Outs_above_average_for_shortstops.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Which Shortstop Converted the Most Outs Above Average?\n",
    "## Chapter 1 Introduction\n",
    "In this task, we will be working with a dataset which has one row per in-play ground ball event for a subset of Major League games. On the basis of this dataset, we will try to figure out which shortstop converted the most **Outs Above Average (OAA)**.\n",
    "\n",
    "According to Statcast, **OAA** is a range-based metric of skill that shows how many outs a player has saved. For example, a fielder who catches a 25% **Out Probability** play gets +0.75; one who fails to make the play gets -0.25.\n",
    "\n",
    "Therefore, in this task, we will develop a framework which is capable of estimating **Out Probability** for each in-play ground ball first fielded by shortstop and further calculating **OAA** for each shortstop."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import collections  as mc\n",
    "from matplotlib.patches import Arc\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.calibration import CalibratedClassifierCV\n",
    "from sklearn.calibration import calibration_curve\n",
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data\n",
    "data = pd.read_excel('shortstopdefense.xlsx', sheet_name='TBJExercise')\n",
    "variable = pd.read_excel('shortstopdefense.xlsx', sheet_name='Variable Dictionary')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 2 Feature Preprocessing\n",
    "Since our main purpose is to calculate **OAA** for each shortstop, we only select in-play ground ball event first fielded by shortstop from the given dataset in the first stage. After that, we \n",
    "check whether there are NaN values in the columns that will be used as features for estimating **Out Probability**. Finally, our final dataset contains 4721 in-play ground ball events."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 4722 entries, 0 to 20146\n",
      "Data columns (total 28 columns):\n",
      " #   Column                   Non-Null Count  Dtype  \n",
      "---  ------                   --------------  -----  \n",
      " 0   id                       4722 non-null   int64  \n",
      " 1   playerid                 4722 non-null   int64  \n",
      " 2   pos                      4722 non-null   int64  \n",
      " 3   player_x                 4722 non-null   float64\n",
      " 4   player_y                 4722 non-null   float64\n",
      " 5   player_vx                4721 non-null   float64\n",
      " 6   player_vy                4721 non-null   float64\n",
      " 7   playtype                 4722 non-null   object \n",
      " 8   eventtype                4722 non-null   object \n",
      " 9   trajectory               4722 non-null   object \n",
      " 10  fielded_pos              4722 non-null   float64\n",
      " 11  fieldingplay             4722 non-null   float64\n",
      " 12  fielded_scoring          4722 non-null   object \n",
      " 13  player_out_credit        4722 non-null   bool   \n",
      " 14  is_runnersgoing          4722 non-null   bool   \n",
      " 15  is_bunt                  4722 non-null   bool   \n",
      " 16  launch_vert_ang          4722 non-null   float64\n",
      " 17  launch_horiz_ang         4722 non-null   float64\n",
      " 18  launch_speed             4722 non-null   float64\n",
      " 19  launch_spin_rate         4722 non-null   float64\n",
      " 20  launch_spin_axis         4722 non-null   float64\n",
      " 21  landing_location_radius  4722 non-null   float64\n",
      " 22  hang_time                4722 non-null   float64\n",
      " 23  landing_location_x       4722 non-null   float64\n",
      " 24  landing_location_y       4722 non-null   float64\n",
      " 25  runner_on_first          4722 non-null   bool   \n",
      " 26  runner_on_second         4722 non-null   bool   \n",
      " 27  runner_on_third          4722 non-null   bool   \n",
      "dtypes: bool(6), float64(15), int64(3), object(4)\n",
      "memory usage: 876.2+ KB\n"
     ]
    }
   ],
   "source": [
    "# Select a subet of the given data\n",
    "shortstop_data = data[data['fielded_pos']==6]\n",
    "shortstop_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop the NaN value\n",
    "shortstop_data = shortstop_data.dropna(how='any',axis=0).reset_index()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before directly starting to build predictive models, we are going to research which factors might be helpful in calculating **OAA** more accurately. According to [Tom Tango's article](http://tangotiger.com/images/uploads/History_of_the_Fielding.pdf?fbclid=IwAR3bIoMfjvaWd_l9ucYVCYRGCJxUHWApJx6I7aQG4KC4KQfBd1pSUvls4m4), the following four factors primarily affect the chance of a play being converted into an out for infielders:\n",
    "\n",
    "1. How far the infielder has to go to reach the ball\n",
    "2. How much time the infielder has to get there\n",
    "3. How far the infielder is from the base the runner is heading to\n",
    "4. How fast the runner is\n",
    "\n",
    "Regarding the first factor, since we have shortstop's location at the moment the ball is hit and location for where the batted ball first landed, we could easily calculate distance that shortstop has to go to reach the ball, named as **distance_ball**. Regarding the second factor, since we have shortstop's velocity and distance that shortstop has to go to reach the ball, we could also easily calculate how much time the infielder has to get there, named as **moving_time**. Regarding the fourth factor, since we don't have detailed physical information about runner, we basically assume that all runners'speed are the same.\n",
    "\n",
    "However, the third factor might be our biggest challenge. Since, from the given dataset, we don't know which base the shortstop would like to throw to, we first figure out every possible sequences of **fieldingplay** and then utilize **plattype**, **eventtype**, **fielded_scoring**, **player_out_credict**, **runner_on_first**, **runner_on_second**, and **runner_on_third** to speculate which base the shortstop would like to throw to against each possible sequence of **fieldingplay**.\n",
    "\n",
    "Nevertheless, we find out that when the sequence of **fieldingplay** equals to **6, 67, 68, or 69**, the corresponding event result is bascially single, double, or error. In that case, it's ambiguous to specify which base the shortstop would like to throw to. Therefore, we decide to slightly transform the third factor. Instead of calcualting  how far the infielder is from the base the runner is heading to, we assume that all shortstops' arm strength are the same (i.e., 90 mph) and further calcuale how much time the infielder is from the base the runner is heading to. Hence, the definition of the modified third factor is how much time the shortstop handle the ball after reaching the ball, named as **handling_time**. Accordingly, when the sequence of **fieldingplay** equals to **6, 67, 68, or 69**, it's more reasonable to determine that the shortstop take zero seconds to deal with the ball after reaching the ball."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate distance between shortstop and ball\n",
    "shortstop_data = shortstop_data.assign(distance_ball=lambda x: ((x['player_x']-x['landing_location_x'])**2+(x['player_y']-x['landing_location_y'])**2)**(0.5))\n",
    "# Calculate moving speed\n",
    "shortstop_data = shortstop_data.assign(moving_speed=lambda x: ((x['player_vx']**2)+(x['player_vy']**2))**(0.5))\n",
    "# Calculate moving time\n",
    "shortstop_data = shortstop_data.assign(moving_time=lambda x: x['distance_ball']/x['moving_speed'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>fieldingplay</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6.0</td>\n",
       "      <td>589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>61.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>62.0</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>63.0</td>\n",
       "      <td>3193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>64.0</td>\n",
       "      <td>396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>65.0</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>67.0</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>68.0</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>69.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>625.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>626.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>632.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>635.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>636.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>642.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>643.0</td>\n",
       "      <td>368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>645.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>646.0</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>653.0</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>654.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>656.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>672.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>6251.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>6253.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>6256.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>6464.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>65256.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    fieldingplay  count\n",
       "0            6.0    589\n",
       "1           61.0      2\n",
       "2           62.0     19\n",
       "3           63.0   3193\n",
       "4           64.0    396\n",
       "5           65.0     34\n",
       "6           67.0     35\n",
       "7           68.0     15\n",
       "8           69.0      5\n",
       "9          625.0      2\n",
       "10         626.0      2\n",
       "11         632.0      1\n",
       "12         635.0      2\n",
       "13         636.0      1\n",
       "14         642.0      1\n",
       "15         643.0    368\n",
       "16         645.0      1\n",
       "17         646.0     20\n",
       "18         653.0     24\n",
       "19         654.0      2\n",
       "20         656.0      2\n",
       "21         672.0      1\n",
       "22        6251.0      1\n",
       "23        6253.0      1\n",
       "24        6256.0      1\n",
       "25        6464.0      2\n",
       "26       65256.0      1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find every possible sequences of fieldingplay\n",
    "shortstop_data.groupby('fieldingplay').size().reset_index().rename(columns={0:'count'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate handling time\n",
    "location_1B = {'x': 45*(2**0.5), 'y': 45*(2**0.5)}\n",
    "location_2B = {'x': 0, 'y': 90*(2**0.5)}\n",
    "location_3B = {'x': -45*(2**0.5), 'y': 45*(2**0.5)}\n",
    "location_HB = {'x': 0, 'y': 0}\n",
    "average_throw_velocity = 90*(5280/3600) # convert 90 mph into 132 ft/s\n",
    "\n",
    "def handling_time_function(data):\n",
    "    \n",
    "    time_list = []\n",
    "    n = data.shape[0]\n",
    "    for i in range(n):\n",
    "        if data['fieldingplay'][i] in [6, 67, 68, 69, 672]:\n",
    "            time_list.append(0)\n",
    "        elif data['fieldingplay'][i] in [61, 63, 632, 635, 636]:\n",
    "            distance = ((data['landing_location_x'][i]-location_1B['x'])**2+(data['landing_location_y'][i]-location_1B['y'])**2)**(0.5)\n",
    "            time = distance/average_throw_velocity\n",
    "            time_list.append(time)\n",
    "        elif data['fieldingplay'][i] in [64, 642, 643, 645, 646, 6464, 653]:\n",
    "            distance = ((data['landing_location_x'][i]-location_2B['x'])**2+(data['landing_location_y'][i]-location_2B['y'])**2)**(0.5)\n",
    "            time = distance/average_throw_velocity\n",
    "            time_list.append(time)\n",
    "        elif data['fieldingplay'][i] in [65, 654, 656, 65256]:\n",
    "            distance = ((data['landing_location_x'][i]-location_3B['x'])**2+(data['landing_location_y'][i]-location_3B['y'])**2)**(0.5)\n",
    "            time = distance/average_throw_velocity\n",
    "            time_list.append(time)\n",
    "        elif data['fieldingplay'][i] in [62, 625, 626, 6251,6253, 6256]:\n",
    "            distance = ((data['landing_location_x'][i]-location_HB['x'])**2+(data['landing_location_y'][i]-location_HB['y'])**2)**(0.5)\n",
    "            time = distance/average_throw_velocity\n",
    "            time_list.append(time)\n",
    "            \n",
    "    return time_list\n",
    "\n",
    "handling_time = handling_time_function(shortstop_data)\n",
    "shortstop_data = shortstop_data.assign(handling_time=handling_time)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 3 Model Evaluation and Parameter Tuning\n",
    "After features preprocessing, we are going to define our machine learning problem. Again, our purpose is to estimate **Out Probability** for each in-play ground ball first fielded by shortstop and further calculating **OAA** for each shortstop. In fact, on the basis of the given dataset, we don't have any information about **Out Probability** which could be used as a response variable.\n",
    "\n",
    "However, we do have a critical variable named **player_out_credict** which represents whether the shortstop is given out credit on the play. Therefore, we would regard **player_out_credict** as a classification label and further consider our problem as a classification problem. After predicting whether the shortstop is given out credit on the play, we could further predict probabilities of the shortstop being given out credit on the play (i.e., **Out Probability**). Hence, in this chapter, we respectively use **Random Forest Classifier**, **k-NN Classifier**, and **XGBoost Classifier** to predict whether the shortstop is given out credit on the play given the features below:\n",
    "\n",
    "1. distance_ball: how far the infielder has to go to reach the ball\n",
    "2. moving_speed: how fast the shortstop is\n",
    "3. moving_time: how much time the shortstop has to go to reach the ball\n",
    "4. handling_time: how much time the shortstop handle the ball after reaching the ball\n",
    "5. hang_time: the time until a batted ball reached the ground\n",
    "6. launch_vert_ang: the intial vertical launch angle of the batted ball\n",
    "7. launch_horiz_ang: the intial horizontal spray angle of the batted ball\n",
    "8. launch_speed: the initial exit velocity of the batted ball\n",
    "\n",
    "In addition, we use **Scikit-learn's Pipeline and GridSearchCV** to tune each model's parameters. Afterward, we compare each model's test accuracy and confusion matrix. Finally, we could verify that **Random Forest Classifier**  has similar performance to **XGBoost Classifier** and they both outperform **k-NN Classifier** on test accuracy and confusion matrix. Therefore, we are able to use **Random Forest Classifier** and **XGBoost Classifier** to predict **player_out_credict** and **Out Probability** later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the data into random train and test subsets\n",
    "features = ['distance_ball', 'moving_speed', 'moving_time', 'handling_time',\n",
    "            'launch_vert_ang', 'launch_horiz_ang', 'launch_speed', 'hang_time']\n",
    "shortstop_data['player_out_credit'] = shortstop_data['player_out_credit'].astype(int)\n",
    "\n",
    "X = shortstop_data[features].to_numpy()\n",
    "y = shortstop_data[['player_out_credit']].to_numpy()\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 646,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tune parameters for random forest classifier\n",
    "pipe = Pipeline([('classifier', RandomForestClassifier())])\n",
    "search_space = [{'classifier': [RandomForestClassifier(random_state=42)],\n",
    "                 'classifier__n_estimators': [200, 600, 1000],\n",
    "                 'classifier__max_features': ['auto', 'sqrt'],\n",
    "                 'classifier__max_depth': [20, 60, 100, None],\n",
    "                 'classifier__min_samples_split': [2, 5, 10],\n",
    "                 'classifier__min_samples_leaf': [1, 2, 4]}]\n",
    "\n",
    "clf = GridSearchCV(pipe, search_space, cv=5, verbose=0)\n",
    "clf = clf.fit(X_train, y_train.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 657,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy: 0.9862433862433863\n",
      "Confusion Matrix:\n",
      "[[136   6]\n",
      " [  7 796]]\n"
     ]
    }
   ],
   "source": [
    "# Calculate test accuracy and confusion matrix for random forest classifier\n",
    "rf_model = clf.best_estimator_['classifier']\n",
    "rf_model.fit(X_train, y_train.ravel())\n",
    "accuracy =  rf_model.score(X_test, y_test.ravel())\n",
    "print(\"Test Accuracy: {}\".format(accuracy))\n",
    "\n",
    "y_pred = rf_model.predict(X_test)\n",
    "cf_matrix = confusion_matrix(y_true=y_test, y_pred=y_pred)\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cf_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 K-NN Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 658,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tune parameters for k-NN Classifier\n",
    "pipe = Pipeline([('classifier', KNeighborsClassifier())])\n",
    "search_space = [{'classifier': [KNeighborsClassifier(algorithm='auto')],\n",
    "                 'classifier__n_neighbors': [5, 7, 9],\n",
    "                 'classifier__weights': ['uniform', 'distance'],\n",
    "                 'classifier__p': [1, 2]}]\n",
    "\n",
    "clf = GridSearchCV(pipe, search_space, cv=5, verbose=0)\n",
    "clf = clf.fit(X_train, y_train.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 660,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy: 0.8539682539682539\n",
      "Confusion Matrix:\n",
      "[[  5 137]\n",
      " [  1 802]]\n"
     ]
    }
   ],
   "source": [
    "# Calculate test accuracy and confusion matrix for k-NN Classifier\n",
