       "      <td>24.998467</td>\n",
       "      <td>0.158736</td>\n",
       "      <td>1937.191351</td>\n",
       "      <td>2034.488071</td>\n",
       "      <td>22156.136093</td>\n",
       "      <td>0.999940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__2</th>\n",
       "      <td>2057.367130</td>\n",
       "      <td>24.865345</td>\n",
       "      <td>0.172443</td>\n",
       "      <td>2009.630549</td>\n",
       "      <td>2106.704445</td>\n",
       "      <td>21122.568430</td>\n",
       "      <td>0.999948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__3</th>\n",
       "      <td>2033.974774</td>\n",
       "      <td>24.876458</td>\n",
       "      <td>0.159948</td>\n",
       "      <td>1984.738341</td>\n",
       "      <td>2082.409690</td>\n",
       "      <td>21113.004346</td>\n",
       "      <td>0.999944</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__4</th>\n",
       "      <td>2034.381207</td>\n",
       "      <td>24.340789</td>\n",
       "      <td>0.165470</td>\n",
       "      <td>1985.690940</td>\n",
       "      <td>2081.029477</td>\n",
       "      <td>20934.261489</td>\n",
       "      <td>0.999939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__5</th>\n",
       "      <td>2050.433952</td>\n",
       "      <td>24.615052</td>\n",
       "      <td>0.179117</td>\n",
       "      <td>2003.226981</td>\n",
       "      <td>2099.493394</td>\n",
       "      <td>18405.593489</td>\n",
       "      <td>0.999988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__6</th>\n",
       "      <td>1947.675731</td>\n",
       "      <td>25.370715</td>\n",
       "      <td>0.172229</td>\n",
       "      <td>1897.679568</td>\n",
       "      <td>1997.294231</td>\n",
       "      <td>22813.355967</td>\n",
       "      <td>0.999987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__7</th>\n",
       "      <td>2022.237612</td>\n",
       "      <td>24.625649</td>\n",
       "      <td>0.177086</td>\n",
       "      <td>1976.524021</td>\n",
       "      <td>2073.036065</td>\n",
       "      <td>20147.598004</td>\n",
       "      <td>0.999961</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__8</th>\n",
       "      <td>2116.452494</td>\n",
       "      <td>26.094669</td>\n",
       "      <td>0.193487</td>\n",
       "      <td>2067.709716</td>\n",
       "      <td>2169.737281</td>\n",
       "      <td>17927.092693</td>\n",
       "      <td>0.999938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__9</th>\n",
       "      <td>2028.985198</td>\n",
       "      <td>24.840915</td>\n",
       "      <td>0.173867</td>\n",
       "      <td>1981.126049</td>\n",
       "      <td>2077.813210</td>\n",
       "      <td>20253.032906</td>\n",
       "      <td>1.000074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__10</th>\n",
       "      <td>1818.665401</td>\n",
       "      <td>30.575416</td>\n",
       "      <td>0.213381</td>\n",
       "      <td>1757.389319</td>\n",
       "      <td>1877.362345</td>\n",
       "      <td>18641.798768</td>\n",
       "      <td>1.000356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>elo_team__11</th>\n",
       "      <td>2021.611648</td>\n",
       "      <td>24.478701</td>\n",
       "      <td>0.197998</td>\n",
       "      <td>1973.838660</td>\n",
       "      <td>2069.686402</td>\n",
       "      <td>20191.938716</td>\n",
       "      <td>0.999938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>noise</th>\n",
       "      <td>42.486276</td>\n",
       "      <td>4.965166</td>\n",
       "      <td>0.043380</td>\n",
       "      <td>33.015826</td>\n",
       "      <td>52.288570</td>\n",
       "      <td>14463.611948</td>\n",
       "      <td>1.000298</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     mean         sd    ...            n_eff      Rhat\n",
       "elo_team__0   1886.902479  26.804245    ...     20662.672671  1.000014\n",
       "elo_team__1   1984.699994  24.998467    ...     22156.136093  0.999940\n",
       "elo_team__2   2057.367130  24.865345    ...     21122.568430  0.999948\n",
       "elo_team__3   2033.974774  24.876458    ...     21113.004346  0.999944\n",
       "elo_team__4   2034.381207  24.340789    ...     20934.261489  0.999939\n",
       "elo_team__5   2050.433952  24.615052    ...     18405.593489  0.999988\n",
       "elo_team__6   1947.675731  25.370715    ...     22813.355967  0.999987\n",
       "elo_team__7   2022.237612  24.625649    ...     20147.598004  0.999961\n",
       "elo_team__8   2116.452494  26.094669    ...     17927.092693  0.999938\n",
       "elo_team__9   2028.985198  24.840915    ...     20253.032906  1.000074\n",
       "elo_team__10  1818.665401  30.575416    ...     18641.798768  1.000356\n",
       "elo_team__11  2021.611648  24.478701    ...     20191.938716  0.999938\n",
       "noise           42.486276   4.965166    ...     14463.611948  1.000298\n",
       "\n",
       "[13 rows x 7 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pm.summary(trace)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "_uuid": "d94e934b397471f3771a9730d493f9c0ddca944e"
   },
   "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>Win</th>\n",
       "      <th>Success</th>\n",
       "      <th>win_index</th>\n",
       "      <th>Elo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>New York Excelsior</td>\n",
       "      <td>138.0</td>\n",
       "      <td>8</td>\n",
       "      <td>2116.452494</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Los Angeles Valiant</td>\n",
       "      <td>108.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2057.367130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>108.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2050.433952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Los Angeles Gladiators</td>\n",
       "      <td>99.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2034.381207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Boston Uprising</td>\n",
       "      <td>98.0</td>\n",
       "      <td>9</td>\n",
       "      <td>2028.985198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Seoul Dynasty</td>\n",
       "      <td>97.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2033.974774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "      <td>94.0</td>\n",
       "      <td>11</td>\n",
       "      <td>2021.611648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Houston Outlaws</td>\n",
       "      <td>93.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2022.237612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>77.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1984.699994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Dallas Fuel</td>\n",
       "      <td>65.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1947.675731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Florida Mayhem</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1886.902479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Shanghai Dragons</td>\n",
       "      <td>23.0</td>\n",
       "      <td>10</td>\n",
       "      <td>1818.665401</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Win  Success  win_index          Elo\n",
       "7       New York Excelsior    138.0          8  2116.452494\n",
       "6      Los Angeles Valiant    108.0          2  2057.367130\n",
       "4          London Spitfire    108.0          5  2050.433952\n",
       "5   Los Angeles Gladiators     99.0          4  2034.381207\n",
       "0          Boston Uprising     98.0          9  2028.985198\n",
       "10           Seoul Dynasty     97.0          3  2033.974774\n",
       "8      Philadelphia Fusion     94.0         11  2021.611648\n",
       "3          Houston Outlaws     93.0          7  2022.237612\n",
       "9      San Francisco Shock     77.0          1  1984.699994\n",
       "1              Dallas Fuel     65.0          6  1947.675731\n",
       "2           Florida Mayhem     43.0          0  1886.902479\n",
       "11        Shanghai Dragons     23.0         10  1818.665401"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wins = wins.sort_values(by='win_index',ascending=True)\n",
    "wins['Elo'] = trace.elo_team.mean(axis=0)\n",
    "wins = wins.sort_values(by='Success',ascending=False)\n",
    "wins.head(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "_uuid": "aa02e4eca0f5f5835234922f746f109013e2990f"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [00:23<00:00, 84.04it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.6567593480345159"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def make_preds(trace,model_name):\n",
    "    ppc = pm.sample_posterior_predictive(trace, model=model_name, samples=500)\n",
    "    return ppc['out'].mean(axis=0)\n",
    "\n",
    "def scoreModel(trace,y,model_name):\n",
    "    ppc = pm.sample_posterior_predictive(trace, model=model_name, samples=2000)\n",
    "    pred = ppc['out'].mean(axis=0)\n",
    "    #return roc_auc_score(y, pred)\n",
    "    return accuracy_score(y,[round(x) for x in pred])\n",
    "\n",
    "scoreModel(trace,X['Success'],elo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "_uuid": "ae3cd96d62e017f8d3369a8a1b960aeecbb5cc69"
   },
   "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>Win</th>\n",
       "      <th>Lose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Win                 Lose\n",
       "38  London Spitfire  Philadelphia Fusion\n",
       "39  London Spitfire  Philadelphia Fusion\n",
       "40  London Spitfire  Philadelphia Fusion\n",
       "41  London Spitfire  Philadelphia Fusion\n",
       "42  London Spitfire  Philadelphia Fusion"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playoffs = pd.read_csv('../input/playoff_win_loss.csv')\n",
    "playoffs.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "_uuid": "279f15b00082975ac11d7f21e41e64d9a07372a8"
   },