    "knn_model = clf.best_estimator_['classifier']\n",
    "knn_model.fit(X_train, y_train.ravel())\n",
    "accuracy =  knn_model.score(X_test, y_test.ravel())\n",
    "print(\"Test Accuracy: {}\".format(accuracy))\n",
    "\n",
    "y_pred = knn_model.predict(X_test)\n",
    "cf_matrix = confusion_matrix(y_true=y_test, y_pred=y_pred)\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cf_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3 XGBoost Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 661,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Tune parameters for XGBoost Classifier\n",
    "pipe = Pipeline([('classifier', XGBClassifier())])\n",
    "search_space = [{'classifier': [XGBClassifier(random_state=42)],\n",
    "                 'classifier__max_depth': [3, 6, 9],\n",
    "                 'classifier__min_child_weight': [3, 6, 9],\n",
    "                 'classifier__gamma': [0.1, 0.3, 0.5],\n",
    "                 'classifier__colsample_bytree': [0.3, 0.6, 0.9]}]\n",
    "\n",
    "clf = GridSearchCV(pipe, search_space, cv=5, verbose=0)\n",
    "clf = clf.fit(X_train, y_train.ravel())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 663,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy: 0.9862433862433863\n",
      "Confusion Matrix:\n",
      "[[135   7]\n",
      " [  6 797]]\n"
     ]
    }
   ],
   "source": [
    "# Calculate test accuracy and confusion matrix for XGBoost Classifier\n",
    "xgb_model = clf.best_estimator_['classifier']\n",
    "xgb_model.fit(X_train, y_train.ravel())\n",
    "accuracy =  xgb_model.score(X_test, y_test.ravel())\n",
    "print(\"Test Accuracy: {}\".format(accuracy))\n",
    "\n",
    "y_pred = xgb_model.predict(X_test)\n",
    "cf_matrix = confusion_matrix(y_true=y_test, y_pred=y_pred)\n",
    "print(\"Confusion Matrix:\")\n",
    "print(cf_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 4 Feature importance and Classification Probability Calibration\n",
    "According to the previous chapter, we find out that **Random Forest Classifier**  has similar performance to **XGBoost Classifier** and they both outperform **k-NN Classifier** on test accuracy and confusion matrix. Therefore, now we are able to use **Random Forest Classifier** and **XGBoost Classifier** to predict **player_out_credict** and **Out Probability**. \n",
    "\n",
    "Hence, in this chapter, we first figure out which features have siginificant impact on predicting **player_out_credict** by calculating feature importance. Regarding both **Random Forest Classifier** and **XGBoost Classifier**, it's obvious that **handling_time**, which represents how much time the shortstop handle the ball after reaching the ball, has the most siginificant impact on predicting **player_out_credict**.\n",
    "\n",
    "After building the machine learning models and understanding the feature importance, we start to focus on which model has a better performance on predict **Out Probability** since both **Random Forest Classifier** and **XGBoost Classifier** achieve a test accuracy of 98.6% on predicting the classification labels, **player_out_credict**.\n",
    "\n",
    "However, in classification tasks, test accuracy only measures how often we are correct and says nothing about how confident we are in being correct. As a result, classification models tend to have much weaker and more systematically biased probability prediction forecasts. Hence, in this chapter, we evaluate probability forecasting performances of both **Random Forest Classifier** and **XGBoost Classifier**. Besides, we try to fix our predicted **Out Probability** by calibrating both **Random Forest Classifier** and **XGBoost Classifier** using sigmoid and isotonic methods. \n",
    "\n",
    "In the blue, we have the calibration curve for the original **Random Forest Classifier** and **XGBoost Classifier**.\n",
    "Meanwhile, in the yellow, we could see that the sigmoid adjusted probabilities are a little bit close to the true mean probabilities (i.e., gray dotted line). However, it seems that **XGBoost Classifier** calibrated by sigmoid method is a little bit better than **Random Forest Classifier** calibrated by sigmoid method since **Random Forest Classifier** calibrated by sigmoid method overshoots the high-probability data by too much. As a result, in the next chapter, we will use **XGBoost Classifier** calibrated by sigmoid method to predict **Out Probability** and further calculate **Outs Above Average**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1 Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Calculate feature importance for random forest classifier\n",
    "rf_model = RandomForestClassifier(n_estimators=600, max_features='auto', max_depth=20, min_samples_split=2, min_samples_leaf=4, random_state=42)\n",
    "rf_model.fit(X_train, y_train.ravel())\n",
    "importances = rf_model.feature_importances_\n",
    "indices = np.argsort(importances)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "plt.title('Feature Importance')\n",
    "plt.barh(range(len(indices)), importances[indices], color='b', align='center')\n",
    "plt.yticks(range(len(indices)), [features[i] for i in indices])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlMAAAEvCAYAAABhSUTPAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAACGX0lEQVR4nOzdd1hUd9bA8e+dYaiCVEWl2VGwAXYxVqwpxmQT0zdt05PNbnazu3nTNtlNtmYTTe9uTNOYmI09FjRqVCyA2BWQLr2XmbnvHxcQlDLAwFDO53l8gDt37j2AMIdfOUdRVRUhhBBCCNE6OlsHIIQQQgjRlUkyJYQQQgjRBpJMCSGEEEK0gSRTQgghhBBtIMmUEEIIIUQbSDIlhBBCCNEGdja8t9RkEEIIIURXoTT2gIxMCSGEEEK0gSRTQgghhBBtIMmUEEIIIUQbSDIlhBBCCNEGtlyAfoWqqipSUlIoLy+3dSiiE3N0dMTPzw+DwWDrUIQQQggUGzY6vuLG58+fx9XVFS8vLxSl0UXzogdTVZWcnByKiooYOHCgrcMRQgjRc3SN3Xzl5eWSSIkmKYqCl5eXjF4KIYToNDpVMgVIIiWaJf9HhBBCdCbNrplSFOVDYDGQpapqaAOPK8B/gIVAKXCXqqqHrB1oR0lMTGTx4sXEx8fXHnv++efp1asXv/3tb61yj7vuuovFixdzww03cO+99/Lkk08ycuRIevXqRXFxcauvu2PHDuzt7ZkyZUqLnhcUFMTBgwfx9vZu9b1FzxXx0hayiyuvOO7dy56Dz8y1QUQW+sYXyjOvPO7YF67P6Ph4OoEO+V76+kJmA1/3vn0ho3t83Vv9dewBX5vuypKRqY+B+U08vgAYWv3vfuCttofVc7z//vuMHDnS4vONRmOjj+3YsYM9e/ZYIywhLNbQi0ZTxzuNhhKppo73AB3yvWwoWWjqeBfU6q9jD/jadFfNjkypqhqtKEpQE6dcC3yqaivZ9ymK4q4oSj9VVdOtFWRDbPHX8IwZM5g4cSLbt28nPz+fDz74gMjISEwmE7///e/ZuHEjOp2O++67j0cffZQXX3yR77//nrKyMqZMmcI777xzxRTVjBkz+Mc//kFERAQAv/71r9m8eTO+vr588cUX+Pj4MGPGDMaOHcvu3btZtmwZw4YN46WXXqKyshIvLy8+++wzysrKePvtt9Hr9fz3v//ljTfeIDg4mAceeIDk5GQAXnvtNaZOnUpOTg7Lli0jNTWVyZMnY8NNCEJ0Tuc+BUUBdKDoAEV729j7tW912vPqvm/xNdpyfp3nNXeN2vOV6reW+Xx/shW+sLCsicf2//FVq9zD1m4+ktLoY8deOI5eUdDrFJTqtzod6BUdfk1cs+zrNegd7NHb26NzdECxtweDAS5/29D7en2LvtedUc1r/oERt+FjyL/yhHzgRduN4FmjNMIA4EKdj1Oqj7VrMmWrv4aNRiP79+9n/fr1vPDCC2zdupV3332XxMREjhw5gp2dHbm5uQA88sgjPPvsswDcfvvt/O9//+Pqq69u9NolJSVERETw73//mxdffJEXXniB5cuXA1BZWcnBgwcByMvLY9++fSiKwvvvv8/f/vY3/vnPf/LAAw/Um4685ZZb+PWvf820adNITk5m3rx5HD9+nBdeeIFp06bx7LPP8sMPP/DBBx+055dMiK5n3522jqCDKPWSr5OhKmZ0mFUFFQUzCmrN+yd09T+uft+MDhVQr3iedrzu9VQUeKn61mYu7emufn+C+jTVF2v4n/my56pXXqPJ5zV1DUvvacFzJyhNXGNHE8+d3Pg9nf56g2XxNhK3UafHpLPDrOgx6fSYdAbMih6z3k47rrdD1dth1hkw21W/rzeg6g2Y7QyodgbQG1AN9tr7dvZgZwCD/aV/dvbo7A2oBi3Z0zk4oNg7orN3QLE3oHfQjuntDegcHdDb22sJoqM9dg4O6BwdsHOw1xLFugmhXl/72l6TSBnNegqNbnja52mfpzs2HcHrVHWm6nrh+2MkpBW26rk3vbO3weMj+7vx3NUhTT63scXNNcevv/56AMLDw0lMTARg69atPPDAA9jZaV9OT09PALZv387f/vY3SktLyc3NJSQkpMlkSqfTcdNNNwFw22231d4LqD0OkJKSwk033UR6ejqVlZWNlgjYunUrCQkJtR8XFhZSXFxMdHQ033zzDQCLFi3Cw8Oj8S+IED3RNWdBNYNa/aqkVr8iqZe9T/U5dd9v6vx6z23t89p4fu37V17jox2nARWdoqKlTqBTzCio3Bg+AKXmOZfdR7k8ljrXVy7/nBPOX1pgUj1gVvOvctIkajMBVb3s2tSPF/XSteucX3OOUpNV1D0Hqh+riemyz6feda88R7myok+XYYcJO0y2C8AMmIASoJgWJbmqCucHaG+rzAZiCsLYkzcVZ30pvwp4B0Wx/ffFGslUKuBf52O/6mNdkpeXF3l5efWO5ebm1iYsDg4OAOj1+ibXL5WXl/PQQw9x8OBB/P39ef7551u8nb9uYufi4lL7/qOPPsqTTz7JNddcw44dO3j++ecbfL7ZbGbfvn04Ojq26L5C9Hi9Btk6Apt45YsfGn3srscWWecmcxqfbrJXG/5DuNNQ67y6N5HMjn5hIwpXJqQKKvv/OKvxxDd4+KXk8vKEM+ag1ZLmliTqqtmEqpoxmc2YzSbM5rrvmzCbVcxmI2pVJWajEdVYpb2tqgJTFWaTCYza+6rJBOYqMJnAbKzz1ghmE5hNKDVv1UsfK6qZ7BIj53TDOFM4khJTLwKdEpnuuVOLvROwRjK1DnhEUZQvgIlAgTXWSzU3ghT0dOM/9F/+anKr79urVy/69evHtm3bmDVrFrm5uWzcuJHHH3+cjz76qMHnzJ07l3feeYeZM2fWTvPpdNpPgre3N8XFxaxevZobbrihyXubzWZWr17NzTffzKpVq5g2bVqD5xUUFDBgwAAAPvnkk9rjrq6uFBZeGs2LiorijTfe4KmnngLgyJEjjB07lunTp7Nq1SqeeeYZNmzYcEXyKERL9HYyUFBWdcVx7172NoimBRz7Nr6br4fycDaQV9rO30snJygru/J43y7wda9dawagb/Q0eyfPRtf04tzEyihz38Z383mGtzhca7g8t7OVO/70CZF2iQx2OEOkZzSBTtZZw2ctlpRG+ByYAXgripICPAcYAFRVfRtYj1YW4QxaaYRftlewHeXTTz/l4Ycf5sknnwTgueeeY/DgwY2ef++993Lq1ClGjx6NwWDgvvvu45FHHuG+++4jNDQUX19fxo8f3+x9XVxc2L9/Py+99BJ9+vThyy+/bPC8559/nhtvvBEPDw9mzZrF+fPnAbj66qu54YYb+O6773jjjTd4/fXXefjhhxk9ejRGo5Hp06fz9ttv89xzz7Fs2TJCQkKYMmUKAQEBrfgqCaGZO7IvG+MzuHtqEK9vO0P8C/Po5dBpVxBcMv072DwJJr4Pg++xdTSdwu2Tg3hj22n2/3EOPq4O1r9Bfr6WkNx9N3TjtZqt3gQl5Q9qlZaW8vPPP+Pm5kZ4eDhnTZ7klztxcOjztg6tQZ2qnczx48cZMWKERU/usrVthFW05P+KaD8lFUbGv7yVq0f3Z/owHx5edYj1j0Uysr+brUNr3oFH4NwHsCQD7HvbOppOYc6/duLlYt+m0f0m/ec/8MQTEBMDYWHtcw/RpRUXF7N3714OHDhAVVUVERERLFq0qLPs5mt0jroL/PnYMEmYhLC9jfEZlFaaWBruh5NBm/ZIzi3p/MmUqRKSv4AB10giVe10ZhFnsoq549qml1i0mtkMK1bA5MmSSIkGHThwgM2bN2MymQgJCSEyMpI+ffoAdV/zm1iW8lD7x9iYLptMCSFsb82hFAI8nRkf5EFhubYhIzm31MZRWSB9I1TkwMDbbR1Jp7E+LgNFgXkhvu1zgy1b4PRpaGTDjOiZ8vPzMRgMuLi44OnpWZtEeXl52Tq0FpFkSgjRKil5pew9l8MTs4ehKAq9nQy4OxtIyukCydT5leDgA/3m2TqSTmNDfDoRgR70dWun3b/Ll2sLqZvZiCN6htzcXHbt2kVsbCwTJ04kKiqKwYMHN7k+uTOTZEoI0SprD6WiqnB92IDaY4Gezp1/ZKoyH1K/hyH3g85g62g6hbMXizmRUcSziy1vbdUi58/DDz/AM89oxRhFj3Xx4kV2795NXFwcer2eiIgIJk6caOuw2kySKSFEi6mqyjeHU5k40BN/T+fa4wFeLhy9kG+7wCyR/DWYK2SKr44NcVo1mwWj2mmK7623QKeDX/2qfa4vuoydO3dy6tQpJk2axJQpU+jVq5etQ7IKSaaEEC12KDmP89klPDSj/pB8oKcz6+PSqTKZMehtXZmmEec/Bbfh4Blh60g6jfVxGYQFuNOvt5P1L15WppVBWLIEBgxo/nzRraSlpREdHc2sWbPo06cPc+fOZcGCBfUKUXcHkkwJIVpsdUwqTgY9C0b1q3c8wMsZk1klLb+MQK9O+Muy+Dxc3A2jX+ryjV+tJTG7hIT0Qp5Z1E6lRr74AnJz4ZFH2uf6olO6cOEC0dHRnDlzBkdHR3JycujTpw+9e3fP3bOd9E9H28nIyODmm29m8ODBhIeHs3DhQk6dOtXkc2qGKdPS0mqrnH/88cc80sZfHq+99hqlpS1bf7Jjxw4WL17c4GOHDx/mnnvuaVN8R44cYf369S1+Xl1Tpkxp8vHf/va3bNu2rU33EO2nvMrE/46msSDU94rinAHVU36ddhH6+f9qb4NutW0cnciGeK0uz/zQdpjiU1V44w0ICYHp061/fdHpqKrK559/zocffkhaWhqzZ8/miSee6PZ1AbtuMuXrq/1lefk/39b/QlBVlSVLljBjxgzOnj1LTEwMf/3rX8m0sBN1//79Wb16dYvuZzabG328NclUU/7yl7/w2GOPteka1kim9uzZ0+Tjjz76KK+88kqb7iHaz+aETIoqjNwQfmVbjEAvLZnqlIvQVRUSV0Kfq6BXkK2j6TQ2xKczxq83fh7OzZ/cUvv2weHD2qiUjAR2W6qqkpycjKqqKIqCn58fUVFRPP7440ybNq22p2131nWTqcYSHAsTn4Zs374dg8HAAw88UHtszJgxREZGUlxczOzZswkLC2PUqFF89913Vzw/MTGR0NDQ2o8vXLjAjBkzGDp0KC+88ELtOcOHD+eOO+4gNDSUCxcu8OCDDxIREUFISAjPPfccAK+//jppaWnMnDmTmTNnArB582YmT55MWFgYN954I8XFxQBs3LiR4OBgwsLC+Oabbxr83IqKioiNjWXMmDFXPPb1118TGhrKmDFjmF7912N5eTm//OUvGTVqFOPGjWP79u1UVlby7LPP8uWXXzJ27Fi+/PJLcnNzue666xg9ejSTJk0iNjYW0Fre3H333cyYMYNBgwbx+uuv196v7oLDV199lVGjRjFmzBiefvppAAIDA8nJySFDWit0SmtiUhjg7sSkQVfWgenr6oi9na5zJlM5+6HotCw8r+NCbimxKQUsvGy61mqWLwc3N7jttva5vrApVVU5ceIE77//Ph999FFta7PIyEgmT56MfQ/audl510w98QQcOdK6586Y0fDxsWPhtdcafVp8fDzh4Q03k3R0dGTt2rW4ubmRnZ3NpEmTuOaaa1Ca+Gtr//79xMfH4+zszPjx41m0aBHe3t6cPn2aTz75hEmTJgHw8ssv4+npiclkYvbs2cTGxvLYY4/xr3/9i+3bt+Pt7U12djYvvfQSW7duxcXFhVdffZV//etf/O53v+O+++5j27ZtDBkyhJtuuqnBWA4ePFgv0avrxRdfZNOmTQwYMID8/HwAVqxYgaIoxMXFceLECaKiojh16hQvvvgiBw8eZPny5YA2ijRu3Di+/fZbtm3bxh133MGR6u/biRMn2L59O0VFRQwfPpwHH3wQg+HSVvQNGzbw3Xff8fPPP+Ps7Exubm7tY2FhYfz0008sXbq00a+v6HiZheXsOn2Rh2YMQae78v++TqcQ4OlMUk6JDaJrxvmVoHcEf6lzVGNDfPUuvtB2SKYyM+Hrr+Ghh6Cb7NgSGlVVSUhIYNeuXWRmZuLh4cHVV19NYGCgrUOzma47MtXBVFXlj3/8I6NHj2bOnDmkpqY2O/03d+5cvLy8cHJy4vrrr2f37t2ANvJSk0gBfPXVV4SFhTFu3DiOHTtGQkLCFdfat28fCQkJTJ06lbFjx/LJJ5+QlJTEiRMnGDhwIEOHDkVRFG5r5C/A9PR0fHx8Gnxs6tSp3HXXXbz33nuYTCYAdu/eXXut4OBgAgMDG1w7tnv3bm6/XftLf9asWeTk5FBYWAjAokWLcHBwwNvbmz59+lzx9dq6dSu//OUvcXbWphc8PT1rH+vTpw9paWkNxitsZ+3hVMwqLG1giq9GoKdz51szJe1jGrQ+LoPQAW4EeLXDFN9770FVlZZMiW7FaDSyceNGjEYj1113HY888ghhYWHo9Xpbh2YznXdkqokRJKDp+fcdO1p1y5CQkEbXPH322WdcvHiRmJgYDAYDQUFBlJeXNxOi0uDHdbeEnj9/nn/84x8cOHAADw8P7rrrrgavq6oqc+fO5fPPP693/IiFo3dOTk6Nxvv222/z888/88MPPxAeHk5MTIxF12xO3XlyvV6P0Wi0+Lnl5eU4ObXDNm3RaqqqsiYmhfBADwZ6N75TL8DLmb3ncmrXT3QK0j7mCqn5ZRy5kM9T84Zb/+JGI7z9NkRFwbBh1r++6FAmk4nY2Fji4+O55ZZbMBgM3HXXXXh4eKDTyZgMyMhUPbNmzaKiooJ333239lhsbCy7du2ioKCAPn36YDAY2L59O0lJSc1eb8uWLeTm5lJWVsa3337L1KlTrzinsLAQFxcXevfuTWZmJhs2bKh9zNXVlaKiIgAmTZrETz/9xJkzZwAoKSnh1KlTBAcHk5iYyNmzZwGuSLZqjBgxova5lzt79iwTJ07kxRdfxMfHhwsXLhAZGclnn30GwKlTp0hOTmb48OH1YgLqnbdjxw68vb1xc7Osye3cuXP56KOPahfZ153mO3XqVKPTksI2YlMKOJ1VzNKwxkelQBuZKq00kV1c2UGRWUDax1yhplBnu6yX+u47SE2VcghdnNFo5MCBA7zxxhusW7eOsrKy2rW6Xl5ekkjV0XlHpprTt2/Di8379m31JRVFYe3atTzxxBO8+uqrODo6EhQUxGuvvcatt97K1VdfzahRo4iIiCA4OLjZ602YMIGlS5eSkpLCbbfdRkREBImJifXOGTNmDOPGjSM4OBh/f/96Cdf999/P/Pnz6d+/P9u3b+fjjz9m2bJlVFRUAPDSSy8xbNgw3n33XRYtWoSzszORkZH1kp0awcHBFBQUUFRUhKura73HnnrqKU6fPo2qqsyePZsxY8YQHBzMgw8+yKhRo7Czs+Pjjz/GwcGBmTNn8sorrzB27Fj+8Ic/1C40Hz16NM7OznzyyScWf73nz5/PkSNHiIiIwN7enoULF/KXv/yFqqoqzpw5Q0SEFFXsTNYcSsHBTsei0U2/+AbU7ugrwce1E+zikfYxDdoQn8GIfm5NjjK22vLlEBQECxda/9qiQ+Tl5fHRRx9RVFSEn58fixYtYsiQIZ1ntLmTUVRVtdW9r7jx8ePHu30tClv697//jaurK/fee6+tQ2nS2rVrOXToEH/+858bPUf+r3SsCqOJiX/5kcihPryxbFyT557JKmbOv3by75vGsGRc06NYHeLMe7D/fpi3H7zG2zqaTiGjoJxJf/2R38wdxqOzh1r34vHxMGoUvPoq/O531r22aFcVFRVkZmYSEBCAqqp8//33hIaGMnDgQEmiNI1+EWSMrgd58MEHu0S9D6PRyG9+8xtbhyHq2HY8i/zSKpaGNd8OxN/TCUXpRIU7z6+U9jGX2Vizi689pvhWrABHR6guECw6v/Lycnbu3Ml//vMfPv/8cyorK1EUhWuuuYZBgwZJImWBrjvNJ1rM0dGxduddZ3bjjTfaOgRxmTWHUujr5kDk0IZ3hNblYKenn5sjyZ0hmSpOhIu7pH3MZdbHZzCsby+G9LFyyYKCAli5EpYtA68r65CJzqWsrIy9e/eyf/9+KioqGDZsGJGRkT2qPpS1SDIlhGjSxaIKtp+8yL2RA9E3UFuqIQFeziR1hsKdidI+5nJZheUcSMzlcWtP7wF88gmUlMDDD1v/2sLqcnNz2bVrFyNGjGD69On4tqGDSE8nyZQQoknfHUnFZFa5oZldfHUFerrw44msdozKAqqqTfH1mS7tY+rYdCwDVW2HXXxmszbFN2kSNFL8WNhWYWEhe/bsQVVVFixYwIABA3j88cdxd3e3dWhdniRTQogmrTmUyhi/3gzt69r8ydUCvJzJLq6gpMKIi4ONfs3kHICiUzDiKdvcv5NaH5fBYB8Xhlp7im/rVjh1Cv77X+teV7RZfn4+u3fv5siRI5jNZsaNG1dbB04SKeuQBeiXycjI4Oabb2bw4MGEh4ezcOHCBit/11XTay4tLY0bbtBaVXz88cc80sYaK61pdLxjxw4WL17c4GOHDx/mnupFoZmZmSxevJgxY8YwcuRIFlZvYa77ObSXgwcPNtpwOSgoiOzsbCorK5k+fXqLCn0K6zuWVsDx9MImK543JMCzEzQ8TlwJOgcIkPYxNbKLK/j5fA4LR/Wz/qLiFSugTx9o598fomWOHj3KG2+8weHDhxk7diyPPvooV199tSwqtzIZmapDVVWWLFnCnXfeyRdffAFo/xEzMzMZZkEV3/79+zdaQb2x+6mq2mjhs9dee43bbruttt1KW/3lL3/hmWeeAeDZZ59l7ty5PP744wC1DYpb+jm0RkRERLM1pOzt7Zk9ezZffvklt94q611sZU1MKga9wtWj+7foeYFel5KpEf0sK+JqVeYqSPoC/K4Be/eOv38ntflYJma1HXrxJSbC99/DH/8IXWDHcHd38eJFAHx8fAgMDCQiIoKpU6daXFBZtFzXHZn6xhdWKVf++6b1C+i2b9+OwWDggQceqD02ZswYIiMjKS4uZvbs2YSFhTFq1Ci+++67K56fmJhYr2r3hQsXmDFjBkOHDuWFF16oPWf48OHccccdhIaGcuHCBR588EEiIiIICQnhueeeA+D1118nLS2NmTNnMnPmTAA2b97M5MmTCQsL48Ybb6ytRLtx40aCg4MJCwvjm2++afBzKyoqIjY2ljFjxgBarz4/v0ujDaNHj77icygtLeUXv/gFI0eOZMmSJUycOJGDBw8C2mjcU089RUhICHPmzGH//v3MmDGDQYMGsW7dOkDbbvvLX/6SUaNGMW7cOLZv3w7UHz3LyckhKiqKkJAQ7r33XurWPbvuuutqq6uLjldlMvPdkVTmjOiLh0vLdvcEemqFIG22oy9tI1RkQ1Dn373akdbHpTPQ24UR/SyfsrXIW2+BTge/+pV1rytaJCMjg6+//po333yTbdu2AeDu7s6CBQskkWpnXTeZKm+kyXBjxy0QHx9PeCMLJx0dHWuLSW7fvp3f/OY3NFfwdP/+/axZs4bY2Fi+/vrr2kTk9OnTPPTQQxw7dozAwEBefvllDh48SGxsLDt37iQ2NpbHHnustvL59u3byc7O5qWXXmLr1q0cOnSIiIgI/vWvf1FeXs59993H999/T0xMDBkZGQ3GcvDgwXqJ3sMPP8w999zDzJkzefnllxtsKvzmm2/i4eFBQkICf/7zn+v17CspKWHWrFkcO3YMV1dXnnnmGbZs2cLatWt59tlnAVixYgWKohAXF8fnn3/OnXfeeUV/wBdeeIFp06Zx7NgxlixZQnJycu1joaGhHDhwoMmvsWg/O09eJKekstn2MQ3p7Wygt5OBpNySdojMAokrwcEb+s+3zf07odySSvaey2FBqK91p3jKyuCDD+C668Df33rXFRZLS0vjiy++4J133uHs2bNERkZy9dVX2zqsHqXzTvPFPAF5R1r33K0zGj7uMRbCX2vVJVVV5Y9//CPR0dHodDpSU1PJzMxscivp3Llz8aqutXL99deze/durrvuOgIDA5k0aVLteV999RXvvvsuRqOR9PR0EhISakeKauzbt4+EhITadjOVlZVMnjyZEydOMHDgQIYO1bY533bbbfV6C9ZIT0/Hx+dSjaB58+Zx7tw5Nm7cyIYNGxg3bhzx8fH1nrN79+7aacDQ0NB6Mdnb2zN/vvZCNWrUKBwcHDAYDIwaNaq2Zc7u3bt59NFHAa2dTWBg4BXrz6Kjo2tH0xYtWoSHh0ftY3q9Hnt7+wZb4Ij2tzomBS8Xe64a3nxtqYYEejnbpnBnZT6krIMh90n7mDq2JGRgMqvW38X35ZeQkyPlEGygZhH5mTNnSEpKYsaMGUycOBFHR0dbh9bjdN5kygZCQkIaXS/02WefcfHiRWJiYjAYDAQFBV0xynK5y//6q/nYxeVSL6zz58/zj3/8gwMHDuDh4cFdd93V4HVVVWXu3LlXNDI+cuSIJZ8aTk5OV1zX09OTW265hVtuuYXFixcTHR3d6Mjc5QwGQ+3no9Ppaiur63Q6qy4ar6iokF8MNpBXUsmPJzK5Y3IQBn3rBrADPJ2JSy2wcmQWSF4N5gqZ4rvM+rgM/D2dCOlvxekeVYU33oCRI2HGDOtdVzRKVVUSExOJjo6uXR4yadIkJk6c2CU6XHRXnXeaL/w1mLOj8X9Naew5zYxKzZo1i4qKinojO7GxsezatYuCggL69OmDwWBg+/btJCUlNfspbNmyhdzcXMrKyvj222/rNTGuUVhYiIuLC7179yYzM5MNGzbUPubq6lrbtHjSpEn89NNPnDlzBtCm2U6dOkVwcDCJiYmcPXsW4Ipkq8aIESNqnwuwbdu22p2CRUVFnD17loCAgHrPmTp1Kl999RUACQkJxMXFNfs51xUZGVm75unUqVMkJyczfPjweudMnz6dVatWAbBhwwby8vJqH8vJycHb2xuDQUYXOtr3sWlUmdRWTfHVCPRyJjWvDKPJbMXILJC4ElyHSR++OvJLK/npTDYLQ628i+/nn+HQIXjkEakw385UVeX06dN89NFHfPrpp2RnZ9cuNbG3t5dEysZkZKoORVFYu3YtTzzxBK+++iqOjo4EBQXx2muvceutt3L11VczatQoIiIiCA4ObvZ6EyZMYOnSpaSkpHDbbbcRERFROwVWY8yYMYwbN47g4GD8/f3rJVz3338/8+fPr1079fHHH7Ns2TIqKioAeOmllxg2bBjvvvsuixYtwtnZmcjIyNoErK7g4GAKCgpqp8xiYmJ45JFHsLOzw2w2c++99zJ+/Ph68T300EPceeedjBw5kuDgYEJCQujdu7fFX8+HHnqIBx98kFGjRmFnZ8fHH398xQ/8c889x7JlywgJCWHKlCn1Errt27ezaNEii+8nrGd1TAoj+rkxsg2jGAGezhjNKmn55QR4WWdHarOKEyErGkb/WV7c69iSkImxPab4VqwANzfoAm2quro1a9Zw7NgxevfuzcKFCxk3bhx2dvIS3lkozS2ibkdX3Pj48eOMGDHCsmd/49vwYnPHvnB9w4uwe7p///vfuLq6cu+991p0vslkoqqqCkdHR86ePcucOXM4efJkh/Vtuv7663nllVcaLEvRov8rokVOZxYx99/R/N/ikdwzbWCrr7P3bA7L3tvHf++ZyLSh3laMsAnxL0PsM3DNeal6XsfdHx/gZEYRu38/03ojU5mZEBCg7eB7/XXrXFPUMpvNnDhxgiFDhmBvb8/x48cpKytjzJgx6PV6W4fXUzX6w9N101pJmFrswQcf5Ouvv7b4/NLSUmbOnElVVRWqqvLmm292WCJVWVnJddddZ1F9L2Fdqw+lYKdTuHZsy2pLXa6m1lRSbgnT6IBkSlW1KT6fSEmk6igsr2LX6YvcOTnIulN8778PlZXw0EPWu6bAbDYTFxfH7t27yc7OZvHixYSHh8sfj51c102mRIs5OjpyewuG411dXWvLOXQ0e3t77rjjDpvcuyczmsysPZTKjOE+ePdq2xoMXzdH7O10HVdrKvcgFJ6ECb/pmPt1ET8ez6TKpLLAmlN8RiO8/TbMnQsWLHkQzVNVlcOHD7N7927y8vLo06cPN9xwgyRRXYQkU0KIWrvPZJNVVMENLWwf0xCdTsHfw6njyiOcr2kfc2PH3K+L+CE2g369HRnn7269i65bBykpsHy59a7ZQ9WUNwCt44ajoyM33XQTw4cPl5YvXUinS6bq/scSoiE2XOfX7a05lIq7s4GZwX2scr1ALxeSOqI/X037mAFXS/uYOorKq4g+fZFbJwag01nx9+ry5dp6qUb6gIrmVVVVcfDgQQ4cOMDdd99Nr169uPnmm3F0dJTXwC6oUyVTjo6O5OTk4OXlJf+ZRINUVSUnJ0dqT7WDgrIqNh3L4Obx/jjYWWeBa4CnMz+fy2n/P5LSN0HFRRgou8rq2nYii0qj2bq7+I4dg+3b4ZVXQBZCt1hFRQUHDhxg7969lJaW1tYs7NWrF05OTrYOT7RSp0qm/Pz8SElJqW3SKERDHB0d6/UVFNbxQ2w6lUazVab4agR4OlNSaSKnpLLNa7CadH4lOHhBP2kfU9eGuAz6uDoQHuDR/MmWevNNrZnxPfdY75o9REVFBa+//jqlpaUMHjyY6dOnX1HfT3RNnSqZMhgMDBzY+q3YQojWW3MohaF9ejFqgOW1xJpTs6MvObe0/ZKpygJI+Q4G3wv6jtlt2hWUVBjZfjKLm8b7W2+Kr6AAPvkEbr4ZvDuo3EUXV1paytmzZ2vbbk2dOpXAwEAGDBhg69CEFXWqZEoIYRvnLhYTk5TH0wuCrTodV5tM5ZQSZs3RkbouVLePkSm+erafzKLC2lN8n34KJSVaxXPRpOLiYvbs2cPBgwcxGo0EBgbi5ubGlClTbB2aaAeSTAkh+OZQKjoFloyz7l/Lfh7OKArtu6Pv/EpwHQpeE9rvHl3QhrgMvHvZMz7I0zoXVFWt4vnEiRARYZ1rdkOlpaXs3LmTQ4cOYTKZCA0NJTIyEjc3K/ZEFJ2OJFNC9HBms8o3h1KIHOpDXzfrLux3NOjxdXMkKbfEqtetVZIEWTth1IvSPqaOskoT205kcX3YAPTWmuL78Uc4eVIbnRJXMJvN6HQ6VFXl6NGjhIaGMm3aNLy8vGwdmugAFiVTiqLMB/4D6IH3VVV95bLHA4BPAPfqc55WVXW9dUMVQrSHvedySCso5+mF7VMcMMDTuf0KdyZqjbQZeFv7XL+L2nkqi7Iqk3Wn+JYvBx8fuFHqeNWVk5PD7t27yc3N5a677sLFxYVf//rX0ni4h2k2mVIURQ+sAOYCKcABRVHWqaqaUOe0Z4CvVFV9S1GUkcB6IKgd4hVCWNmamBRcHe2IGtm3Xa4f6OXM9pPtsENXVbUpPp9p0Es2rtT1Q1wGni72TBxopSm+pCT4/nt4+mmQsiQAZGVlsXv3buLj49Hr9YSHh2MymbCzs5NEqgeyZGRqAnBGVdVzAIqifAFcC9RNplSgZkK4N5BmzSCFEO2juMLIhvgMrhs3AEdD+9QMCvB05mJRBaWVRpztrbiyIDcGCk/AhHesd81uoLzKxLbjmVwztj92ep11Lvr229rbBx6wzvW6uNOnT7Nq1SoMBgOTJ09m8uTJ9OrVy9ZhCRuy5DfbAOBCnY9TgImXnfM8sFlRlEcBF2COVaITQrSr9XHplFWZuCG8/bZpB3i5AHAht4zhvq7Wu/D5laCzl/Yxl4k+dZGSShMLQq00xVdeDu+9B9deC/7+1rlmF5SamkpZWRlDhgwhKCiImTNnEhERgbOzs61DE52Atf5MXAZ8rKrqPxVFmQysVBQlVFVVs5WuL4RoB2tiUhjo7dJ+ZQuAQE/txSYpp8R6yZS5CpI+r24f036xd0Ub4jPo7WRg8mArLXz+8kvIyemx5RCSk5OJjo7m7Nmz9OvXjyFDhmAwGJg+fbqtQxOdiCXJVCpQ988Rv+pjdd0DzAdQVXWvoiiOgDeQZY0ghRDWdyG3lJ/P5/LbqGHt2uqlbuFOq0nfLO1jGlBhNLE1IZP5ob4YrDXFt3w5jBgBM2da53pdREpKClu3biUpKQkXFxfmzJlDhJSEEI2wJJk6AAxVFGUgWhJ1M3DLZeckA7OBjxVFGQE4AtITRohObM2hFBQFloS1b2sed2d73BztrFtrqrZ9zALrXbMb2H06m6IKIwtHW2mKb/9+OHhQqy/VA0pPqKqK2WxGr9dTXFxMTk4O8+bNIzw8HIPBYOvwRCfWbDKlqqpRUZRHgE1oZQ8+VFX1mKIoLwIHVVVdB/wGeE9RlF+jLUa/S1VVtT0DF0K0nqqqfHMolSmDvRjg3v7NVQO9XEiy1shUZQGkfgeD7pb2MZdZH5eBq6MdUwdbqdXL8uXg6gq3d+8RQFVVOXnyJNHR0QwdOpSZM2cyfPhwhgwZgp2dlGMUzbPof0l1zaj1lx17ts77CcBU64YmhGgvBxLzSM4t5Yk5QzvkfgFezhxLLbDOxS6sAVO5TPFdptJoZktCBnNH9sXezgpTfFlZ2nqp++/XEqpuyGw2k5CQwK5du8jKysLDwwPv6p6DiqJIIiUsJv9ThOiBVsdcwMVez/xQ3w65X4CnM5viMzCazG3frl/bPubyTcU9256z2RSWG1lorV18H3wAlZXw8MPWuV4ntGHDBg4ePIi3tzdLliwhNDQUnc5Ka81EjyLJlBA9TFmlifVxGSwc1c+6dZ+aEOjpjNGskl5Qjr9nG7aSlyRD1g4Y9UKPWMPTEhviMujlYMe0oVaY4jMa4a23YPZsCA5u+/U6CZPJxNGjRwkKCsLT05Pw8HCCgoIYMWKEJFGiTSSZEqKH2XQsg+IKI0vD23fheV0BdXb0tSmZkvYxDaoymdmUkMGcEX2sU3z1++/hwgV4/fW2X6sTMBqNHD58mJ9++omCggJmzJjBVVddha+vL76+HTM6K7o3SaaE6GFWx6Tg5+HEhCArtRqxQGB14c6knFKmDmnlRWrbx0yFXoOsF1w3sO9cDvmlVSywVi++5cshIAAWL7bO9Wxo//797Nq1i+LiYvz9/Vm8eDGDBw+2dViim5FkSogeJC2/jJ/OZvPYrKHodB03Tebr5oi9XkdSbknrL5J3CAqPw/i3rRdYN7E+LgNnez1XDfNp+8WOH4dt2+Cvf4UuugC7qqqqtpRBVlYW3t7eXH/99QQFBbVrTTXRc3XNnxQhRKusPZyKqsLSdq4tdTm9TsHP04nkttSaqmkfE/gL6wXWDRhNZjYfy2BWsJWm+FasAHt7uOeetl+rg5WVlbF//3727dvHsmXLCAgIYMGCBej17dN3UogakkwJ0UOoqsqamBQmBHnWrmHqSAGezq0v3Gk2VrePWSztYy6z/3wuOSWVLLTGFF9hIXzyCdx8M/hYYZSrg5SWlrJ3714OHDhARUUFw4cPx9HREUASKdEhJJkSooc4fCGfc9klPHCVbdaLBHo6czAxD1VVWz7Vkr4ZyrMgSGpLXW59fDpOBj0zh/dp+8U+/RSKi7tUHz6z2cy7775LQUEBI0eOJDIyUhaViw4nyZQQPcSamBQcDToWjLLNC02AlwvFFUbySqvwdGlh5fLElWDvCf0Xtk9wXZTJrLIxPpOZwT442bdxBEZVtSm+8eO1f51YQUEBhw8fZvr06eh0OubPn4+Xlxc+XWg0TXQvkkwJ0QOUV5n4/mga80N8cXW0TY+xwOqSCEk5JS1LpqoKIeVbGPRLaR9zmYOJuWQXV7DAGoU6t22DEye0ab5OKi8vj927d3PkyBEAhg4dyoABAwjuRrWwRNckyZQQPcDW45kUlhu5IdzfZjEE1qk1NS6gBeuekqvbx8gU3xU2xGfgYKdjVrAVpviWLwdvb/hF51vgX15ezqZNmzh69Cg6nY6wsDCmTp2Ku7u7rUMTApBkSogeYU1MCv16OzJ5sJfNYvCvHZlq4SL0xJXQawh4T2qHqLous1llQ3w6M4b74OLQxl/lycmwbh38/vdQvXC7MygvL8fR0RF7e3vS09OZMGECU6ZMwc3NzdahCVGPJFNCdHNZheXsPHWRB64ajL4Da0tdztGgx9fNsWXJVMkFyNwBo56T9jGXOZScR2ZhhXV28b1dXbvrgQfafi0rSE9PZ9euXSQnJ/PYY49hb2/P/fffLy1fRKclyZQQ3dy3R1Ixq3Ro+5jGBHg6k9ySwp2JnwEqBEn7mMutj8vAXm+FKb7ycnjvPbjmGq3quQ2lpKSwa9cuTp06hYODAxMmTEBVVQBJpESnJsmUEN2YVlsqlXEB7gz26WXrcAjwcib61EXLTlZVbYrPewq4SvuPumqm+KYP8277hoKvvoLsbJuXQ0hPT+eDDz7AycmJmTNnMmHChNpaUUJ0dpJMCdGNHUsr5GRmES9dF2rrUABtR19WUQVllabmt/LnHYaCBBj/VscE14UcTcknvaCc30YNb/vFVqyA4GCYNavt12oBVVU5f/48OTk5jB8/Hl9fX6677jqCg4NxcHDo0FiEaCtJpoToxlbHpGBvp+Pq0f1tHQpAbeX1C3mlDOvr2vTJNe1jAjrf7jJbWx+XjkGvMGdk37ZdaP9+7d8bb3TYmjRVVTlz5gzR0dGkpKTg4eFBWFgYer2eMWPGdEgMQlibJFNCdFOVRjPfHUll7si+9Ha2TW2pywV6uQDajr4mk6ma9jH9F4GDZwdF1zWoqsr6uAymDfGmt1Mbv68rVkCvXnDHHdYJrhmpqan88MMPpKen07t3bxYtWsTYsWOl5Yvo8iSZEqKb2n4yi7zSKm7o4KbGTalbuLNJGVugPBMGSm2py8WlFpCaX8bjc4a27UIXL8KXX8K990I7lhowm81UVFTg5OSEg4MDlZWVXHPNNYwePVqSKNFtSDIlRDe1OiYFH1cHIod62zqUWu7OBlwd7UjObaY8wvmVWkNjaR9zhfVxGdjpFKLaOsX3wQdQUQEPPWSdwC5jNpuJi4tj165d9OnTh1/84hd4e3vz8MMPt7w3oxCdnCRTQnRDOcUVbD+Rxd3TBmKn7zxbyhVFIcDTuelaU1VFWvuYgXeCXhYi16Wq2i6+yYO9cHduQ2sdoxHeektbdD5ypPUCBEwmE0eOHOGnn34iLy+Pvn37Ehp6aQOEJFKiO5JkSohuaN3RNIxmlaWdaIqvRqCXMyfSixo/4cIaMJXJFF8DjqUVkpRTyoNXtbFUxP/+p1U9f+01q8RV108//cT27dvp378/8+bNY9iwYZJAiW5PkikhuqHVMSmMGtCb4b7N7JizgQBPF7YkZGIyqw1XZD+/EnoNBu/JHR9cJ7chPh29TiEqxLdtF1qxAvz94eqr2xxTZWUlMTEx9O3bl0GDBhEeHk7//v0ZPHiwJFGix+g84/9CCKs4kVHIsbRCloYNsHUoDQr0cqbKpJJeUHblg6UpkLldq3guL8T11OzimzTIE0+XNkzxHT8OW7dqrWPsWv/3dEVFBbt27eI///kPmzdv5tSpUwC4uLgwZMgQSaREjyIjU0J0M2tiUjDoFa4Z20mTqeodfck5pfh5ONd/sKZ9zEBpH3O5k5lFnM8u4Z5pA9t2oTffBHt7bRdfK+3fv5/t27dTXl7OkCFDmD59Ov7+/m2LS4guTJIpIboRo8nM2sNpzAru07bRCwBfX8jMvPJ4376QkdHqy9YU7kzKLWVK3QdUVZvi854MrkNaff3uan1sOjoF5rVliq+oCD75BG66Cfq0rKdfSUkJDg4O2FWPZgUFBREZGUn//p2jIKwQtiTJlBDdSPTpi2QXV1hn4XlDiVRTxy3Ur7cTBr1y5Y6+vCNQcAzGv9mm63dX6+MzmDDQEx/XNuxwXLlSS6geftjipxQVFbF3714OHjzIvHnzCA8PZ/z48UyYMKH1cQjRzUgyJUQ3siYmFU8Xe2YMb9moQ0fS6xT8PJxJzr2scOf5laAzSPuYBpzOLOJMVjF3TA5p/UVUFZYvh4gIsCARKigo4KeffuLQoUOYzWZGjRpFYGAgIOUNhLicJFNCdBP5pZVsScjk1kkB2Nt17r0lAZ7O9Qt3mo2QtKq6fYyX7QLrpNbHZaC0dYpv+3Zt8fnHH1u0uH/16tWkpaUxevRoIiMj8fSUtj5CNEaSKSG6ie9j06k0mTtlbanLBXo5cyg5D1VVtVGOjK3SPqYJG+LTiQj0oK+bY+svsmIFeHlp66UakJOTw549e5gzZw5OTk4sXLgQJycn3N3dW39PIXoISaaE6CbWxKQQ7OtKSH8r9Fkzm9t+jSYEeDpTVG4kv7QKDxf7Ou1jFrXrfbuisxeLOZFRxHNXt6FSeXIyfPstPPUUONZPyLKysti1axfHjh1Dr9cTHBzM0KFD6devX9sCF6IHkWRKiG7gTFYxRy7k88yiEdZZz/LPfzb+WN829oQDAr1cAG1Hn4d9BaSshYF3SPuYBmyISwdgfmgbpvjeeUd7+8ADtYdMJhNr1qzh+PHj2NvbM2XKFCZPnoyLi0tbwhWiR+rcCyuEEBZZcygFvU7hWmvUltq7F/74R7jhBm2ESlXhyBHtsY8+alNZhBqBNeURckrgwjfSPqYJ6+MyCAtwp19vp9ZdoKIC3ntPq3YeFEReXh4Aer0eg8HA9OnTefzxx5kzZ44kUkK0kiRTQnRxJrPK2kOpXDXMp23b5gHy8uDmm7VWI++9d2mh8ujR2ojUpk1tDxjw97hUuFNrHzMIvKc086yeJzG7hIT0QhaOasOU29dfw8WLJN12GytXruSNN94gNzcXgCVLljBz5kycnZ2buYgQoikyzSdEF/fTmWwyCst5ti1rakAbgbr7bkhPh59+groLjxUFoqJg/XpttErXtr/DnOz19HF1ID/nPBi3Qeiz0j6mARvitVHA1k7xqarK+c8/J/qhh0g6dgwXFxfmzJlDr169rBmmED2eJFNCdHFrDqXQ28nA7BFtrC21fLm2SPlf/4Lx4698PCpKK/p4+DCEh7ftXmhTfQNLPgMHaR/TmPVx6Yzxd7+y7Y6FSvbsYVVYGM4GA/PnzSMsLAyDwWDlKIUQkkwJ0YUVllex6VgGN4T74WCnb/2FYmLgt7+FxYvhiScaPmfuXO3tpk1WSaYCPJyZVLZe2sc04kJuKXGpBfxhQbDFz1FVlRMnTnD+/HkWLlxIr/ff5/a9exmwZw92UidKiHYjyZQQXdj62HTKq8zcEN6GJrOFhZd6tTVV0LFvXxg7FjZv1haot1F47ySGmBKpCvgtMlZypQ3x2i6+BaHNr5cym80kJCSwa9cusrKy8PT0pCwlBafPPyfw7rtBEikh2pUkU0J0YWsOpTDYx4Uxfr1bdwFVhfvvh8RE2LFDK+rYlHnztLIJRUXg6tq6e1abwHoqzXZc6LWIwW26Uve0Pi6D0AFutY2hG5OVlcVXX31FTk4OPj4+XH/99YSEhKD7+9+1nXwt6MMnhGgd2c0nRBeVmF3CgcQ8lob7tb621Pvvw5dfwp//DNOmNX9+VBQYjVri1RZmI4FF37G9KILzha3c8t+NpeaXceRCfqO7+IxGIzk5OQD07t0bNzc3brzxRh588EFGjRqFTlXhrbdg5kwIaUM/PyGERSxKphRFma8oyklFUc4oivJ0I+f8QlGUBEVRjimKssq6YQohLvfNoRR0Clw/rpXtY+Li4LHHtLVQv/+9Zc+ZOhWcnbWpvrbI+BFDVSbf5M0iqW6PPgFcKtR5+RRfVVUV+/fv54033uCLL77AbDbj4ODAHXfcwciRIy8l1T/8AElJMiolRAdpdppPURQ9sAKYC6QABxRFWaeqakKdc4YCfwCmqqqapyhK521ZL0Q3YDarrDmUytQh3vj2bkW/tpIS+MUvtPIHK1daXurAwQFmzGh7vanElaj2HuyvnIRvTknbrtUNbYjPYEQ/NwZ6a0U0KysrOXjwIHv37qW4uJiAgACmT5/e+Ijk8uXg5wfXXtuBUQvRc1nyG3QCcEZV1XOqqlYCXwCX/4TeB6xQVTUPQFXVLOuGKYSo6+fzuaTml3FDeCtHpR55BE6ehM8+a3l7mKgoOH0azp9v3b2riuHCWpSAX9DPw51kGZmqJ6OgnJikPBbWqS118uRJtmzZgo+PD3feeSd33XUXgwcPbjiZOnECtmzRWsfYybJYITqCJT9pA4ALdT5OASZeds4wAEVRfgL0wPOqqm60SoRCiCusjknB1cGOqJGtKOb46afarr1nn4VZs1r+/HnztLdbtmiL11vqwjdgKoWBtxN4ypGTmUUtv0Y3tjE+HXuMDCg/z/79eUyYMIGQkBA8PDzw87MgeX7zTbC3h/vua/9ghRCA9Rag2wFDgRnAMuA9RVHcrXRtIUQdJRVGNsSns2h0P5zsW1hb6sQJeOghuOoqLZlqjeHDtXYzrZ3qS7zUPibAy5mU3DJMZrV11+pmSkpKOLhnFzc5xRF3cC+ZmZkA6HQ6yxKpoiL45BO48Uat1IUQokNYMjKVCtQtYuNXfayuFOBnVVWrgPOKopxCS64OWCVKIUStjfEZlFaaWNrSKb6yMq2elJOTNr2nb2WRz5rWMqtXazv7WjKVVJoKGT9C6P+BohDo6UKlyUxGYTkD3Hv2rr5Dhw6xYcNGvKuqsPcO4J4bF9K3pVOw//2vVjfskUfaJ0ghRIMsGZk6AAxVFGWgoij2wM3AusvO+RZtVApFUbzRpv3OWS9MIUSN1TEpBHo5ExHo0bInPvkkxMZq03wDBrQtiHnzoKAA9u9v2fMSVwEqBGntYwKraygl9dBF6AUFBRQXFwPg6emJo48/aytCuOHGG1qeSKmqtvA8PBwmXr4SQwjRnppNplRVNQKPAJuA48BXqqoeUxTlRUVRrqk+bROQoyhKArAdeEpV1Zz2ClqIniolr5S953JYGtbC2lJffQVvvw2/+x0sWND2QGbP1nYAtrREQuJK8JoEbkMBCPDUkqnknJ61CD0vL4/vv/+e119/nV27dgEQFBTEQYbi7e3N0D6taES8cyckJGjlEKRptBAdyqLxeVVV1wPrLzv2bJ33VeDJ6n9CiHay9pA2w75kXAtGls6e1RYjT5oEL71knUA8PbVmyJs3w/PPW/acvKOQHwcRK2oP9evtiJ1O6TG1prKzs9m9ezexsbHodDrCwsKYMmWK9lhxBT+fz+HhmUNaV4R1+XLt+3LzzVaOWgjRHNk3K0QXoaoqaw6lMHmQF/6eTbcYqVVRoa2T0ungiy/AYMUueFFR8PLLkJcHHhZMOZ5fCToDBN5Ue8hOr8PPw6nHlEfYvXs3x44dY+LEiUyZMgXXOi15Nh3LwKzSaNXzJl24AN9+C7/5jbYmTgjRoaSdjBBdRExSHok5pS1beP700xATAx99BIGB1g0oKgrMZti2rflzzSZIWgX9F4JD/f5/AV4u3XaaLz09nS+//JL0dK2i+axZs3jiiSeYN29evUQKYENcBgO9XQj2bUXPw3fe0b4XDz5ojbCFEC0kyZQQXcSaQyk42+tZEGphbal16+C117SWMdddZ/2AJk4ENzfLSiRk/ghl6RB0+xUPBXo6d7sF6CkpKaxatYp3332X8+fPk5ubC4CbmxsuLi5XnJ9bUsneczksCPVt+RRfRQW89x4sXgxBQVaIXgjRUjLNJ0QXUF5l4n9H01kQ2g8XBwt+bJOT4a67ICwM/va39gnKYNCKfm7erO0kayoJOL8SDO4wYPEVDwV6OVNYbiS/tBJ3Z/v2ibUDffXVVxw/fhwnJydmzZrF+PHjcXRsuuXPloQMTGa1dVN8q1dDVpaUQxDChiSZEqIL2HQsg6IKI0vDLVh4XlUFy5ZpNaC+/FLrp9deoqK0tTqnT8OwYY3EU6xVPR94G+ivjKVmR19STmmXTKZUVSU5OZmAgAAURcHPzw8/Pz8iIiKwt7fs8/khLoMAT2dC+ru1PIDly7Wv/Zw5LX+uEMIqJJkSogtYcyiVAe5OTBro1fzJzz4Le/bA55/DkCHtG1hNa5lNmxpPplLWau1jGpjiAwj00qa9knJLGePv3g5Btg9VVTl9+jTR0dGkpqZyyy23MHTo0NrdeZbKL61kz5ls7okc2PIpvpgY2LdPm861tFm1EMLqJJkSopPLKChn9+mLPDJzCDpdMy+2mzbBK69oPfM6Yov8oEEweLA21ffoow2fc34luAwEn6kNPuzvqe0+S+4i66ZUVeX48ePs2rWLjIwM3N3dWbx4MQMHDmzV9bYkZGI0qywMbcUU34oV4OICd97ZqnsLIaxDkikhOrm1h1Mxq3B9WDO7+NLS4PbbITRUG6noKPPmaf3gKiu1Brt1laZpi89D/tTomipnezt8XB26THkEs9nMpk2bsLOz49prr2XUqFHoW9uaB9gQn8EAdydG+/Vu2RNzcmDVKvjlL8HdvdX3F0K0nYwLC9GJ1dSWigj0IMj7yl1gtUwmuPVWKCnRqp13ZK2hqCjtvnv2XPlY0ipQzY1O8dXQdvR1zmTKZDJx5MgRPvnkE4xGI3q9njvuuIOHH36YsWPHtimRKiyvYtfpi63bxffBB9pOvocfbvX9hRDWISNTQnRiR1MKOJNVzCvXj2r6xJdegh074OOPYcSIjgjtkpkztWbHmzfDjBn1Hzu/Erwm1raPaUyAlzN7z3auDlRGo5GjR4+ye/du8vPz8fX1paioCA8PD7y8LFi7ZoGtCZlUmVQWjm7hFJ/JBG+9BVddpY1ECiFsSpIpITqxNTEpONjpmn6x3b4dXngB7rjDNmtn3Nxg8mRtvdZf/nLpeF4s5MdCxPJmLxHo6cLaw6mUV5lwNLR+pMdaCgoK+PDDDyksLGTAgAEsWLCAoUOHtq7NSxPWx2XQr7cjY/3cW/jE9ZCYCH//u1XjEUK0jiRTQnRSFUYT646mMS/EFzfHRtrAZGVp03vDhmmLkW0lKgr+7//g4kXw8dGOJa4ExQ4Cbmr6uWi1plRVa+Q8pE8rKoBbQWVlJenp6QQGBuLm5saQIUMYOXIkgwYNsnoSBVBUXkX06YvcOjGg+Y0Fl1u+HAYMgGuvtXpcQoiWkzVTQnRSPx7PoqCsihsaax9jNmujUbm52jqpXr06NsC6akokbN1aHZsJEqvbxzh6N/t0/zq1pjpaRUUFu3bt4rXXXmPVqlVUVFSgKApXX301gwcPbpdECmDbiSwqjeaWF+o8dUqbUn3gAev2WhRCtJqMTAnRSa2JScHXzZGpQxpJRv7+d21q7e23YfTojg3ucmFh4OmpxbNsGWRug7I0GNj0wvMagV4dn0yVlZXx888/8/PPP1NeXs7QoUOJjIzEoT2LnNaxIS6DPq4OhAdY0CS6rjff1JKo++5rn8CEEC0myZQQndDFogp2nLrIfZGD0Dc0BbRnD/zpT/CLX2g1pWxNr9cqcNe0ljm/Egy9G2wf0xAvF3tc7PUdWh6hoKCAnTt3EhwcTGRkJP379++we5dUGNl+Moubx/u3bIqvuFhrWn3jjdC3b/sFKIRoEUmmhOiEvjuSismsckND7WNyc7WCnIGB8O67TffE60jz5mnTjbH7IeUbCLwF9E33pKuhKAoBXi7tmkwVFRWxZ88eqqqqWLx4Mb6+vjz++OO426BG0/aTWVQYzSxo6RTff/8LhYXSh0+ITkaSKSE6odUxKYzxd79yMbaqakUaMzK00aneLSz02J7mztXe/vQfcC+xeIqvRqCnM6eziqweVkFBAbt37+bw4cOYzWbGjBmDqqooimKTRAq0KT7vXvaMD/K0/Emqqm0yGDcOJk1qv+CEEC0myZQQncyxtAJOZBTx52tDrnzw9ddh3TqtwnlERIfH1qhvfKE8Ez4D+Fw7tnU6OPaF6zMsukSglzPbTmZhNqst393WiPj4eNauXQvA2LFjmTZtGh4eLVyjZGVllSa2ncji+rABDU/hNiY6GuLjtWKdnWU0UggBSDIlRKezOiYFe72Oq8dctobn4EF46im45hp47DHbBNeY8syWHW9AgJczlUYzGYXl9HdvfQX37OxsTCYTffv2JSAggPDwcKZOnUrvTjKKt+NkFmVVJha1dIpv+XJtkf+yZe0TmBCi1SSZEqITqTKZWXckjTkj++DuXKfPXUEB3HQT+PpqC5A708iEarbKZQLqlEdoTTKVmZnJrl27OHbsGEOGDOHWW2/Fzc2NhQsXWiU+a1kfn4Gniz0TBrZgii8lBdauhSef7NhWQUIIi0gyJUQnsuPkRXJKKllat6mxqmo79pKStKkezxa8CLcHVYWCBK38QeZ2yNpplcsGemq9B5NzS5g82PJ2Lenp6URHR3PixAns7e2ZNm0akzrpmqLyKhPbjmdyzdj+2OlbUObv3Xe1umIPPth+wQkhWk2SKSE6kTUxKXj3cmD6MJ9LB999V9sl98orMGVKxwelqlB0WkucMrdB1g4oz9IecwkCv2vh3Edtvk1/d0fsdIrFO/pqFpGfO3eOxMRErrrqKiZOnIhTJx65iT51kZJKEwtCWzDFV1EB77wDixbBwIHtF5wQotUkmRKik8grqeTHE5ncOTkIQ82oRWwsPP64Vnbgqac6Lpji85eSp8ztWgFOAKcB4BsFfWdq/3pVv7hbIZmy0+sY4OHUZOFOVVVJSkoiOjqasWPHMnr0aCZMmEBERESHFdtsiw3xGbg7G1o08saaNVrbICmHIESnJcmUEJ3EuqNpVJlUlta0jyku1opyenrCp5+Crh27P5WmVCdP1QlUSZJ23LEP9KlOnPrOAtchDa/Xcuzb8GJzx5YVlgzwdG5wZEpVVc6dO0d0dDTJycm4uLigqioAhi7SUqXCaGJrQiYLRvleSpYtsWIFDBlyqfSEEKLTkWRKiE5izaEUQvq7MaKfm3bg4Yfh9Gn48Ufo08e6NyvLrF7vtB0ytkHxGe24vSf0nQHBvwXfWeA2wrLF7nXLH5jN0L8/zJoFq1a1KKxAL2e+P5p+xfFvv/2W2NhYXF1dmT9/PmFhYV0miaqx+3Q2RRXGlhXqPHRIqyf273+3bzIthGgTSaaE6AROZRYRm1LAs4tHagc++UQbjXr+eZgxo+03qMiBzB2XEqiCBO24wQ36XAXDHtJGn9xHg9LGF22dDqKiYMMGLbFqQRIQ6OlCQVkV+SWVpCefZdCgQTg4ODBy5EgCAgIYM2YMdnZd89fW+rgMXB3tmDq4+cbPtVasAGdnuOuudotLCNF2XfO3khDdzJqYFOx0CteO7Q/Hj8NDD2lJ1DPPtO6ClfmQFX1p6i7/qHbczgV8ImHgnVry5DEOdO3wayAqClauhMOHITzc4qf5eTgySJ/Dh++/Q1F+LgsWLGDChAkMHz7c+jF2oEqjmS0JGcwd2Rd7OwuTy5wcbWTvzjvBRpXahRCWkWRKCBszmsysPZzKzOA+eOnN2jopZ2f47DOtgbAlqorh4u5LC8bzDmn1n/SO4D0FRr+kJU9e40HXAdNjNet7Nm+2KJlSVZWjR4+SsGMnV9nnU2XyYOnSpYwcObKdA+0Ye85mU1huZGFLdvF99BGUl2vTvUKITk2SKSFsbNeZbLKKKrTaUk88obUM2bBBW3fUGGMZZO+5tGA85wCoRi1R8poEIf+nJU/eEy1uNmxVffvC2LGwaRP84Q+NnlZT3kBRFGJjY3F2dGBd1mBujJxIaOjQjou3na2PS6eXgx2Rwyyc4jOZ4M03Yfp0GDWqfYMTQrSZJFNC2NiamBQ8nA3MObpNqyn1+9/D/Pn1TzJVQM7Pl5Kn7H1grgRFD57jYeTvqpOnKWDnbJtP5HJRUdrC6aIicK3fsLmqqopDhw7x888/c+edd9K7d29uvPFGHB0d+eDlH7mQW2ajoK2vymRmc0Imc0b0wcHOwpHGDRvg/Hl49dX2DU4IYRWSTAlhQwVlVWxOyOQRf7B78AGYPBn+/GcwV0HOwUu77bL3gKkMUMAzDIY/ppUq8JkGBtdm72MTUVHwt7/Bzp2weDEAlZWVHDhwgL1791JSUkJgYCAVFRUAtcU2A72cScotsVnY1rbvXA75pVUt28W3fLk2Mnndde0WlxDCeiSZEsKG/hebBuUV3PfmczAI+PsM2H0tXNwFxmLtJPfRMOR+LXnqMx3s3W0YcQtMm6b1kdu0CRYvprKyktdff52SkhIGDRrE9OnTCQwMvOJpgZ7O7DuXY4OA28f6uAxc7PVcVbeqfVNOn9a+Zi+8AF2s/IMQPZUkU0LYgmqG/Dgq4j9kp8fXOD2SDi5A0l+12k4D76hOnq4CxxZspe9MHBwomz2bU6dOMQZq++b5+fnh5+fX6NMCvJxZeySVCqPJ8mmxTspoMrP5WAazRvTF0WDh5/Lmm1oSdf/97RucEMJqJJkSoiOoKhSeqNMceAdU5HC3E+AElITC3D9qBTOdWjAd1EmVlJSwZ88eDkZEUKmqBMTF4TFqlEUNiAM8nVFVuJBbxpA+vTog2vaz/3wuOSWVLAz1tewJJSXaLr4bbgBfC58jhLA5SaaEaA+qCsVn6/S32wHl1VXCnQNgwNVsO92HiN+9gZPvCAz79kAX6C3XnNLSUqKjo4mJicFoNBLq70/kH/6AR//+Fu9KC/TSFtAn55Z0+WRqfXw6TgY9M4ZbWMH+s8+goEDKIQjRxUgyJURLfePbcB86B28Y9/dLCVRpinbcqT/4zr7UHNhlIObKKvr8YiyGfAXD7q+6fCJlNpvR6XQoisLRo0cJCQlh2rRpeHt5wXPPafWmLJy2CvB0ASC5iYbHXYHJrLIxPpOZwT442Vswxaeq2sLzsWNhypR2j08IYT2STAnRUg0lUgAV2bDvl+Dgcylx6jsTXIdd0d8u7dHfEnrhODF/e4vwwYM7IOj2kZuby+7du8nKyuKee+7BycmJJ554Aoe6yeG8ebB6NRiNYEErGO9e9jjb60lqoOFxV3IwMZfs4goWWFqoc9cuiIuD99+3rB+iEKLTkGRKCGtaGAe9Q5p+MdywAb/33uCr8IVc88R9HRebFV28eJHdu3cTFxeHTqcjLCwMo9GIwWCon0iBViLhgw/gwAGt9EMzFEUhwNO5y49MrY9Lx8FOx6xgC6f4li8HDw9Ytqx9AxNCWJ0kU0JYk3to04+npqLecQen+gQR/9vn+YWlO7w6kXPnzrFy5UoMBgOTJk1i8uTJuLo2Uetqzhwtudy82aJkCrR1U2cvdt1aU2azyob4DGYM98HFwYJfs6mpsHYtPP641kpICNGltLE9vBDCYkYj3HILpuISHrrm91w3ZYitI7JYWloap06dAiAwMJCZM2fy+OOPExUV1XQiBeDpCePHa7WTLBTo5UJybilms9qWsG3mUHIeWUUVLLS0UOe772otZB58sH0DE0K0CxmZEqIl1Da8uP/5zxAdzdt3/R/q8GDG+btbLaz2cuHCBaKjozlz5gx9+vRh6NCh6PV6pk+f3rILRUXBX/4C+fng7t7s6f6ezlQazWQWldOvt1OrYrel9XEZ2OstnOKrrIR33oGFC6ELr58ToiezaGRKUZT5iqKcVBTljKIoTzdx3lJFUVRFUSKsF6IQnUjqusYfc+zb+GPbtsGf/0zJzbfyj74TWRruh9KJFxmnpaXx6aef8uGHH5KWlsasWbO4++67Wx/zvHlgNsOPP1p0eqBndXmELrhuSpviS2f6MG9cHS2oYP7NN5CZKeUQhOjCmh2ZUhRFD6wA5gIpwAFFUdapqppw2XmuwOPAz+0RqBA2ZyyBg49B71BYcAh0Frb6yMyEW2+F4cP5+JbfoOxJY8m4Ae0bayuoqorZbEav11NSUsLFixeJiooiPDwce3v7tl184kSt2fHmzbB0abOn19SaSsotZeIgr7bdu4MdScknvaCcp+YNt+wJy5drI1Lz5rVvYEKIdmPJNN8E4IyqqucAFEX5ArgWSLjsvD8DrwJPWTVCITqL+D9DaTLM2WV5ImU2w+23Q34+5g0b+WJzDlMHe9PfvfNMXamqyqlTp9i1axdBQUHMmTOHIUOG8Pjjj2NnQSkDixgMMHu2tm5KVZvd+t/f3Qm9TumSI1Mb4tIx6BVmj2hipLLG4cPw00/wr3+BTpawCtFVWfLTOwC4UOfjlOpjtRRFCQP8VVX9wYqxCdF55B+D4/+EQb+EPtMsf96rr8KWLfCf/3DAzY8LuWUsDe8co1KqqpKQkMA777zDF198QUlJCT4+WjNeRVGsl0jViIqCpCStkW8zDHodA9ydulytKVVVWR+XwbQh3vR2siDhXrFC2713113tHpsQov20+beloig64F/AXW2ORojOSFXh4ENgcIOxf7P8ebt3w//9H9x0E9x3H2vWxNLLwY55IZ2j59qmTZv4+eef8fLy4rrrriM0NBS9vh1LNURFaW83b4Zhw5o9PdDLmeScrlUeIS61gNT8Mh6fM7T5k3NzYdUquO02rb6UEKLLsiSZSgX863zsV32shisQCuyoXpzqC6xTFOUaVVUPWitQIWzm/ErIioYJ74Kjt2XPycnRii8GBcG771JaZeKH2HQWje6Hs71tNtGaTCZiY2Px9/fH29ubsLAw/Pz8GDlyJLqOmGIaPFj7t3kzPPJIs6cHeDrzQ1x6+8dlRevjMrDTKUSNtGCK76OPoKxMFp4L0Q1Y8lv9ADBUUZSBaEnUzcAtNQ+qqloA1L7CKIqyA/itJFKiW6jIhcO/Be/JMPgey56jqtq0TWYm7N0Lbm5sOpxCSaWJpWF+7RpuQ4xGI0eOHGH37t0UFBQQGRnJrFmz6NOnD336WFid21qiomDlSq0cQDOL2gM8nckvraKgrMqyKTMb06b40pkyxBt352YW7JtM8OabEBkJY8Z0TIBCiHbT7J+jqqoagUeATcBx4CtVVY8pivKioijXtHeAQtjU0T9CZQ6MfwsUC0dvXnsN/vc/+Mc/IDwcgDUxqQR4OjM+yLP9Ym3AwYMHef311/nhhx9wdXXllltuYebMmR0aQz1RUVBcrCWZzajZ0Xehi6ybOpZWSHJuKQtDLZjG3bgRzp2TUSkhugmL5htUVV0PrL/s2LONnDuj7WEJ0Qlk/wxn3oXhj4OHhaMH+/fD738P114Ljz4KQFp+GT+dzebx2UPR6dq/tlRlZSUGgwFFUcjOzsbT05PrrruOgQMH2r621axZoNdru/quuqrJUwM8XQBIyikldEDvjoiuTTbEp6PXKURZsiZuxQro1w+WLGn/wIQQ7U4qoAvRELMRDjwITv1g9IuWPSc/H26+WXuR/PDD2u3/aw+noqq0+xRfeXk5+/fvZ9++fdx4440MHDiQuXPntu+i8pZyc9P6823erFVEb0JAba2pzr8IvWYX36RBnni6NDPFd+YMbNgAzz/f7FSnEKJrkGRKiIacfhPyDsO0r8DQTO850NZJ3XcfJCfDrl1aPzq0F9k1MSlMHOiJv2f7NLAtLS1l37597N+/n4qKCoYNG4ZzdbPcTpVI1Zg3D559Fi5ehOpSDA3p5WCHdy/7LlFr6kRGEeezS7g3cmDzJ7/5JtjZwf33t39gQogOIcmUEJcrS4ejz4BvFPjfYNlz3n4bVq/W6kpNnlx7+FByPueyS3hgRvv0XFNVlffff5+8vDxGjBhBZGQk/fpZ2FzXVqKitJIRW7dqOx6bEODpTFIXSKY2xKWjUyBqZDNTfCUl2i6+pUu1EUwhRLcgyZQQlzv0JJgrYfyKZit1A3DkCPz61zB/Pvz2t/UeWnMoBSeDnoWjrPfCWVhYyKFDh5g+fTo6nY758+fj7u7e8TvzWis8XBu527y52WQq0MuF/edzOyiw1lsfn8GEgZ74uDo0feKqVdp0sAWlIYQQXYckU0LUlbEVkr6AUc+D65Dmzy8q0opyennBp5/WawlSXmXi+6NpLAj1pZdD23/U8vPz2b17N0eOHMFsNjNo0CACAgIYZkEBzE5Fr4c5c7RkqpnWMv6eznx7JJUKowkHu044ZQmcziziTFYxd0wOafpEVdX68I0ZA1OndkxwQogOIcmUEDVMFXDgYeg1BEb+vvnzVRUefFBbULxt2xXrf7YkZFJUbmRpeNsWnldUVLBx40ZiY2MBGDt2LNOmTcOjK1fNjoqCr76CY8cgNLTR0wI9nVFVSM0rY5BPrw4M0HLr4zJQFJjf3C6+3bshNhbefdeyEU8hRJchyZQQNRL+BkWnYOYm0Ds2f/7HH8Nnn8ELLzS4zX/NoRT693Zk8iCvVoVTXl6Oo6Mj9vb2ZGZmEhERwZQpU+jdu/OXCWhW3dYyTSVTtTv6SjtxMpXO+EBP+rg1839mxQpwd4dbbmn6PCFElyNtyoUAKDoLx16GgBuhX1Tz5yckaAUXZ86EP/3pioezCsuJPnWR68P8WlxbKjMzk6+//prXX3+diooKFEXh3nvvZcGCBd0jkQLw94cRI7R6U02oKY/QWXf0nckq5mRmEQtGNTMqlZYGa9bA3XeDi0vHBCeE6DAyMiWEqsLBR0FngLB/N3yOr6/WHuZy8fHaGqDLrD2cilmF68MGWBxGWloa0dHRnDx5Ent7eyZMmICqqgAd0zuvo0VFwTvvaP3pnJwaPMWnlwPO9vpOu6NvY7zWO3B+c1XP331XayHz4IMdEJUQoqNJMiXEhW8gfYOWSDk3kvw0lEiBVivpMqqqsuZQCmEB7hZPTV28eJH33nsPR0dHZsyYwYQJE3BqJMHoNubNg//8R1tLNHdug6coikKApzPJnbRw5/q4DMIC3OnXu4nvVWWlljTOnw9DLNjUIITociSZEj1bVRHEPA7uY2CYdbarx6cWciqzmL8sGdXoOaqqkpiYSFZWFhMnTsTHx4clS5YwfPhwHBya2V7fXUyfrlUA37Sp0WQKtFpT57M7XzKVmF1CQnohzywa0fSJa9dCRoaUQxCiG5NkSvRscS9AWSpM+xp01vlxWB1zAXs7HYtGX1lbSlVVzp49S3R0NBcuXKB3796Eh4djZ2fH6NGjrXL/LsPFBaZN0xahNyHA05mdpy5iNqsd0tvQUuurp/gWNFdDbPlyGDRIG5kSQnRLkkyJnisvFk6+BoPvA5/JzZ5uiUqjmXVH04ga2ZfeToZ6j6Wnp/O///2PtLQ03NzcWLhwIePGjcPOrgf/GM6bpzWGTkuD/v0bPCXQy5kKo5mLxRX0bW7HXAfaEJfBGH93Brg3McV39Kg2jfmPf9SrQSaE6F7kp1v0TKpZa2Rs7wFj/9r0uWazxZfddiKLvNIqbqiuLaWqKqWl2uJpR0dHKioquPrqq3nssccYP358z06k4FKJhC1bGj0lwEvb/daZFqFfyC0lLrWAhc0tPF+xQltc/8tfdkxgQgibkGRK9EznPobsPTD2b+DQTB2oF19s/LG+fet9uDomhT6uDkwd7EVsbCxvvvkm3377LQAeHh48/PDDhIWFdc4GxLYwerT2NWxiqi+wukF0Uk7nWTe1oXqKr8k2QXl58N//wq231ja+FkJ0T5JMiZ6nIgeO/A58psGgO5s+93//04py3nmnNkKlqvX/ZWTUnppTXEH0yQyu8avgrTdXsHbtWnQ6Xb21UIpUvq5Pp9MWn2/e3OgI4AAPJ/Q6heTczjMy9UNcBqMG9Ma/OtFr0EcfaWUfHn644wITQtiEJFOi5znyNFTmw/g3QWniR+D0abjtNhg3Dt56q9kWIN8dSWOYLgNT4kEcHR256aabeOCBBwhtosK3QJvqy87WGkY3wKDX0d/dsdNM86XklXL0Qn7ThTrNZnjzTa0H39ixHRabEMI2eviCDdHjXNwDZ9+HEb8F98ZLF1BcDNdfD3Z28M03jRaVrKqqIiYmBi8vL1bHZODQZzC3LLiKIUOGyCiUpWrKImzeDGFhDZ4S4OlMUicZmdoYr41GLghtYopv0yY4exZeeqmDohJC2JIkU6LnMBu1RefOfhD6XOPnqSrcc4/WMmbjRggKuuKUiooKDh48yN69eykpKWHwyNEkpNvzwjUhDB165fmiCb6+MGaMloA8/XSDpwR4utRWG7e1DfEZjOjnxkDvJtrCLF+ufV7XX99xgQkhbEaSKdFznHoD8mMhcg0YmqhM/q9/wVdfwSuvNFhMMiYmhh9//JGysjIGDx5MZGQkK+NKMOgTuWZMw9v7RTPmzYN//1sbEex15fcm0MuZvNIqCsurcHM0NHCBjpFRUE5MUh6/mTus8ZPOnoUNG+D//k8rSiqE6PZkzZToGUpTIPZZ6L8Q/JY0ft62bfC738HSpdrbmqeXllJVVQVoi8j9/f255557uO222+jv58+3R1KZHdwXDxd58WyVqCioqoIdOxp8uGZHn60bHtfu4mugIGutt97S+jX+6lcdFJUQwtYkmRI9w6EnQTVCxBuNLyS/cAFuugmGDdN2YikKxcXFbNmyhddee41Dhw4BMG7cOJYtW4afn1ZLKvrURbKLK1laXVtKtMLUqdq6tEZKJAR4VSdTNl43tSEug+F9XRncWM/F0lL44ANteq+RIqRCiO5HpvlE95e2CZK/htF/hl6DGj6nvFwbjaqogLVrKVRV9mzcSExMDCaTidDQUAYN0p57+cLyNYdS8HKxZ8Zwn/b+TLovR0eYMaPRZCqwExTuzCos50BSLo/PHtr4SatWQX6+9OETooeRZEp0b8YyOPgwuA6DEU81ft6jj8KBA1pT2uBgvvn4Yy5cuMDo0aOZNm0aXl4NF/bML61ka0IWt00KxKCXgd42iYqCX/8akpIgMLDeQ70c7PBysSc513aFOzcdy0BVmyjUqarawvNRo7Seg0KIHkOSKdG9JbwKxWdh1lbQOzR8znvvkfvNN/z07LPMnDOHXsD8+fNxdHTE3d29yct/fzSNSpOZpeEDrB56j1PTWmbzZrjvvise9vd0tunI1Pq4DAb7uDC0TyNTfHv2aL343nmn2ZpkQojuRf6UFt1X4WlI+CsELgPf2Q2ecvHHH/lm82aWP/oosfb2pKamAuDr69tsIgWw+lAqI/q5EdK/tzUj75lGjAA/P61EQgMCvWyXTGUXV/Dz+RwWjerXeP2w5cuhd2+tfYwQokeRkSnRPamqNr2nd4Swf17xsNls5ptVqzh25gyG4cOZFBbGlFmz6NXAtvzGnMkq4uiFfJ5ZNMKakfdciqKNTn3zDRiNWsHUOgI9nbWRQKMZe7uO/Ttw07EMzCosaGyKLz0dVq/Wpotdmqg/JYTolmRkSnRPyV9DxhYY/RI4XXoBzM3NBUBnMmHYtYvIPXt44uqribrmmhYlUgCrY1LR6xSuHStTfFYzb562gPvgwSseCvBywaxCan5Zh4e1IS6Dgd4uBPu6NnzCe+9pCeCDD3ZsYEKITkGSKdH9VBXCoSfAIwyGPgRAcnIy//3vf3njjTe4ePEi/P73XPvuu8y64w6cJ05s8S1MZpW1h1OYOdwHH9dG1mKJlps9WxuhamCqL7C6PEJSTscuQs8tqWTvuRwWhPo2PMVXVQVvvw3z58PQJnb6CSG6LZnmE91P7HNQloE6bS2JSclER0eTmJiIs7Mzs2fPxm3zZq3a9qOPao2MW+GnM9lkFlbw/NVSW8qqvLwgIkJbhP5c/ZY/tYU7O7jW1OZjGZjMauO7+Nau1ab53nuvQ+MSQnQekkyJ7iXvCJx6HYb8ijKXUax69984Ojoyb948wsPDMRw/ru0UmzYN/nnlWipLrY5JobeTgVkj+lgvdqGZNw/++ldtuq/OJgAfVwccDboOX4S+Pj6DAE9nQvq7NXzC8uUwcKA2MiWE6JFkmk90G6rZxImNz/G/i9ejjnkZZ2dnbr/9dh5//HEmTZqEobgYlizRXqC//hoMrevxVlhexaZjGVw7tj8OdnrrfhJCW4RuMmmtfepQFIWADi6PkF9ayZ4z2SwY1cgUX2ws7NoFDz2ktZARQvRIMjIlujyz2czx48eJ3votWflhePSyo9TogIsDBAQE1JykbVm/cAF27gRf31bfb31sOhVGM0vDZIqvXUyaBK6u2lTf9dfXeyjA06VDC3duScjEaFZZGNrIFN+KFVr19rvv7rCYhBCdjyRTokvLzs7myy+/JDs7Gy/7XK4bmsGom1ahu3yU4IUXYMMGePNNmDy5TfdcHZPCkD69GO0ntaXahcEAs2Zpi9BVtV4BzEAvZ3afuYiqqo3Xe7KiDfEZDHB3avh7nZcH//2vlqR7erZ7LEKIzkum+USXYzKZyMnJAaB37964urpyw5gsHgp8izGLnrsykfr+e3jxRbjrLnjggTbdOzG7hINJedwQ7tchL+Y9VlQUJCbCmTP1Dgd6OVNeZeZiUUW7h1BQVsWu0xdZ2NgU38cfa42NH3643WMRQnRukkyJLsNoNHLgwAHeeOMNPvvsM8xmMwaDgTvmDySk9E10I5+E3iPrP+n0aW3HXliYNirVxgTom0Mp6BRYMk5qS7WrefO0t5c1Pg6o3tGX1AE7+n48nkmVSW24UKfZrP1/mjIFxo1r91iEEJ2bJFOi06uqqmLfvn28/vrrrF+/HldXVxYsWKCNFpir4MBD4BIIof9X/4k1C84NBq2qtpNTm+Iwm1XWHEpl2lAf+ro5tulaohmDB8OgQVfUmwr00qqLd8Qi9PVxGfTr7chYP/crH9y8WRs1e+SRdo9DCNH5yZop0emdPn2aTZs2ERQUxJIlSwgKCro07XL8NSiIh+nfgV2dNh6qCvfcA8ePay/IgYFtjmPf+RxS88v43fzhbb6WsMC8ebByJVRWgr09AAPcndApkNzOhTuLyquIPn2RWycGoNM1MJq5fDn07QtLl7ZrHEKIrkFGpkSnU15ezs6dO9m3bx8AI0aM4J577uHOO+9k4MCBlxKpkgsQ9zwMuBr8rql/kX/+E776SqtXNGeOVeJaE5OKq4Md80JavxNQtEBUlDa6uHdv7SF7Ox39eju1+zTfthNZVBrNDRfqPHcO1q+H+++vTfKEED2bjEyJTqO0tJR9+/axf/9+KioqGDNmDKDVF/Lza6AMQczjgArhr9c/vm0b/P732qjBU09ZJbaSCiMb4tO5dmx/HA1ST6hDzJyp1W7avBmuuqr2cKBX+9eaWh+XTh9XB8IDPK588K23QKeDX/2qXWMQQnQdFiVTiqLMB/4D6IH3VVV95bLHnwTuBYzAReBuVVWTrByr6MaOHDnC+vXrqaqqYuTIkURGRuLbVC2o1B8gZS2M+Sv0Crp0PDkZbroJhg+Hjz5q84LzGhviMyitNEltqY7Uu7dWxmLzZnj55drDgV7ObD6W2W63LakwsuPkRW4e73/lFF9pKXzwgVb/aoBsQhBCaJpNphRF0QMrgLlACnBAUZR1qqom1DntMBChqmqpoigPAn8DbmqPgEX3UVhYiKIouLq64uXlRXBwMJGRkfj4+DT9RGMpHHwU3EZA8JOXjpeXa6NRFRVavzRXV6vFuiYmhSAvZ8IDGxipEO0nKkrr0ZedDd7egFa4M6ekkuIKI70crD+4vv1kFhVGc8O7+D7/XKsvJeUQhBB1WLJmagJwRlXVc6qqVgJfANfWPUFV1e2qqtaMu+8D5M930aj8/Hz+97//8frrr7Nz504A/P39uf7665tPpACO/QVKzsP4N0FfvWZFVbUXuIMH4dNPtZEpK7mQW8reczksDZPaUh0uKkr73m7dWnso0Ku6PEI7LULfEJeBdy8HxgddVohTVbWF56GhMH16u9xbCNE1WfJn3QDgQp2PU4CJTZx/D7ChLUGJ7iknJ4fdu3cTGxuLoiiMHTuWadOmtewihSfh+N8g6HboO+PS8ffegw8/hD/9Ca67zpphs/ZwKgBLwmRap8NFRICHhzbVd/PNwKVaU8k5pYT0t24V+rJKE9tOZLE0fAD6y6f49u6FI0fg7betNn0shOgerDpGrijKbUAEcFVz54qeZ8+ePcTHxzN+/HimTJmCm5tbyy6gqlpNKb0LjPv7peP79mn1fubP19rGWJGqqqw5lMKUwV74eThb9drCAnq9thuzTmuZAK/2K9y542QWZVWmhnvxLV8Obm5a+xghhKjDkmQqFfCv87Ff9bF6FEWZA/wJuEpV1fbv9SA6vYyMDKKjo5kyZQp+fn7MnDmTmTNn0qtXr9ZdMOlzyNymTe859dWOZWZq66T8/OCzz7QXXys6mJRHUk4pj80aatXrihaYNw++/hoSEiAkBDdHAx7OhnbZ0bc+PgNPF3smDLxsii8jA1avhocegtb+/xVCdFuWJFMHgKGKogxES6JuBm6pe4KiKOOAd4D5qqpmWT1K0aWkpqYSHR3NqVOncHBwIDg4GD8/v9YnUQCVBXDoSfCMgMH3a8eqquAXv9AWBO/d2y7NZtfEpOBsr2d+qNSWspm5c7W3mzZBSAgAAV4uJOdad81UeZWJbcczuWZsf+z0ly0nfe897f/bQw9Z9Z5CiO6h2WRKVVWjoiiPAJvQSiN8qKrqMUVRXgQOqqq6Dvg70Av4unqBbrKqqtc0elHRba1evZpjx47h5OTEjBkzmDhxIo6OVmi9EvsMVFyEGT+Arnr06Xe/g+horUp2dU0qa4h4aQvZxZX1joU8twnvXvYcfGau1e4jLBQQAMHB2rqpJ7Xdm4Gezhy+kGfV20SfukhJpYkFl0/xVVVp66TmzYNhw6x6TyFE92DRmilVVdcD6y879myd961TYlp0OaqqkpycTEBAQG1xzX79+hEREYGDg4N1bpIbA6ffhKEPgWe4dmzVKnjtNXjsMa2RsRVdnkg1d1x0gHnz4J13tPIXjo4EejnzQ1w6VSYzhstHkVppfVw67s4GJg/2qv/At99CWpqWUAkhRAOknYxoFVVVOX36NB9++CEff/wxJ0+eBGDSpElMnTrVeomU2QT7HwQHHxj9Z+1YbCzcey9ERsI//mGd+4jOLSpKS6R27QK0HX0ms0pqXplVLl9hNLH1eBZRI/temZytWAFBQbBwoVXuJYTofqSdjGgRVVU5efIk0dHRpKen07t3bxYuXMiQIUPa54Zn34XcAzDlM7B319ZHLVmibZf/6iswGNrnvqJzueoqrQ/e5s0wdy6BXlpT66TcUoK8XZp5cvN2n86muMJ4ZaHOuDjYuRP+9jerb24QQnQfkkyJFjGbzWzatAlFUbj66qsZM2YM+vZ6kSnLhCN/gL6zIHAZmEzatvQLF7QXuKbazbTS+rh0q19TWIGLC0ybpi1C//vfawt3JueUABYUem3G+rgMXB3tmDrYu/4DK1aAoyPcfXeb7yGE6L4kmRJNMpvNxMXFcfjwYW699VYMBgO333477u7u6HTtPEt8+CkwlULECq1I4gsvwIYNWqPZyZOtequswnKe/e4YG49lWPW6woqiouDppyE9HZ++vjjY6axSHqHSaGZLQgZzR/bF3q7O/+n8fG1zw7Jl4OXV6POFEELWTIkGmUwmDh06xPLly/n2228pLy+nsLAQAE9Pz/ZPpDJ3QOJKGPE76B0M69bBn/8Mv/wl/OpXVruNqqp8ffACc/61k20ns3h6QTDevewbPLex46KDzJunvd2yBZ1OIcDTmWQrFO786Ww2heVGFl0+xffJJ1pj40ceafM9hBDdm6Kqqq3ubbMbi6YVFxfz/vvvU1BQQP/+/Zk+fTrDhg3ruL50pkrYMBZM5bDoGJy7AOPHw9Ch2gJkJyer3CYlr5Q/ro0n+tRFxgd58OrS0QzykYKMnZbZDP36aRXRP/uMez85QEpeGRufaFufvN+tPsr6uAxi/m8ODnb6S/cKDtaaK+/ZY4XghRDdQKMvgjLNJwCorKwkLS2NoKAgXFxcGDp0KMOHD2fw4MEd39z3xL+g8Dhc9T8oN2kLzg0GWLPGKomU2ayycl8Sr248gQK8eG0It00MRHd5LzbRueh0WgHPzZvBbCbA04U9Z3NQVbXV/0erTGY2J2QyZ0SfS4kUwJYtcPo0PPeclYIXQnRnkkz1cBUVFRw4cIC9e/dSWVnJk08+iZOTE4sWLbJNQMWJEP8i+C2B/gvhppvgxAlt4XFgYJsvf/ZiMU+vieVAYh7Th/nwlyWh0nOvK5k3T2sbdPQogV4elFaauFhcQR/X1hWG3Xcuh/zSqit38a1YAX36wA03WCFoIUR3J8lUD1VeXs6+ffv4+eefKS8vZ8iQIURGRuJkpSm0Vot5HFAg/DWthtTXX2vb0ue0rS6s0WTmvV3n+ffWUzgZ9PzjxjEsDRvQ8aNuom3qtJYJWKLtsEvOKW11MrU+LgMXez1XDauzI/D8efjf/+BPfwJr1UsTQnRrkkz1UEVFRURHRzNs2DAiIyMZMGCArUOClHWQug7G/g32ndZ2bt1wA/z2t226bEJaIb9fE0tcagHzQ3x58bqQVr/4Chvz9dVaB23eTOA92sLwpJxSIoJa3pfRaDKz6VgGs0b0xdFQZ4rvrbe0KUUrbnQQQnRvkkz1EMXFxezZs4eysjKuvfZafHx8eOyxx3B3d7d1aBpjCRx8FHqHgNMSuGmStgD4ww+1sgitUGE0sXzbGd7acRZ3Z3veujXsyukc0fVERcFrrzHAYEJRtMKdrbH/fC65JZUsrNvEuqwMPvgArrsO/PysE68QotuTZKqbKygoYM+ePRw6dAiTycTo0aMxm83odLrOk0gBxL8EpckwfQtcd7PWXPabb8DVtVWXO5Scx+9Xx3I6q5jrwwbw7OKRuDtLaYNuISoK/v53HH7aTf/eTlxoZTK1Pj4dJ4OeGcP7XDr4xReQmyvlEIQQLSLJVDd2/PhxVq9eDcDo0aOJjIzE07Pl0yHtriABjv8DBt4Fz66CmBituezw4S2+VGmlkX9uPsWHP52nn5sjH/1yPDPrvliKrm/aNG1X5+bNBIQsIymnpMWXMJlVNsZnMjPYByf76ik+VYU33oCQEK19jRBCWEiSqW4mJyeHqqoqfH198ff3JywsjKlTp3auUai6VBUOPAQGV4gLgY+egmeegWuvbfGl9pzJ5ulv4kjOLeX2SYH8fkEwvRzkv3i34+ioJTubNhE4/R62Hs9s8SUOJOaSXVzBwrrTvvv2weHD8OabrZ5aFkL0TPJK001kZWWxa9cujh07RlBQEHfccQe9evWyXYkDSyX+F7J2gtfv4JY/wvz58PzzLbpEYXkVf11/nM/3XyDIy5kv75/ExEHS/qNbmzcPfv1rRlbl8UVxJcUVxhYlzhvi0nGw09UftVyxAtzc4Pbb2yFgIUR3JslUF5eZmcnOnTs5fvw4BoOByZMnM9nKfevaTWUeHPoNuIXDPSvB31+rIdSCxsk/Hs/kT2vjySoq51fTB/HrucPq78wS3VNUFABjjx8ARpCcU8rI/m4WPdVsVtkQn8GM4T641CRgmZnw1Vfw4IPQS6rgCyFaRpKpLqqm6vP58+c5d+4ckZGRTJo0CWfnLlSA8uifoCIH/usHufmwdwNYuKYrp7iCF75PYN3RNIJ9XXnn9nDG+Lu3a7iiExkxAgYMIDBmN4SOIDm3xOJk6lByHllFl03xvfeetunhoYfaKWAhRHcmyVQXk5SURHR0NCEhIYSFhREeHs7YsWNxdOxidZOy98PptyFjDHx7GP77X61+UDNUVeX72HSeX3eMovIqfj1nGA/OGIy9nfTs7lEUBebNw+2bb9CNvJukHMt39K2Py8DeTses4OopPqMR3n5bKwjaik0PQgghyVQXoKoq58+fJzo6mqSkJFxcXAgNDQXAYDBgMBhsHGELmU1w4EFQ3eGZI/D443Drrc0+LaOgnGe+jWfr8UzG+Lvzt6WjGe7butIJohuIikL58EOm5p0nOXegRU/RpvjSmT7UB1fH6p+b776D1FRt4bkQQrSCJFNdwPfff8/hw4dxdXVl3rx5hIeHd70Eqq7Tb0HeIXjHABGR8Pe/N3m6qqp8eeACL68/TpXJzJ8WjuDuaQPRS2Pinm3OHFAUFqTFsiF3kkVPOZKST3pBOU/NqzMCtXy51vexs2/WEEJ0WpJMdUKqqnLy5EkCAwNxcnJi5MiR9O/fn7Fjx2Jn18W/ZWXpcPSPcMYRznpAzFfQRGJ4IbeUp7+J5aczOUwa5Mkr148myNulAwMWnZaXF0REMPHMId62cJpvQ1w6Br3C7BF9tQPx8bBjB7zySos2PgghRF1d/JW5ezGbzSQkJLBr1y6ysrKIiopi8uTJDBkyxNahWU/Mk1BRDO/pYfUarddaA0xmlU/2JPL3TSfR6xReXhLKsvEB6GQ0StQVFcXAv75CQUY2VSYzBn3ja+dUVWV9XAbThnjT26k6gV+xQmtmfM89HRSwEKI7kmSqE1BVldjYWHbt2kVOTg7e3t4sWbKkdl1Ut5HxIyR/Ad8BzyyHRko4nMkq4nerYzmUnM/M4T68vGQU/d2dOjZW0TXMm4fu5ZeZlHiEtPyFBHo1PmoZl1pAan4ZT8wZqh0oKICVK2HZMvD27qCAhRDdkSRTNlRT3kBRFOLj47Gzs+PGG29kxIgRKN2tArOpArbfARmAxx1w//1XnFJlMvPOzrO8/uMZXBz0vHbTWK4d27/7fS2E9UyahMnFhennD5GUU9pkMvVDXDp2OoW5I6un+D75BEpKpA+fEKLNJJmyAaPRyKFDh9i3bx+33347Hh4eXH/99Tg6OnavxMHXVyuGCHAt8AvgEyBz0xXtOuJTC/jd6lgS0gtZNLofL1wTgncvh46OWHQ1BgNV02cwfc9BduSUAD4NnqaqKhviMpgyxFtreG02a1N8EydCeHjHxiyE6HYkmepAlZWVxMTEsGfPHoqLi/H396eyshIAJ6duOI1Vk0j5ANcBPwOxAJd6qZVXmXj9x9O8E30OLxd73rk9nHkhDa+jEqIh9osW4L/hB4qPnYDJQQ2ecyytkOTcUh6aMVg78OOPcOqUNs0nhBBtJMlUBzEajSxfvpyioiIGDhzI0qVLCQwM7F4jUY25EzABl71uHUzM5XdrYjl3sYRfRPjxp4Uj6e3chUs+CJvQzZsHgPuubXDv/AbP2RCfjl6nEFWTqC9fDj4+cOONHRWmEKIbk2SqHZWVlXHy5MnakgbTpk2jX79++Pv72zo06ysqgthYOHJE+3f0qHY8AhiHlkjlXTr9+XXH+GRvIv17O7HynglEDm14ekaIZg0ZQpbPAIIO7Wnw4ZpdfJMHeeHpYg+JifD99/CHP2g7+YQQoo0kmWoHJSUl7Nu3j/3791NZWYmfnx/e3t5MmDDB1qG1narChQtaslSTNB05AmfPXjrHywvGjgUH4A4gCdhc/zKf7E3kzslBPDVv+KVms0K0UnL4VEZv+x61ogLlsgTpREYR57NLuDeyukr6W29pa/YeeMAGkQohuiN5FbOisrIyoqOjiYmJoaqqipEjRzJ9+nS8u+q268pKSEi4lDDVJE951UNMigJDhkBYGPzyl1oCNWYMDBigPXaLAl7AG4C5/qW//tVkIoIsa2osRHMKI2fisvEr8rfvwn3+nHqPbYhLR6dA1EhfKCuDDz6A666D7jhCLISwCUmmrMBsNqPT6dDpdMTGxjJixAimTZuGj08XmrrKydESpbqJ0/HjUFWlPe7kBKNHa2tMxo7V/o0aBb16XbrGN74QfWlxOYur3z4BPHzpsNq3ryRSwqr0c2djfEZH2Q8brkim1sdnMGGgJz6uDvDxx9r/dSmHIISwIkVVVVvd22Y3tpa8vDx2795NWloa999/P4qiUFFRgUNnXodhNsO5c/Wn6I4cgZSUS+f063cpYRozRns7ZEjz7TZWNb6YfuHFaP52w2hCB/Ru86cgxOXOZBWTFzaRIb3t8Dh2tPb46cwi5v47mhevDeGOSYEwfrw2OhUff0V5DiGEaEajvzRkZKoVcnJy2LVrF7Gxseh0OsaNG0dVVRX29vadK5EqLdVeNOomTrGxUFysPa7XQ3AwXHXVpcRpzBjo06f5a6tmKE2F4jNQVP2vCd89MrXJVh9CtIW/pxPrBo4j4qdVkJ1dW9H8h7h0FAXmh/jC/v0QE6PVl5JESghhRZJMtVBiYiKffvoper2eCRMmMHXqVFxdXW0dFmRk1F/XdOSIVkfHXL1Yyc1NS5Tqrm0KCQFHx8avaTZCaTIUna2fNBWf0Y6ZK2pPrTDb4dBEriSJlGhPDnZ6EkZNQtn9mVZD6qabANgQl8H4QE/6uDlq5RBcXeH2220crRCiu5FkygLp6ekUFhYyfPhw/P39mTFjBuHh4bi4NN66ot0YjXDy5JW76bKyLp0TFKQlSzfddClxCgpq+K9xUyWUJNZJkuokTMXnQTVeOlfvBL0Gg+sw6L+QP24pI7GyH0kV/Umv8uLc6Gvb8zMXokklo8dS5OyK66ZNcNNNnMkq5mRmEc9dPVIrIPvVV/CrX2kJlRBCWJEkU01ISUkhOjqa06dP4+XlxbBhw9Dr9UyfPr1jAigsvLJ2U3w8lJdrj9vbQ2goLFpUf5rO3b3+dYxlUHj8spGl6n+lSdqUXQ27XuA6FDzGgv8N4DqkOoEaAk79QLk0wrTqyx/a9/MXogUCvN3YN3AcczdvBlVlY3w6APNDfWH5v7TdqQ89ZOMohRDdkSRTDUhPT2fr1q2cO3cOJycnZs6cyYQJE9qvWrmqQnLylSUIzp27dI63t5YwPfzwpcQpOBgM1RXDq4qh+CwU/ghpl40wlabUv5/BXUuYvCeB623Qa4iWLLkOAQefeiNYqqqSXVxJUlYJiTlpJOWUkJhTSlJOyRWfxsUqd3wM+Vd+fo592/b1EcICAV7ObPUfw9xj0ZCQwPq4XMIDPejnYoC334Y5c7SfGSGEsDJJpqqpqorJZMLOzo6ysjIyMzOZO3cuERER2NvbW+9GFRUN127Kz9ceVxQYOhQiIuCeey7tquvXD6oKtISp6AwUrYOYf11Kmsoz6t/HsY+WJPWdVZ0sDb6UNDnUL0ugqipZRRUkppWQlJNCYk4JSTmlnM8uISmnhJJKU+25OgX8PJwJ9HK+4lMbf/y/te8nvrLIKl8uISwV4OnMZwPHAZC79n8kFIfyzKIRsG6dtlt1+XIbRyiE6K56fGkEVVU5ffo00dHR+Pn5MX/+/HqJVZtkZzdcu8lYvQ7J2Vmr3VR3im54fzCnXzklV3wWKrLrX9+pf/U0XJ2RpZrEyeBW71SzWSW9sJyk7EsjSzVJU1JOKWVVlxImO52Cv6eWMAV5udR76+fhjL2dNtUX9HTj03ySTImOFpuSzzXLfyL+qyfI8enPVTN/x09Pz2LA9Yu0Cv3nzjVf3kMIIRonpREup6oqJ06cIDo6moyMDNzd3fH11ZqgKopyKZHy9dUWr16ub194Eyhv4LEye3jOB1JTLx3r319Lmq5eDKODYEgv6F0OJeeqk6X3IeUMnC+ocyEFXAK0BMl/af31S70GgV39BfBGk5n0gnISE0tIzEmqTZwSc0pIzi2l0nhpbZS9XkeAlzNBXs5MHeJNkJczgV4uBHm50N/dETsLdt9597Inu7iyweNCdLRAT+3nITFsKkPXfUHEMicGpJ6D7dvhr3+VREoI0W56bDK1detW9uzZg6enJ9deey2jRo1C39Av24YSqZrj5Y1c3KkSFkyA0X0hyAm8qsBUPdpUvBOMJXCy+lxFDy5BWoLkNan+CFOvgaCvX7eqymQmJa+MxLMlJGVn1Y4yJeWUciGvlCrTpQE/R4OOQE8XBnm7MCu4D0FeLlrS5O2Cr5sjel3b1oAdfGZum54vhDX1djbQ28nAJr8xhFZ9wl2mFHhzjdbM+J57bB2eEKIbs2iaT1GU+cB/AD3wvqqqr1z2uAPwKRAO5AA3qaqa2MxlO3Saz2QyERcXR//+/enTpw/Z2dmkp6cTEhKCTtfEKExTi84/s+zeFWY7LlT6kljZjyzVj1tmz740wuQSCDpD/fONJi7kltVb7K2tXyolNb8Mk/nSl87FXq+NKHnXjCxdGmHq4+qAro0JkxBdQcRLW2pHSQ+8cSs+pQVXntS3r1aPTQghWqf103yKouiBFcBcIAU4oCjKOlVVE+qcdg+Qp6rqEEVRbgZeBW5qW8zWYTQaOXr0KLt37yY/P5/JkycTFRWFt7d3uzYg/mPKw/VqMJm5NOp1y7BFlFWaSM4tJTE5p17SlJhdSlpBGXVzXFdHOwZ6uzDG351rx/avlzR597Jvv12GQnQRdaebG0ykoPFRZiGEaCNLpvkmAGdUVT0HoCjKF8C1QN1k6lrg+er3VwPLFUVRVBuubgc4dOgQO3fupLCwkAEDBrBgwQKGDh3aIfdelbug0ccm/eVHMgrrzxF6OBsI9HJhfJAHgV5+dUaaXPBwNkjCJIQQQnRSliRTA4ALdT5OASY2do6qqkZFUQoAL+Cy7WcdKycnB3d3d6655hoGDRrUaRKS2gXf3tUjTJ4u9HY2NP9EIYQQQnQ63XoB+qxZs9DpdG1Lovr2bXw3nyMN7+ZrpkjlP38xpvXxCCGEEKJTsSSZSgX863zsV32soXNSFEWxA3qjLUS3qQZ357VUaxes7pdWK0IIIURP0HwxITgADFUUZaCiKPbAzcC6y85ZB9xZ/f4NwDZbr5eytcZqLUkNJiGsr+7P1UUX94ZP6ittjYQQ7cPS0ggLgdfQSiN8qKrqy4qivAgcVFV1naIojsBKYByQC9xcs2C9CT062RJCCCFEl9LomqEe305GCCGEEMICjSZTlkzzCSGEEEKIRkgyJYQQQgjRBpJMCSGEEEK0gSRTQgghhBBtIMmUEEIIIUQbSDIlhBBCCNEGkkwJIYQQQrSBLXvzdY6uw0IIIYQQbSAjU0IIIYQQbSDJlBBCCCFEG0gyJYQQQgjRBpJMCSGEEEK0gSRTQgghhBBtIMmUEEIIIUQbdItkSlGU+YqinFQU5YyiKE838LiDoihfVj/+s6IoQTYIs8ex4PvypKIoCYqixCqK8qOiKIG2iLMnae57Uue8pYqiqIqiRHRkfD2RJd8TRVF+Uf2zckxRlFUdHWNPZMHvrwBFUbYrinK4+nfYQlvE2ZMoivKhoihZiqLEN/K4oijK69Xfs1hFUcI6KrYun0wpiqIHVgALgJHAMkVRRl522j1AnqqqQ4B/A692bJQ9j4Xfl8NAhKqqo4HVwN86NsqexcLvCYqiuAKPAz93bIQ9jyXfE0VRhgJ/AKaqqhoCPNHRcfY0Fv6sPAN8parqOOBm4M2OjbJH+hiY38TjC4Ch1f/uB97qgJiAbpBMAROAM6qqnlNVtRL4Arj2snOuBT6pfn81MFtRFCka2r6a/b6oqrpdVdXS6g/3AX4dHGNPY8nPCsCf0f7gKO/I4HooS74n9wErVFXNA1BVNauDY+yJLPm+qIBb9fu9gbQOjK9HUlU1Gsht4pRrgU9VzT7AXVGUfh0RW3dIpgYAF+p8nFJ9rMFzVFU1AgWAV4dE13NZ8n2p6x5gQ7tGJJr9nlQPi/urqvpDRwbWg1nyczIMGKYoyk+KouxTFKWpv8yFdVjyfXkeuE1RlBRgPfBox4QmmtDS1x2rsWU7GSEAUBTlNiACuMrWsfRkiqLogH8Bd9k4FFGfHdq0xQy00dtoRVFGqaqab8ugBMuAj1VV/aeiKJOBlYqihKqqarZ1YKLjdYeRqVTAv87HftXHGjxHURQ7tCHZnA6Jruey5PuCoihzgD8B16iqWtFBsfVUzX1PXIFQYIeiKInAJGCdLEJvV5b8nKQA61RVrVJV9TxwCi25Eu3Hku/LPcBXAKqq7gUcAe8OiU40xqLXnfbQHZKpA8BQRVEGKopij7YQcN1l56wD7qx+/wZgm6qqagfG2BM1+31RFGUc8A5aIiXrQNpfk98TVVULVFX1VlU1SFXVILR1bNeoqnrQNuH2CJb8/voWbVQKRVG80ab9znVgjD2RJd+XZGA2gKIoI9CSqYsdGqW43DrgjupdfZOAAlVV0zvixl1+mk9VVaOiKI8AmwA98KGqqscURXkROKiq6jrgA7Qh2DNoi9dutl3EPYOF35e/A72Ar6v3AySrqnqNzYLu5iz8nogOZOH3ZBMQpShKAmACnlJVVUbW25GF35ffAO8pivJrtMXod8kf6e1LUZTP0f6w8K5eq/YcYABQVfVttLVrC4EzQCnwyw6LTb73QgghhBCt1x2m+YQQQgghbEaSKSGEEEKINpBkSgghhBCiDSSZEkIIIYRoA0mmhBBCCCHaQJIpIYQQQog2kGRKCCGEEKINJJkSQvz/RsEoGAWjYBRQAADjip4AqHNn0QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Classification probability calibration for random forest classifier\n",
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "\n",
    "# Uncalibrated\n",
    "clf = RandomForestClassifier(n_estimators=600, max_features='auto', max_depth=20, min_samples_split=2, min_samples_leaf=4, random_state=42)\n",
    "clf.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', label='Uncalibrated')\n",
    "\n",
    "# Calibrated by isotonic method\n",
    "clf_isotonic = CalibratedClassifierCV(clf, cv=3, method='isotonic')\n",
    "clf_isotonic.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf_isotonic.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', color='red', label='Calibrated (Isotonic)')\n",
    "\n",
    "# Calibrated by sigmoid method\n",
    "clf_sigmoid = CalibratedClassifierCV(clf, cv=3, method='sigmoid')\n",
    "clf_sigmoid.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf_sigmoid.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', color='orange', label='Calibrated (Sigmoid)')\n",
    "\n",
    "plt.plot([0, 1], [0, 1], '--', color='gray')\n",
    "sns.despine(left=True, bottom=True)\n",
    "plt.gca().xaxis.set_ticks_position('none')\n",
    "plt.gca().yaxis.set_ticks_position('none')\n",
    "plt.gca().legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.2 XGBoost Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Calculate feature importance for XGBoost classifier\n",
    "xgb_model = XGBClassifier(max_depth=6, min_child_weight=9, gamma=0.1, colsample_bytree=0.6, random_state=42)\n",
    "xgb_model.fit(X_train, y_train.ravel())\n",
    "importances = xgb_model.feature_importances_\n",
    "indices = np.argsort(importances)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "plt.title('Feature Importance')\n",
    "plt.barh(range(len(indices)), importances[indices], color='b', align='center')\n",
    "plt.yticks(range(len(indices)), [features[i] for i in indices])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Classification probability calibration for XGBoost classifier\n",
    "fig, ax = plt.subplots(figsize=(10, 5))\n",
    "\n",
    "# Uncalibrated\n",
    "clf = XGBClassifier(max_depth=6, min_child_weight=9, gamma=0.1, colsample_bytree=0.6, random_state=42)\n",
    "clf.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', label='Uncalibrated')\n",
    "\n",
    "# Calibrated by isotonic method\n",
    "clf_isotonic = CalibratedClassifierCV(clf, cv=5, method='isotonic')\n",
    "clf_isotonic.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf_isotonic.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', color='red', label='Calibrated (Isotonic)')\n",
    "\n",
    "# Calibrated by sigmoid method\n",
    "clf_sigmoid = CalibratedClassifierCV(clf, cv=5, method='sigmoid')\n",
    "clf_sigmoid.fit(X_train, y_train.ravel())\n",
    "y_test_predict_proba = clf_sigmoid.predict_proba(X_test)[:, 1]\n",
    "fraction_of_positives, mean_predicted_value = calibration_curve(y_test, y_test_predict_proba, n_bins=8)\n",
    "plt.plot(mean_predicted_value, fraction_of_positives, 's-', color='orange', label='Calibrated (Sigmoid)')\n",
    "\n",
    "plt.plot([0, 1], [0, 1], '--', color='gray')\n",
    "sns.despine(left=True, bottom=True)\n",
    "plt.gca().xaxis.set_ticks_position('none')\n",
    "plt.gca().yaxis.set_ticks_position('none')\n",