   "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>Playoff_Success</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Win</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Boston Uprising</th>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>London Spitfire</th>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Los Angeles Gladiators</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Los Angeles Valiant</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>New York Excelsior</th>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Philadelphia Fusion</th>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        Playoff_Success\n",
       "Win                                    \n",
       "Boston Uprising                     5.0\n",
       "London Spitfire                    18.0\n",
       "Los Angeles Gladiators              3.0\n",
       "Los Angeles Valiant                 1.0\n",
       "New York Excelsior                  2.0\n",
       "Philadelphia Fusion                14.0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playoffs['Playoff_Success'] = 1.0\n",
    "playoff_wins=playoffs.groupby(\"Win\").agg({\"Playoff_Success\":'sum'})\n",
    "playoff_wins.head(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "_uuid": "af94e3240421667c1338aabf4d50ad69bc8bbafd"
   },
   "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>Win</th>\n",
       "      <th>Success</th>\n",
       "      <th>win_index</th>\n",
       "      <th>Elo</th>\n",
       "      <th>Playoff_Success</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>New York Excelsior</td>\n",
       "      <td>138.0</td>\n",
       "      <td>8</td>\n",
       "      <td>2116.452494</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Los Angeles Valiant</td>\n",
       "      <td>108.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2057.367130</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>London Spitfire</td>\n",
       "      <td>108.0</td>\n",
       "      <td>5</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>18.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Los Angeles Gladiators</td>\n",
       "      <td>99.0</td>\n",
       "      <td>4</td>\n",
       "      <td>2034.381207</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Boston Uprising</td>\n",
       "      <td>98.0</td>\n",
       "      <td>9</td>\n",
       "      <td>2028.985198</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Seoul Dynasty</td>\n",
       "      <td>97.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2033.974774</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Philadelphia Fusion</td>\n",
       "      <td>94.0</td>\n",
       "      <td>11</td>\n",
       "      <td>2021.611648</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Houston Outlaws</td>\n",
       "      <td>93.0</td>\n",
       "      <td>7</td>\n",
       "      <td>2022.237612</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>San Francisco Shock</td>\n",
       "      <td>77.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1984.699994</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Dallas Fuel</td>\n",
       "      <td>65.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1947.675731</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Florida Mayhem</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1886.902479</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Shanghai Dragons</td>\n",
       "      <td>23.0</td>\n",
       "      <td>10</td>\n",
       "      <td>1818.665401</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       Win       ...         Playoff_Success\n",
       "0       New York Excelsior       ...                     2.0\n",
       "1      Los Angeles Valiant       ...                     1.0\n",
       "2          London Spitfire       ...                    18.0\n",
       "3   Los Angeles Gladiators       ...                     3.0\n",
       "4          Boston Uprising       ...                     5.0\n",
       "5            Seoul Dynasty       ...                     NaN\n",
       "6      Philadelphia Fusion       ...                    14.0\n",
       "7          Houston Outlaws       ...                     NaN\n",
       "8      San Francisco Shock       ...                     NaN\n",
       "9              Dallas Fuel       ...                     NaN\n",
       "10          Florida Mayhem       ...                     NaN\n",
       "11        Shanghai Dragons       ...                     NaN\n",
       "\n",
       "[12 rows x 5 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "For our original batch of teams, only the top performing survived\n",
    "into the playoff season. We have made the very naive assumption that\n",
    "the best regular season teams would become the best post-season teams.\n",
    "\n",
    "This has turned out to be only partially true. The teams with the most wins\n",
    "in the regular season (Excelsior, Valiant, Gladiators) were quickly removed \n",
    "from the playoffs.\n",
    "\n",
    "Teams with high ELO standings, but low wins, (Fusion) ultimately had 14 wins in the playoffs \n",
    "and ended up making it to 2nd place overall.\n",
    "\"\"\"\n",
    "\n",
    "wins = pd.merge(wins,playoff_wins, on=['Win'], how = 'left')\n",
    "wins.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "_uuid": "d73510a90fabf72de8b607ed71fe56f18a722518"
   },
   "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>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   win_index  lose_index\n",
       "0         11           9\n",
       "1         11           9\n",
       "2         11           9\n",
       "3         11           9\n",
       "4         11           9"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "playoffs = pd.merge(playoffs,\\\n",
    "                    teams.rename(index=str,\\\n",
    "                    columns={\"team\": \"Win\", \"index\": \"win_index\"}),\n",
    "                    on = ['Win'] )\n",
    "\n",
    "playoffs = pd.merge(playoffs,\\\n",
    "                    teams.rename(index=str,\\\n",
    "                    columns={\"team\": \"Lose\", \"index\": \"lose_index\"}),\n",
    "                    on = ['Lose'] )\n",
    "\n",
    "playoffs = playoffs.drop(['Win','Lose','Playoff_Success'], axis=1)\n",
    "\n",
    "playoffs.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "_uuid": "43263b57c3c1e69ff0c10c5363005c616963ab9d",
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n",
      "100%|██████████| 2000/2000 [00:22<00:00, 88.12it/s]\n",
      "  2%|▏         | 11/500 [00:00<00:04, 101.86it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch AUC:  0 0.1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 91.82it/s]\n",
      "  1%|          | 11/2000 [00:00<00:19, 100.76it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   win_index  lose_index  Success  predictions\n",
      "0         11           9      1.0        0.508\n",
      "6          9          11      0.0        0.534\n",
      "6         11           9      1.0        0.454\n",
      "5         11           9      1.0        0.524\n",
      "9         11           8      1.0        0.394\n",
      "3         11           9      1.0        0.438\n",
      "8          8          11      0.0        0.624\n",
      "9          8          11      0.0        0.642\n",
      "4          9          11      0.0        0.514\n",
      "5          9          11      0.0        0.496\n",
      "7          8          11      0.0        0.586\n",
      "7         11           8      1.0        0.342\n",
      "2         11           9      1.0        0.450\n",
      "1          9          11      0.0        0.484\n",
      "1         11           9      1.0        0.490\n",
      "4         11           9      1.0        0.484\n",
      "0          9          11      0.0        0.558\n",
      "2          9          11      0.0        0.476\n",
      "3          9          11      0.0        0.518\n",
      "8         11           8      1.0        0.344\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [00:22<00:00, 89.58it/s]\n",
      "  2%|▏         | 11/500 [00:00<00:04, 105.62it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch AUC:  10 0.15\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 94.18it/s] \n",
      "  1%|          | 11/2000 [00:00<00:19, 102.88it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    win_index  lose_index  Success  predictions\n",
      "11          8          11      0.0        0.612\n",
      "12          8          11      0.0        0.626\n",
      "15          4           5      1.0        0.428\n",
      "14          5           4      0.0        0.526\n",
      "18          9          11      1.0        0.522\n",
      "19          9          11      1.0        0.488\n",
      "10          8          11      0.0        0.676\n",
      "18         11           9      0.0        0.504\n",
      "15          5           4      0.0        0.532\n",
      "16          5           4      0.0        0.512\n",
      "19         11           9      0.0        0.498\n",
      "16          4           5      1.0        0.490\n",
      "17          5           2      0.0        0.502\n",
      "12         11           8      1.0        0.350\n",
      "14          4           5      1.0        0.508\n",
      "11         11           8      1.0        0.384\n",
      "17          2           5      1.0        0.524\n",
      "13         11           5      1.0        0.462\n",
      "10         11           8      1.0        0.360\n",
      "13          5          11      0.0        0.524\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [00:21<00:00, 92.56it/s]\n",
      "  2%|▏         | 11/500 [00:00<00:04, 107.08it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch AUC:  20 0.95\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 92.53it/s]\n",
      "  0%|          | 10/2000 [00:00<00:20, 97.51it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    win_index  lose_index  Success  predictions\n",
      "22         11           9      0.0        0.446\n",
      "27          5          11      1.0        0.560\n",
      "21          9          11      1.0        0.484\n",