    "plt.gca().legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chpater 5 Outs above average Leaderboard\n",
    "Now, we decide to use **XGBoost Classifier** calibrated by sigmoid method to predict **Out Probability** and further calculate **Outs Above Average**. Let's first recall what is **Outs Above Average**. According to Statcast, **Outs Above Average** is a range-based metric of skill that shows how many outs a player has saved. For example, a fielder who catches a 25% **Out Probability** play gets +0.75; one who fails to make the play gets -0.25. \n",
    "\n",
    "Therefore, on the basis of the definition, we first use **XGBoost Classifier** calibrated by sigmoid method to predict **Out Probability** for each in-play ground ball first fielded by shortstop and further calculating **Outs Above Average** for each shortstop. Finally, we have a leaderboard of **Outs Above Average** for shortstop including the playerid, number of opportunities, and the outs above average. From the leaderboard, we could recognize that the shortstop whose playerid equals to 162393 has the highest **Outs Above Average**, which equals to 3.67. Please refer to Appendix for the full leaderboard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a function to calculate out probability and oaa\n",
    "def OAA_function(data, model):\n",
    "    \n",
    "    X = data[features].to_numpy()\n",
    "    y = data[['player_out_credit']].to_numpy()\n",
    "    y_predict_prob = model.predict_proba(X)[:,1]\n",
    "    out_prob = y_predict_prob\n",
    "    \n",
    "    OAA_score = []\n",
    "    n = len(y_predict_prob)\n",
    "    for i in range(n):\n",
    "        if y.ravel()[i] == 1:\n",
    "            score = round(1-y_predict_prob[i], 2)\n",
    "            OAA_score.append(score)\n",
    "        elif y.ravel()[i] == 0:\n",
    "            score = round(-y_predict_prob[i], 2)\n",
    "            OAA_score.append(score)\n",
    "    \n",
    "    return out_prob, OAA_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate out probability and oaa for each play\n",
    "xgb_model = XGBClassifier(max_depth=6, min_child_weight=9, gamma=0.1, colsample_bytree=0.6, random_state=42)\n",
    "xgb_sigmoid = CalibratedClassifierCV(xgb_model, cv=5, method='sigmoid')\n",
    "xgb_sigmoid.fit(X_train, y_train.ravel())\n",
    "OAA = OAA_function(shortstop_data, xgb_sigmoid)\n",
    "shortstop_data = shortstop_data.assign(outs_above_average=OAA[1])\n",
    "shortstop_data = shortstop_data.assign(out_probability=OAA[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>playerid</th>\n",
       "      <th>number of opportunities</th>\n",
       "      <th>outs above average</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>162393</td>\n",
       "      <td>148</td>\n",
       "      <td>3.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>167960</td>\n",
       "      <td>196</td>\n",
       "      <td>2.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>206311</td>\n",
       "      <td>133</td>\n",
       "      <td>1.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>154448</td>\n",
       "      <td>166</td>\n",
       "      <td>1.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2087</td>\n",
       "      <td>62</td>\n",
       "      <td>1.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9424</td>\n",
       "      <td>46</td>\n",
       "      <td>1.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>5393</td>\n",
       "      <td>129</td>\n",
       "      <td>0.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>9210</td>\n",
       "      <td>48</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5419</td>\n",
       "      <td>129</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>169390</td>\n",
       "      <td>13</td>\n",
       "      <td>0.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>9074</td>\n",
       "      <td>126</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>6619</td>\n",
       "      <td>38</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>158358</td>\n",
       "      <td>23</td>\n",
       "      <td>0.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>171885</td>\n",
       "      <td>68</td>\n",
       "      <td>0.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>161551</td>\n",
       "      <td>157</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>9148</td>\n",
       "      <td>111</td>\n",
       "      <td>0.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>207428</td>\n",
       "      <td>46</td>\n",
       "      <td>0.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>7580</td>\n",
       "      <td>82</td>\n",
       "      <td>0.38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>132551</td>\n",
       "      <td>137</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>162294</td>\n",
       "      <td>42</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>4379</td>\n",
       "      <td>12</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>113301</td>\n",
       "      <td>76</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>159128</td>\n",
       "      <td>34</td>\n",
       "      <td>0.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>3295</td>\n",
       "      <td>16</td>\n",
       "      <td>0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>11869</td>\n",
       "      <td>135</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>162648</td>\n",
       "      <td>145</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>172568</td>\n",
       "      <td>25</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>167746</td>\n",
       "      <td>20</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>168425</td>\n",
       "      <td>17</td>\n",
       "      <td>0.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>3843</td>\n",
       "      <td>33</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>184486</td>\n",
       "      <td>7</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>159560</td>\n",
       "      <td>7</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>6731</td>\n",
       "      <td>7</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>159953</td>\n",
       "      <td>6</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>163723</td>\n",
       "      <td>6</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>9425</td>\n",
       "      <td>122</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>4955</td>\n",
       "      <td>5</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>203336</td>\n",
       "      <td>5</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>206262</td>\n",
       "      <td>18</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>4331</td>\n",
       "      <td>8</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    playerid  number of opportunities  outs above average\n",
       "0     162393                      148                3.67\n",
       "1     167960                      196                2.87\n",
       "2     206311                      133                1.78\n",
       "3     154448                      166                1.60\n",
       "4       2087                       62                1.06\n",
       "5       9424                       46                1.04\n",
       "6       5393                      129                0.86\n",
       "7       9210                       48                0.77\n",
       "8       5419                      129                0.76\n",
       "9     169390                       13                0.73\n",
       "10      9074                      126                0.72\n",
       "11      6619                       38                0.65\n",
       "12    158358                       23                0.62\n",
       "13    171885                       68                0.51\n",
       "14    161551                      157                0.48\n",
       "15      9148                      111                0.47\n",
       "16    207428                       46                0.39\n",
       "17      7580                       82                0.38\n",
       "18    132551                      137                0.32\n",
       "19    162294                       42                0.24\n",
       "20      4379                       12                0.24\n",
       "21    113301                       76                0.22\n",
       "22    159128                       34                0.18\n",
       "23      3295                       16                0.17\n",
       "24     11869                      135                0.16\n",
       "25    162648                      145                0.14\n",
       "26    172568                       25                0.10\n",
       "27    167746                       20                0.10\n",
       "28    168425                       17                0.09\n",
       "29      3843                       33                0.08\n",
       "30    184486                        7                0.07\n",
       "31    159560                        7                0.07\n",
       "32      6731                        7                0.07\n",
       "33    159953                        6                0.06\n",
       "34    163723                        6                0.06\n",
       "35      9425                      122                0.05\n",
       "36      4955                        5                0.05\n",
       "37    203336                        5                0.05\n",
       "38    206262                       18                0.04\n",
       "39      4331                        8                0.04"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build a leaderboard (the first one)\n",
    "leaderboard = shortstop_data.groupby('playerid').agg({'outs_above_average': ['count', 'sum']}).reset_index()\n",
    "leaderboard.columns = ['playerid', 'number of opportunities', 'outs above average']\n",
    "leaderboard.sort_values(by='outs above average', ascending=False).reset_index(drop=True)[0:40]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 6 Shifting\n",
    "Since now we understand how to calculate **Outs Above Average** and know exactly where each shortstop stands, which is helpful in a baseball world where shifting and out-of-position defenders are commonplace. Therefore, after visualizing each shortstop's location, we could easily recognize whether the shorstop is standing in regular spot or between the first base and the second base. Hence, It also allows for an further **Outs Above Average** breakdown.\n",
    "\n",
    "In our case, when the shorstop is standing between the second base and the third base, we define this situation as **no shifting**. On the other hand, when the shorstop is standing between the first base and the second base, we define this situation as **shifting**. For example, from the previous chapter, the shortstop whose playerid equals to 162393 has **+3.67 OAA**. Due to shifting, this shortstop's **+3.67 OAA** actually breaks down into **+3.33 OAA** in regular spot and **+0.34 OAA** when shifting. Please also refer to Appendix for the full leaderboard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize each shortstop's location\n",
    "x1 = shortstop_data['player_x']\n",
    "y1 = shortstop_data['player_y']\n",
    "\n",
    "lines = [[(location_HB['x'], location_HB['y']), (160*(2**0.5), 160*(2**0.5))],\n",
    "         [(location_HB['x'], location_HB['y']), (-160*(2**0.5), 160*(2**0.5))],\n",
    "         [(location_1B['x'], location_1B['y']), (location_2B['x'], location_2B['y'])],\n",
    "         [(location_2B['x'], location_2B['y']), (location_3B['x'], location_3B['y'])]]\n",
    "lc = mc.LineCollection(lines, colors='black', linewidths=1)\n",
    "arc_element1 = Arc((0, 0), 548, 800, theta1=45, theta2=135, linewidth=1, color='black')\n",
    "arc_element2 = Arc((0, 60.5), 180, 180, theta1=17, theta2=163, linewidth=1, color='black')\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(8,8))\n",
    "ax.axis('scaled')\n",
    "ax.set_xlim([-250, 250])\n",
    "ax.set_ylim([-10, 430])\n",
    "ax.add_collection(lc)\n",
    "ax.add_patch(arc_element1)\n",
    "ax.add_patch(arc_element2)\n",
    "ax.scatter(x1, y1, color='red', alpha=1, s=6, label='shortstop location')\n",
    "ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a function to check whether the shortstop shifts\n",
    "def shift_function(data):\n",
    "    \n",
    "    shift = []\n",
    "    n = data.shape[0]\n",
    "    for i in range(n):\n",
    "        if data['player_x'][i] >= 0:\n",
    "            shift.append('shifting')\n",
    "        elif data['player_x'][i] < 0:\n",
    "            shift.append('no shifting')\n",
    "            \n",
    "    return shift"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Check whether the shortstop shifts\n",
    "shift_or_not = shift_function(shortstop_data)\n",
    "shortstop_data = shortstop_data.assign(shift_or_not=shift_or_not)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>playerid</th>\n",
       "      <th>number of opportunities | no shifting</th>\n",
       "      <th>number of opportunities | shifting</th>\n",
       "      <th>number of opportunities</th>\n",
       "      <th>outs above average | no shifting</th>\n",
       "      <th>outs above average | shifting</th>\n",
       "      <th>outs above average</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>162393</td>\n",
       "      <td>121</td>\n",
       "      <td>27</td>\n",
       "      <td>148</td>\n",
       "      <td>3.33</td>\n",
       "      <td>0.34</td>\n",
       "      <td>3.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>167960</td>\n",
       "      <td>176</td>\n",
       "      <td>20</td>\n",
       "      <td>196</td>\n",
       "      <td>1.90</td>\n",
       "      <td>0.97</td>\n",
       "      <td>2.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>206311</td>\n",
       "      <td>127</td>\n",
       "      <td>6</td>\n",
       "      <td>133</td>\n",
       "      <td>1.72</td>\n",
       "      <td>0.06</td>\n",
       "      <td>1.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>154448</td>\n",
       "      <td>138</td>\n",
       "      <td>28</td>\n",
       "      <td>166</td>\n",
       "      <td>1.21</td>\n",
       "      <td>0.39</td>\n",
       "      <td>1.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2087</td>\n",
       "      <td>48</td>\n",
       "      <td>14</td>\n",
       "      <td>62</td>\n",
       "      <td>1.13</td>\n",
       "      <td>-0.07</td>\n",
       "      <td>1.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9424</td>\n",
       "      <td>42</td>\n",
       "      <td>4</td>\n",
       "      <td>46</td>\n",
       "      <td>1.05</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>5393</td>\n",
       "      <td>109</td>\n",
       "      <td>20</td>\n",
       "      <td>129</td>\n",
       "      <td>1.04</td>\n",
       "      <td>-0.18</td>\n",
       "      <td>0.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>9210</td>\n",
       "      <td>44</td>\n",
       "      <td>4</td>\n",
       "      <td>48</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5419</td>\n",
       "      <td>112</td>\n",
       "      <td>17</td>\n",
       "      <td>129</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>169390</td>\n",
       "      <td>10</td>\n",
       "      <td>3</td>\n",
       "      <td>13</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>9074</td>\n",
       "      <td>107</td>\n",
       "      <td>19</td>\n",
       "      <td>126</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>6619</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>38</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>158358</td>\n",
       "      <td>18</td>\n",
       "      <td>5</td>\n",
       "      <td>23</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.57</td>\n",
       "      <td>0.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>171885</td>\n",
       "      <td>55</td>\n",
       "      <td>13</td>\n",
       "      <td>68</td>\n",
       "      <td>-0.92</td>\n",
       "      <td>1.43</td>\n",
       "      <td>0.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>161551</td>\n",
       "      <td>120</td>\n",
       "      <td>37</td>\n",
       "      <td>157</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>9148</td>\n",
       "      <td>95</td>\n",
       "      <td>16</td>\n",
       "      <td>111</td>\n",
       "      <td>0.38</td>\n",
       "      <td>0.09</td>\n",
       "      <td>0.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>207428</td>\n",
       "      <td>31</td>\n",
       "      <td>15</td>\n",
       "      <td>46</td>\n",
       "      <td>-0.49</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>7580</td>\n",
       "      <td>63</td>\n",
       "      <td>19</td>\n",
       "      <td>82</td>\n",
       "      <td>0.51</td>\n",
       "      <td>-0.13</td>\n",
       "      <td>0.38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>132551</td>\n",
       "      <td>113</td>\n",
       "      <td>24</td>\n",
       "      <td>137</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>162294</td>\n",
       "      <td>40</td>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>4379</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>113301</td>\n",
       "      <td>61</td>\n",
       "      <td>15</td>\n",
       "      <td>76</td>\n",
       "      <td>0.32</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>159128</td>\n",
       "      <td>34</td>\n",
       "      <td>0</td>\n",
       "      <td>34</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>3295</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>11869</td>\n",
       "      <td>109</td>\n",
       "      <td>26</td>\n",
       "      <td>135</td>\n",
       "      <td>0.35</td>\n",
       "      <td>-0.19</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>162648</td>\n",
       "      <td>111</td>\n",