      "28         11           5      0.0        0.446\n",
      "29         11           8      0.0        0.350\n",
      "21         11           9      0.0        0.496\n",
      "28          5          11      1.0        0.534\n",
      "22          9          11      1.0        0.486\n",
      "25          5          11      1.0        0.536\n",
      "20         11           9      0.0        0.482\n",
      "24         11           5      0.0        0.430\n",
      "20          9          11      1.0        0.494\n",
      "23         11           5      0.0        0.438\n",
      "23          5          11      1.0        0.564\n",
      "27         11           5      0.0        0.486\n",
      "29          8          11      1.0        0.678\n",
      "24          5          11      1.0        0.548\n",
      "25         11           5      0.0        0.476\n",
      "26          5          11      1.0        0.546\n",
      "26         11           5      0.0        0.482\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [00:23<00:00, 84.46it/s]\n",
      "  2%|▏         | 10/500 [00:00<00:04, 99.06it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch AUC:  30 0.75\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 92.19it/s] \n",
      "  1%|          | 11/2000 [00:00<00:19, 102.96it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    win_index  lose_index  Success  predictions\n",
      "30         11           8      0.0        0.388\n",
      "33          5           4      1.0        0.512\n",
      "31          4           5      0.0        0.462\n",
      "37          2           5      0.0        0.538\n",
      "35          5           4      1.0        0.524\n",
      "39          2           5      0.0        0.538\n",
      "36          5           4      1.0        0.494\n",
      "34          4           5      0.0        0.462\n",
      "31          5           4      1.0        0.538\n",
      "38          2           5      0.0        0.492\n",
      "37          5           2      1.0        0.462\n",
      "36          4           5      0.0        0.486\n",
      "32          4           5      0.0        0.472\n",
      "33          4           5      0.0        0.458\n",
      "35          4           5      0.0        0.496\n",
      "38          5           2      1.0        0.490\n",
      "34          5           4      1.0        0.524\n",
      "30          8          11      1.0        0.630\n",
      "39          5           2      1.0        0.470\n",
      "32          5           4      1.0        0.526\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 2000/2000 [00:21<00:00, 91.53it/s]\n",
      "  2%|▏         | 11/500 [00:00<00:04, 107.23it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Batch AUC:  40 0.16666666666666666\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 93.68it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    win_index  lose_index  Success  predictions\n",
      "40          2           5      0.0        0.486\n",
      "42          2           5      0.0        0.498\n",
      "41          2           5      0.0        0.526\n",
      "42          5           2      1.0        0.492\n",
      "41          5           2      1.0        0.480\n",
      "40          5           2      1.0        0.488\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "THIS seems backwards. Bad teams are winning?\n",
    "\"\"\"\n",
    "\n",
    "for i in range(0,playoffs.count()[0],10):\n",
    "    sample = playoffs.iloc[i:i+10]\n",
    "    batch = build_training_df(sample,add_losses=True)\n",
    "    batch = batch.sample(frac=1.0)\n",
    "    winner_index.set_value(np.array(batch['win_index']))\n",
    "    loser_index.set_value(np.array(batch['lose_index']))\n",
    "    model_output.set_value(np.array(batch['Success']))\n",
    "    \n",
    "    print (\"Batch AUC: \",i, scoreModel(trace,batch['Success'],elo) )\n",
    "    batch['predictions']=make_preds(trace,elo)\n",
    "    print (batch.head(20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "_uuid": "aef3d5f6f4ae5a852055d5380ec20dec4c0612e7"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 500/500 [00:05<00:00, 90.44it/s] \n"
     ]
    }
   ],
   "source": [
    "X_test = build_training_df(playoffs,add_losses=True)\n",
    "winner_index.set_value(np.array(X_test['win_index']))\n",
    "loser_index.set_value(np.array(X_test['lose_index']))\n",
    "model_output.set_value(np.array(X_test['Success']))\n",
    "X_test['preds'] = make_preds(trace,elo)\n",
    "X_test['prediction'] = [round(x) for x in X_test['preds'].values ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "_uuid": "185f80607c87767d8b4fd735ea60b74bf8c2d839"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5116279069767442"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(X_test['Success'], X_test['prediction'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "_uuid": "f9a5d129a6fc6d3b0039db91db209729da94c015"
   },