       "      <td>34</td>\n",
       "      <td>145</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>167746</td>\n",
       "      <td>17</td>\n",
       "      <td>3</td>\n",
       "      <td>20</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>172568</td>\n",
       "      <td>21</td>\n",
       "      <td>4</td>\n",
       "      <td>25</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>168425</td>\n",
       "      <td>15</td>\n",
       "      <td>2</td>\n",
       "      <td>17</td>\n",
       "      <td>0.15</td>\n",
       "      <td>-0.06</td>\n",
       "      <td>0.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>3843</td>\n",
       "      <td>31</td>\n",
       "      <td>2</td>\n",
       "      <td>33</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>159560</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>184486</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>6731</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>159953</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>163723</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>9425</td>\n",
       "      <td>105</td>\n",
       "      <td>17</td>\n",
       "      <td>122</td>\n",
       "      <td>0.95</td>\n",
       "      <td>-0.90</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>4955</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>203336</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>206262</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>4331</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    playerid  number of opportunities | no shifting  \\\n",
       "0     162393                                    121   \n",
       "1     167960                                    176   \n",
       "2     206311                                    127   \n",
       "3     154448                                    138   \n",
       "4       2087                                     48   \n",
       "5       9424                                     42   \n",
       "6       5393                                    109   \n",
       "7       9210                                     44   \n",
       "8       5419                                    112   \n",
       "9     169390                                     10   \n",
       "10      9074                                    107   \n",
       "11      6619                                     36   \n",
       "12    158358                                     18   \n",
       "13    171885                                     55   \n",
       "14    161551                                    120   \n",
       "15      9148                                     95   \n",
       "16    207428                                     31   \n",
       "17      7580                                     63   \n",
       "18    132551                                    113   \n",
       "19    162294                                     40   \n",
       "20      4379                                     11   \n",
       "21    113301                                     61   \n",
       "22    159128                                     34   \n",
       "23      3295                                     16   \n",
       "24     11869                                    109   \n",
       "25    162648                                    111   \n",
       "26    167746                                     17   \n",
       "27    172568                                     21   \n",
       "28    168425                                     15   \n",
       "29      3843                                     31   \n",
       "30    159560                                      5   \n",
       "31    184486                                      7   \n",
       "32      6731                                      7   \n",
       "33    159953                                      5   \n",
       "34    163723                                      4   \n",
       "35      9425                                    105   \n",
       "36      4955                                      4   \n",
       "37    203336                                      5   \n",
       "38    206262                                     18   \n",
       "39      4331                                      6   \n",
       "\n",
       "    number of opportunities | shifting  number of opportunities  \\\n",
       "0                                   27                      148   \n",
       "1                                   20                      196   \n",
       "2                                    6                      133   \n",
       "3                                   28                      166   \n",
       "4                                   14                       62   \n",
       "5                                    4                       46   \n",
       "6                                   20                      129   \n",
       "7                                    4                       48   \n",
       "8                                   17                      129   \n",
       "9                                    3                       13   \n",
       "10                                  19                      126   \n",
       "11                                   2                       38   \n",
       "12                                   5                       23   \n",
       "13                                  13                       68   \n",
       "14                                  37                      157   \n",
       "15                                  16                      111   \n",
       "16                                  15                       46   \n",
       "17                                  19                       82   \n",
       "18                                  24                      137   \n",
       "19                                   2                       42   \n",
       "20                                   1                       12   \n",
       "21                                  15                       76   \n",
       "22                                   0                       34   \n",
       "23                                   0                       16   \n",
       "24                                  26                      135   \n",
       "25                                  34                      145   \n",
       "26                                   3                       20   \n",
       "27                                   4                       25   \n",
       "28                                   2                       17   \n",
       "29                                   2                       33   \n",
       "30                                   2                        7   \n",
       "31                                   0                        7   \n",
       "32                                   0                        7   \n",
       "33                                   1                        6   \n",
       "34                                   2                        6   \n",
       "35                                  17                      122   \n",
       "36                                   1                        5   \n",
       "37                                   0                        5   \n",
       "38                                   0                       18   \n",
       "39                                   2                        8   \n",
       "\n",
       "    outs above average | no shifting  outs above average | shifting  \\\n",
       "0                               3.33                           0.34   \n",
       "1                               1.90                           0.97   \n",
       "2                               1.72                           0.06   \n",
       "3                               1.21                           0.39   \n",
       "4                               1.13                          -0.07   \n",
       "5                               1.05                          -0.01   \n",
       "6                               1.04                          -0.18   \n",
       "7                               0.73                           0.04   \n",
       "8                               0.67                           0.09   \n",
       "9                               0.70                           0.03   \n",
       "10                              0.53                           0.19   \n",
       "11                              0.63                           0.02   \n",
       "12                              0.05                           0.57   \n",
       "13                             -0.92                           1.43   \n",
       "14                              0.08                           0.40   \n",
       "15                              0.38                           0.09   \n",
       "16                             -0.49                           0.88   \n",
       "17                              0.51                          -0.13   \n",
       "18                              0.07                           0.25   \n",
       "19                              0.22                           0.02   \n",
       "20                              0.23                           0.01   \n",
       "21                              0.32                          -0.10   \n",
       "22                              0.18                           0.00   \n",
       "23                              0.17                           0.00   \n",
       "24                              0.35                          -0.19   \n",
       "25                              0.10                           0.04   \n",
       "26                              0.07                           0.03   \n",
       "27                              0.06                           0.04   \n",
       "28                              0.15                          -0.06   \n",
       "29                              0.06                           0.02   \n",
       "30                              0.05                           0.02   \n",
       "31                              0.07                           0.00   \n",
       "32                              0.07                           0.00   \n",
       "33                              0.05                           0.01   \n",
       "34                              0.04                           0.02   \n",
       "35                              0.95                          -0.90   \n",
       "36                              0.04                           0.01   \n",
       "37                              0.05                           0.00   \n",
       "38                              0.04                           0.00   \n",
       "39                              0.02                           0.02   \n",
       "\n",
       "    outs above average  \n",
       "0                 3.67  \n",
       "1                 2.87  \n",
       "2                 1.78  \n",
       "3                 1.60  \n",
       "4                 1.06  \n",
       "5                 1.04  \n",
       "6                 0.86  \n",
       "7                 0.77  \n",
       "8                 0.76  \n",
       "9                 0.73  \n",
       "10                0.72  \n",
       "11                0.65  \n",
       "12                0.62  \n",
       "13                0.51  \n",
       "14                0.48  \n",
       "15                0.47  \n",
       "16                0.39  \n",
       "17                0.38  \n",
       "18                0.32  \n",
       "19                0.24  \n",
       "20                0.24  \n",
       "21                0.22  \n",
       "22                0.18  \n",
       "23                0.17  \n",
       "24                0.16  \n",
       "25                0.14  \n",
       "26                0.10  \n",
       "27                0.10  \n",
       "28                0.09  \n",
       "29                0.08  \n",
       "30                0.07  \n",
       "31                0.07  \n",
       "32                0.07  \n",
       "33                0.06  \n",
       "34                0.06  \n",
       "35                0.05  \n",
       "36                0.05  \n",
       "37                0.05  \n",
       "38                0.04  \n",
       "39                0.04  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build a leaderboard (the second one)\n",
    "raw_oaa_leaderboard = shortstop_data.groupby(['playerid', 'shift_or_not']).agg({'outs_above_average': ['sum', 'count']}).reset_index().droplevel(0,axis=1)\n",
    "raw_oaa_leaderboard.columns = ['playerid', 'shifting', 'outs above average', 'number of opportunities']\n",
    "\n",
    "oaa_leaderboard = pd.pivot_table(raw_oaa_leaderboard, values=['number of opportunities','outs above average'],\n",
    "               index='playerid', columns='shifting', aggfunc=np.sum, fill_value=0).reset_index()\n",
    "oaa_leaderboard.columns = oaa_leaderboard.columns.map(' | '.join).str.strip(' | ')\n",
    "\n",
    "n = oaa_leaderboard['number of opportunities | no shifting'] + oaa_leaderboard['number of opportunities | shifting']\n",
    "oaa_leaderboard.insert(loc=3, column='number of opportunities', value=n)\n",
    "oaa_leaderboard['outs above average'] = oaa_leaderboard['outs above average | no shifting'] + oaa_leaderboard['outs above average | shifting']\n",
    "oaa_leaderboard.sort_values(by='outs above average', ascending=False).reset_index(drop=True)[0:40]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 7 Out Probability\n",
    "After building the leaderboard of **Outs Above Average** for each shortstop, we are interested in focusing on each shortstop's performance. According to the leaderboard in the previous chapter, we are really curious about why the shorstop (id: 162393) and the shortstop (id: 162648) have such difference in **OAA**. In detail, given that they have similar number of opportunities, the shorstop (id: 162393) has **+3.67 OAA** while the shortstop (id: 162648) only has **+0.14 OAA**.\n",
    "\n",
    "Therefore, we decide to visualize every locations for where the batted ball first landed sperately for these two shortstops. However, on the basis of the two scatter plots, there is no essential difference. Hence, we start to recall the definition of **Outs Above Average**. That is, a fielder who catches a 25% **Out Probability** play gets +0.75; one who fails to make the play gets -0.25.\n",
    "\n",
    "As a result, we speculate that the shorstop (id: 162393) might be given out credits on more plays which have low **Out Probability**. Hence, we decide to separate each play's **Out Probability** into 5 groups. In detail,**Out Probability** larger than 80% would be groups 5; **Out Probability** between 60% and 80% would be groups 4; **Out Probability** between 40% and 60% would be groups 3; **Out Probability** between 20% and 40% would be groups 2; **Out Probability** less than 20% would be groups 1.\n",
    "\n",
    "Consequently, we build a new leaderboard of **Outs Above Average** including the playerid, the outs above average,  number of opportunities for each group, number of success for each group, and success rate for each group. According to the new leaderboard, we could figure that the reason that shorstop (id: 162393) has the highest **OAA**  is that  he is given out credits on more plays whose **Out Probability** are lower than 60%. Please also refer to Appendix for the full leaderboard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize landing location for shortstop 162393 and 162648\n",
    "compared_id = [162393, 162648]\n",
    "\n",
    "fig, ax = plt.subplots(1, 2, figsize=(16,8))\n",
    "for i in range(len(compared_id)):\n",
    "    \n",
    "    lines = [[(location_HB['x'], location_HB['y']), (160*(2**0.5), 160*(2**0.5))],\n",
    "             [(location_HB['x'], location_HB['y']), (-160*(2**0.5), 160*(2**0.5))],\n",
    "             [(location_1B['x'], location_1B['y']), (location_2B['x'], location_2B['y'])],\n",
    "             [(location_2B['x'], location_2B['y']), (location_3B['x'], location_3B['y'])]]\n",
    "    lc = mc.LineCollection(lines, colors='black', linewidths=1)\n",
    "    arc_element1 = Arc((0, 0), 548, 800, theta1=45, theta2=135, linewidth=1, color='black')\n",
    "    arc_element2 = Arc((0, 60.5), 180, 180, theta1=17, theta2=163, linewidth=1, color='black')\n",
    "  \n",
    "    x1 = shortstop_data[(shortstop_data['playerid']==compared_id[i]) & (shortstop_data['player_out_credit']==1)]['landing_location_x']\n",
    "    y1 = shortstop_data[(shortstop_data['playerid']==compared_id[i]) & (shortstop_data['player_out_credit']==1)]['landing_location_y']\n",
    "    x2 = shortstop_data[(shortstop_data['playerid']==compared_id[i]) & (shortstop_data['player_out_credit']==0)]['landing_location_x']\n",
    "    y2 = shortstop_data[(shortstop_data['playerid']==compared_id[i]) & (shortstop_data['player_out_credit']==0)]['landing_location_y']\n",
    "    \n",
    "    ax[i].axis('scaled')\n",
    "    ax[i].set_xlim([-250, 250])\n",
    "    ax[i].set_ylim([-10, 430])\n",
    "    ax[i].add_collection(lc)\n",
    "    ax[i].add_patch(arc_element1)\n",
    "    ax[i].add_patch(arc_element2)\n",
    "    ax[i].scatter(x1, y1, color='red', alpha=1, s=6, label='out')\n",
    "    ax[i].scatter(x2, y2, color='blue', alpha=1, s=6, label='hit / error')\n",
    "    ax[i].legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a function to separte each play's out probability into 5 groups\n",
    "def out_probability_group_function(data):\n",
    "    \n",
    "    group = []\n",
    "    n = data.shape[0]\n",
    "    for i in range(n):\n",
    "        if data['out_probability'][i] < 0.2:\n",
    "            group.append('group 1')\n",
    "        elif (data['out_probability'][i] >= 0.2) & (data['out_probability'][i] < 0.4):\n",
    "            group.append('group 2')\n",
    "        elif (data['out_probability'][i] >= 0.4) & (data['out_probability'][i] < 0.6):\n",
    "            group.append('group 3')\n",
    "        elif (data['out_probability'][i] >= 0.6) & (data['out_probability'][i] < 0.8):\n",
    "            group.append('group 4')\n",
    "        elif data['out_probability'][i] >= 0.8:\n",
    "            group.append('group 5')\n",
    "    return group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separte each play's out probability into 5 groups\n",
    "out_probability_group = out_probability_group_function(shortstop_data)\n",
    "shortstop_data = shortstop_data.assign(out_probability_group=out_probability_group)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>playerid</th>\n",
       "      <th>number of success | group 1</th>\n",
       "      <th>number of opportunities | group 1</th>\n",
       "      <th>success rate | group 1</th>\n",
       "      <th>number of success | group 2</th>\n",
       "      <th>number of opportunities | group 2</th>\n",
       "      <th>success rate | group 2</th>\n",
       "      <th>number of success | group 3</th>\n",
       "      <th>number of opportunities | group 3</th>\n",
       "      <th>success rate | group 3</th>\n",
       "      <th>number of success | group 4</th>\n",
       "      <th>number of opportunities | group 4</th>\n",
       "      <th>success rate | group 4</th>\n",
       "      <th>number of success | group 5</th>\n",
       "      <th>number of opportunities | group 5</th>\n",
       "      <th>success rate | group 5</th>\n",
       "      <th>outs above average</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>162393</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>8.33</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>40.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>128</td>\n",
       "      <td>128</td>\n",
       "      <td>100.00</td>\n",
       "      <td>3.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>167960</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>0.00</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>100.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>164</td>\n",
       "      <td>165</td>\n",
       "      <td>99.39</td>\n",
       "      <td>2.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>206311</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>9.52</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>109</td>\n",
       "      <td>109</td>\n",
       "      <td>100.00</td>\n",
       "      <td>1.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>154448</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>50.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>147</td>\n",
       "      <td>147</td>\n",
       "      <td>100.00</td>\n",
       "      <td>1.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2087</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52</td>\n",
       "      <td>52</td>\n",
       "      <td>100.00</td>\n",
       "      <td>1.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>9424</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>41</td>\n",
       "      <td>41</td>\n",
       "      <td>100.00</td>\n",
       "      <td>1.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>5393</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>25.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>112</td>\n",
       "      <td>112</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.86</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>9210</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>37</td>\n",
       "      <td>37</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>5419</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>113</td>\n",
       "      <td>113</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>169390</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>9074</td>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>12.50</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>114</td>\n",
       "      <td>114</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>6619</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>158358</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>18</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>171885</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>50.0</td>\n",
       "      <td>51</td>\n",
       "      <td>51</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>161551</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>141</td>\n",
       "      <td>142</td>\n",
       "      <td>99.30</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>9148</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>102</td>\n",
       "      <td>102</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>207428</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>50.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>39</td>\n",
       "      <td>39</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>7580</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>77</td>\n",
       "      <td>77</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>132551</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>50.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>121</td>\n",
       "      <td>121</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>162294</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>38</td>\n",
       "      <td>38</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>4379</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>113301</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>69</td>\n",
       "      <td>69</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>159128</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30</td>\n",
       "      <td>30</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>3295</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>11869</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>114</td>\n",
       "      <td>114</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>162648</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>133</td>\n",
       "      <td>133</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>172568</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22</td>\n",
       "      <td>22</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>167746</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18</td>\n",
       "      <td>18</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>168425</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16</td>\n",
       "      <td>16</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>3843</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28</td>\n",
       "      <td>28</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>159560</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>184486</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>6731</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>163723</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>159953</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>9425</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>100.0</td>\n",
       "      <td>102</td>\n",
       "      <td>103</td>\n",
       "      <td>99.03</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>4955</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>203336</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>4331</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>7393</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>100.00</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    playerid  number of success | group 1  number of opportunities | group 1  \\\n",
       "0     162393                            1                                 12   \n",
       "1     167960                            0                                 26   \n",
       "2     206311                            2                                 21   \n",
       "3     154448                            0                                 16   \n",
       "4       2087                            0                                  8   \n",
       "5       9424                            0                                  4   \n",
       "6       5393                            0                                 12   \n",
       "7       9210                            0                                 10   \n",
       "8       5419                            0                                 15   \n",
       "9     169390                            0                                  2   \n",
       "10      9074                            1                                  8   \n",
       "11      6619                            0                                  9   \n",
       "12    158358                            0                                  4   \n",
       "13    171885                            0                                 13   \n",
       "14    161551                            0                                 13   \n",
       "15      9148                            0                                  8   \n",
       "16    207428                            0                                  5   \n",
       "17      7580                            0                                  4   \n",
       "18    132551                            0                                 11   \n",
       "19    162294                            0                                  4   \n",
       "20      4379                            0                                  2   \n",
       "21    113301                            0                                  7   \n",
       "22    159128                            0                                  4   \n",
       "23      3295                            0                                  0   \n",
       "24     11869                            0                                 19   \n",
       "25    162648                            0                                 11   \n",
       "26    172568                            0                                  3   \n",
       "27    167746                            0                                  2   \n",
       "28    168425                            0                                  1   \n",
       "29      3843                            0                                  5   \n",
       "30    159560                            0                                  0   \n",
       "31    184486                            0                                  0   \n",
       "32      6731                            0                                  0   \n",
       "33    163723                            0                                  0   \n",
       "34    159953                            0                                  0   \n",
       "35      9425                            0                                 16   \n",
       "36      4955                            0                                  0   \n",
       "37    203336                            0                                  0   \n",
       "38      4331                            0                                  1   \n",
       "39      7393                            0                                  0   \n",
       "\n",
       "    success rate | group 1  number of success | group 2  \\\n",
       "0                     8.33                            2   \n",
       "1                     0.00                            3   \n",
       "2                     9.52                            0   \n",
       "3                     0.00                            1   \n",
       "4                     0.00                            0   \n",
       "5                     0.00                            1   \n",
       "6                     0.00                            1   \n",
       "7                     0.00                            1   \n",
       "8                     0.00                            0   \n",
       "9                     0.00                            1   \n",
       "10                   12.50                            0   \n",
       "11                    0.00                            1   \n",
       "12                    0.00                            1   \n",
       "13                    0.00                            1   \n",
       "14                    0.00                            0   \n",
       "15                    0.00                            0   \n",
       "16                    0.00                            1   \n",
       "17                    0.00                            0   \n",
       "18                    0.00                            0   \n",
       "19                    0.00                            0   \n",
       "20                    0.00                            0   \n",
       "21                    0.00                            0   \n",
       "22                    0.00                            0   \n",
       "23                    0.00                            0   \n",
       "24                    0.00                            0   \n",
       "25                    0.00                            0   \n",
       "26                    0.00                            0   \n",
       "27                    0.00                            0   \n",
       "28                    0.00                            0   \n",
       "29                    0.00                            0   \n",
       "30                    0.00                            0   \n",
       "31                    0.00                            0   \n",
       "32                    0.00                            0   \n",
       "33                    0.00                            0   \n",
       "34                    0.00                            0   \n",
       "35                    0.00                            1   \n",
       "36                    0.00                            0   \n",
       "37                    0.00                            0   \n",
       "38                    0.00                            0   \n",
       "39                    0.00                            0   \n",
       "\n",
       "    number of opportunities | group 2  success rate | group 2  \\\n",
       "0                                   5                    40.0   \n",
       "1                                   3                   100.0   \n",
       "2                                   2                     0.0   \n",
       "3                                   2                    50.0   \n",
       "4                                   0                     0.0   \n",
       "5                                   1                   100.0   \n",
       "6                                   4                    25.0   \n",
       "7                                   1                   100.0   \n",
       "8                                   0                     0.0   \n",
       "9                                   1                   100.0   \n",
       "10                                  4                     0.0   \n",
       "11                                  1                   100.0   \n",
       "12                                  1                   100.0   \n",
       "13                                  1                   100.0   \n",
       "14                                  0                     0.0   \n",
       "15                                  1                     0.0   \n",
       "16                                  2                    50.0   \n",
       "17                                  1                     0.0   \n",
       "18                                  2                     0.0   \n",
       "19                                  0                     0.0   \n",
       "20                                  1                     0.0   \n",
       "21                                  0                     0.0   \n",
       "22                                  0                     0.0   \n",
       "23                                  0                     0.0   \n",
       "24                                  1                     0.0   \n",
       "25                                  1                     0.0   \n",
       "26                                  0                     0.0   \n",
       "27                                  0                     0.0   \n",
       "28                                  0                     0.0   \n",
       "29                                  0                     0.0   \n",
       "30                                  0                     0.0   \n",
       "31                                  0                     0.0   \n",
       "32                                  0                     0.0   \n",
       "33                                  0                     0.0   \n",
       "34                                  0                     0.0   \n",
       "35                                  1                   100.0   \n",
       "36                                  0                     0.0   \n",
       "37                                  0                     0.0   \n",
       "38                                  0                     0.0   \n",
       "39                                  0                     0.0   \n",
       "\n",
       "    number of success | group 3  number of opportunities | group 3  \\\n",
       "0                             3                                  3   \n",
       "1                             2                                  2   \n",
       "2                             1                                  1   \n",
       "3                             1                                  1   \n",
       "4                             2                                  2   \n",
       "5                             0                                  0   \n",
       "6                             1                                  1   \n",
       "7                             0                                  0   \n",