   "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>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "      <th>Success</th>\n",
       "      <th>preds</th>\n",
       "      <th>prediction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.522</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.498</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.520</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.496</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.492</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    win_index  lose_index  Success  preds  prediction\n",
       "38          2           5      0.0  0.522         1.0\n",
       "39          2           5      0.0  0.498         0.0\n",
       "40          2           5      0.0  0.520         1.0\n",
       "41          2           5      0.0  0.496         0.0\n",
       "42          2           5      0.0  0.492         0.0"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "_uuid": "b067e1d6f75066cb90f916754aa429bda3fb3f30"
   },
   "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>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "      <th>Success</th>\n",
       "      <th>preds</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Elo_Win</th>\n",
       "      <th>Elo_Lose</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.464</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.452</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.506</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.478</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.478</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    win_index  lose_index     ...          Elo_Win    Elo_Lose\n",
       "81          5           2     ...      2050.433952  2057.36713\n",
       "82          5           2     ...      2050.433952  2057.36713\n",
       "83          5           2     ...      2050.433952  2057.36713\n",
       "84          5           2     ...      2050.433952  2057.36713\n",
       "85          5           2     ...      2050.433952  2057.36713\n",
       "\n",
       "[5 rows x 7 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test = pd.merge(X_test,\\\n",
    "                    wins[['win_index','Elo']],\n",
    "                    on = ['win_index'],how='inner')\n",
    "X_test = X_test.rename(index=str,\\\n",
    "                    columns={\"Elo\": \"Elo_Win\"})\n",
    "\n",
    "X_test = pd.merge(X_test,\\\n",
    "                    wins[['win_index','Elo']].rename(index=str,\\\n",
    "                       columns={\"win_index\": \"lose_index\"}),\n",
    "                    on = ['lose_index'],how='inner')\n",
    "X_test = X_test.rename(index=str,\\\n",
    "                    columns={\"Elo\": \"Elo_Lose\"})\n",
    "X_test.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "_uuid": "b61ae5ca325efcc4a530690f2730fc0f5f6cd5f0"
   },
   "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>win_index</th>\n",
       "      <th>lose_index</th>\n",
       "      <th>Success</th>\n",
       "      <th>preds</th>\n",
       "      <th>prediction</th>\n",
       "      <th>Elo_Win</th>\n",
       "      <th>Elo_Lose</th>\n",
       "      <th>raw_expected</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.464</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "      <td>0.490024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.452</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "      <td>0.490024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.506</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "      <td>0.490024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.478</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "      <td>0.490024</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.478</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2050.433952</td>\n",
       "      <td>2057.36713</td>\n",
       "      <td>0.490024</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    win_index  lose_index      ...         Elo_Lose  raw_expected\n",
       "81          5           2      ...       2057.36713      0.490024\n",
       "82          5           2      ...       2057.36713      0.490024\n",
       "83          5           2      ...       2057.36713      0.490024\n",
       "84          5           2      ...       2057.36713      0.490024\n",
       "85          5           2      ...       2057.36713      0.490024\n",
       "\n",
       "[5 rows x 8 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test['raw_expected'] = Expected(X_test['Elo_Win'],X_test['Elo_Lose'])\n",
    "X_test.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "_uuid": "9267fbdb7a11be98e48370a7cbb8587bd654b99c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5154137371552191"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "roc_auc_score(X_test['Success'], X_test['raw_expected'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "_uuid": "33b74a09978c2b877e77a743c72c7d86b1b3393b"
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}