       "8                             0                                  0   \n",
       "9                             0                                  0   \n",
       "10                            0                                  0   \n",
       "11                            0                                  0   \n",
       "12                            0                                  0   \n",
       "13                            1                                  1   \n",
       "14                            1                                  1   \n",
       "15                            0                                  0   \n",
       "16                            0                                  0   \n",
       "17                            0                                  0   \n",
       "18                            1                                  2   \n",
       "19                            0                                  0   \n",
       "20                            1                                  1   \n",
       "21                            0                                  0   \n",
       "22                            0                                  0   \n",
       "23                            0                                  0   \n",
       "24                            0                                  0   \n",
       "25                            0                                  0   \n",
       "26                            0                                  0   \n",
       "27                            0                                  0   \n",
       "28                            0                                  0   \n",
       "29                            0                                  0   \n",
       "30                            0                                  0   \n",
       "31                            0                                  0   \n",
       "32                            0                                  0   \n",
       "33                            0                                  0   \n",
       "34                            0                                  0   \n",
       "35                            0                                  1   \n",
       "36                            0                                  0   \n",
       "37                            0                                  0   \n",
       "38                            0                                  0   \n",
       "39                            0                                  0   \n",
       "\n",
       "    success rate | group 3  number of success | group 4  \\\n",
       "0                    100.0                            0   \n",
       "1                    100.0                            0   \n",
       "2                    100.0                            0   \n",
       "3                    100.0                            0   \n",
       "4                    100.0                            0   \n",
       "5                      0.0                            0   \n",
       "6                    100.0                            0   \n",
       "7                      0.0                            0   \n",
       "8                      0.0                            1   \n",
       "9                      0.0                            0   \n",
       "10                     0.0                            0   \n",
       "11                     0.0                            0   \n",
       "12                     0.0                            0   \n",
       "13                   100.0                            1   \n",
       "14                   100.0                            1   \n",
       "15                     0.0                            0   \n",
       "16                     0.0                            0   \n",
       "17                     0.0                            0   \n",
       "18                    50.0                            1   \n",
       "19                     0.0                            0   \n",
       "20                   100.0                            0   \n",
       "21                     0.0                            0   \n",
       "22                     0.0                            0   \n",
       "23                     0.0                            0   \n",
       "24                     0.0                            1   \n",
       "25                     0.0                            0   \n",
       "26                     0.0                            0   \n",
       "27                     0.0                            0   \n",
       "28                     0.0                            0   \n",
       "29                     0.0                            0   \n",
       "30                     0.0                            0   \n",
       "31                     0.0                            0   \n",
       "32                     0.0                            0   \n",
       "33                     0.0                            0   \n",
       "34                     0.0                            0   \n",
       "35                     0.0                            1   \n",
       "36                     0.0                            0   \n",
       "37                     0.0                            0   \n",
       "38                     0.0                            0   \n",
       "39                     0.0                            0   \n",
       "\n",
       "    number of opportunities | group 4  success rate | group 4  \\\n",
       "0                                   0                     0.0   \n",
       "1                                   0                     0.0   \n",
       "2                                   0                     0.0   \n",
       "3                                   0                     0.0   \n",
       "4                                   0                     0.0   \n",
       "5                                   0                     0.0   \n",
       "6                                   0                     0.0   \n",
       "7                                   0                     0.0   \n",
       "8                                   1                   100.0   \n",
       "9                                   0                     0.0   \n",
       "10                                  0                     0.0   \n",
       "11                                  0                     0.0   \n",
       "12                                  0                     0.0   \n",
       "13                                  2                    50.0   \n",
       "14                                  1                   100.0   \n",
       "15                                  0                     0.0   \n",
       "16                                  0                     0.0   \n",
       "17                                  0                     0.0   \n",
       "18                                  1                   100.0   \n",
       "19                                  0                     0.0   \n",
       "20                                  0                     0.0   \n",
       "21                                  0                     0.0   \n",
       "22                                  0                     0.0   \n",
       "23                                  0                     0.0   \n",
       "24                                  1                   100.0   \n",
       "25                                  0                     0.0   \n",
       "26                                  0                     0.0   \n",
       "27                                  0                     0.0   \n",
       "28                                  0                     0.0   \n",
       "29                                  0                     0.0   \n",
       "30                                  0                     0.0   \n",
       "31                                  0                     0.0   \n",
       "32                                  0                     0.0   \n",
       "33                                  0                     0.0   \n",
       "34                                  0                     0.0   \n",
       "35                                  1                   100.0   \n",
       "36                                  0                     0.0   \n",
       "37                                  0                     0.0   \n",
       "38                                  0                     0.0   \n",
       "39                                  0                     0.0   \n",
       "\n",
       "    number of success | group 5  number of opportunities | group 5  \\\n",
       "0                           128                                128   \n",
       "1                           164                                165   \n",
       "2                           109                                109   \n",
       "3                           147                                147   \n",
       "4                            52                                 52   \n",
       "5                            41                                 41   \n",
       "6                           112                                112   \n",
       "7                            37                                 37   \n",
       "8                           113                                113   \n",
       "9                            10                                 10   \n",
       "10                          114                                114   \n",
       "11                           28                                 28   \n",
       "12                           18                                 18   \n",
       "13                           51                                 51   \n",
       "14                          141                                142   \n",
       "15                          102                                102   \n",
       "16                           39                                 39   \n",
       "17                           77                                 77   \n",
       "18                          121                                121   \n",
       "19                           38                                 38   \n",
       "20                            8                                  8   \n",
       "21                           69                                 69   \n",
       "22                           30                                 30   \n",
       "23                           16                                 16   \n",
       "24                          114                                114   \n",
       "25                          133                                133   \n",
       "26                           22                                 22   \n",
       "27                           18                                 18   \n",
       "28                           16                                 16   \n",
       "29                           28                                 28   \n",
       "30                            7                                  7   \n",
       "31                            7                                  7   \n",
       "32                            7                                  7   \n",
       "33                            6                                  6   \n",
       "34                            6                                  6   \n",
       "35                          102                                103   \n",
       "36                            5                                  5   \n",
       "37                            5                                  5   \n",
       "38                            7                                  7   \n",
       "39                            4                                  4   \n",
       "\n",
       "    success rate | group 5  outs above average  \n",
       "0                   100.00                3.67  \n",
       "1                    99.39                2.87  \n",
       "2                   100.00                1.78  \n",
       "3                   100.00                1.60  \n",
       "4                   100.00                1.06  \n",
       "5                   100.00                1.04  \n",
       "6                   100.00                0.86  \n",
       "7                   100.00                0.77  \n",
       "8                   100.00                0.76  \n",
       "9                   100.00                0.73  \n",
       "10                  100.00                0.72  \n",
       "11                  100.00                0.65  \n",
       "12                  100.00                0.62  \n",
       "13                  100.00                0.51  \n",
       "14                   99.30                0.48  \n",
       "15                  100.00                0.47  \n",
       "16                  100.00                0.39  \n",
       "17                  100.00                0.38  \n",
       "18                  100.00                0.32  \n",
       "19                  100.00                0.24  \n",
       "20                  100.00                0.24  \n",
       "21                  100.00                0.22  \n",
       "22                  100.00                0.18  \n",
       "23                  100.00                0.17  \n",
       "24                  100.00                0.16  \n",
       "25                  100.00                0.14  \n",
       "26                  100.00                0.10  \n",
       "27                  100.00                0.10  \n",
       "28                  100.00                0.09  \n",
       "29                  100.00                0.08  \n",
       "30                  100.00                0.07  \n",
       "31                  100.00                0.07  \n",
       "32                  100.00                0.07  \n",
       "33                  100.00                0.06  \n",
       "34                  100.00                0.06  \n",
       "35                   99.03                0.05  \n",
       "36                  100.00                0.05  \n",
       "37                  100.00                0.05  \n",
       "38                  100.00                0.04  \n",
       "39                  100.00                0.04  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build a leaderboard (the third one)\n",
    "raw_out_prob_leaderboard = shortstop_data.groupby(['playerid', 'out_probability_group']).agg({'player_out_credit': ['sum', 'count'], 'outs_above_average': 'sum'}).reset_index()#.droplevel(0,axis=1)\n",
    "raw_out_prob_leaderboard.columns = ['playerid', 'out probability group', 'number of success', 'number of opportunities', 'outs above average']\n",
    "\n",
    "out_prob_leaderboard = pd.pivot_table(raw_out_prob_leaderboard, values=['number of success', 'number of opportunities', 'outs above average'],\n",
    "               index='playerid', columns='out probability group', aggfunc=np.sum, fill_value=0).reset_index()\n",
    "out_prob_leaderboard.columns = out_prob_leaderboard.columns.map(' | '.join).str.strip(' | ')\n",
    "\n",
    "out_prob_leaderboard['outs above average'] = out_prob_leaderboard['outs above average | group 1'] + out_prob_leaderboard['outs above average | group 2'] + out_prob_leaderboard['outs above average | group 3'] + out_prob_leaderboard['outs above average | group 4'] + out_prob_leaderboard['outs above average | group 5']\n",
    "out_prob_leaderboard.drop(['outs above average | group 1', 'outs above average | group 2', 'outs above average | group 3', 'outs above average | group 4', 'outs above average | group 5'], axis=1)\n",
    "\n",
    "out_prob_leaderboard = out_prob_leaderboard[['playerid',\n",
    "                                             'number of success | group 1',\n",
    "                                             'number of opportunities | group 1',\n",
    "                                             'number of success | group 2',\n",
    "                                             'number of opportunities | group 2',\n",
    "                                             'number of success | group 3',\n",
    "                                             'number of opportunities | group 3',\n",
    "                                             'number of success | group 4',\n",
    "                                             'number of opportunities | group 4',\n",
    "                                             'number of success | group 5',\n",
    "                                             'number of opportunities | group 5',\n",
    "                                             'outs above average']]\n",
    "\n",
    "for i in range(1,6):\n",
    "    r = round(out_prob_leaderboard['number of success | group {}'.format(6-i)]*100/out_prob_leaderboard['number of opportunities | group {}'.format(6-i)], 2)\n",
    "    out_prob_leaderboard.insert(loc=13-2*i, column='success rate | group {}'.format(6-i), value=r)\n",
    "\n",
    "out_prob_leaderboard.fillna(0).sort_values(by='outs above average', ascending=False).reset_index(drop=True)[0:40]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chapter 8 Conclusions\n",
    "In this task, in order to estimate **Out Probability** for each in-play ground ball first fielded by shortstop and further calculating **Outs Above Average** for each shortstop, we basically develop a reliable framework including creating features, building predictive models, tuning parameters, evaluating the models, calculating feature importance, building **Outs Above Average** leadboards and trying to get some insights from the leadboards.\n",
    "\n",
    "Apparently, infielders pose a more complex challenge than outfielders on calculating **Outs Above Average** since we have to consider more situations on the field. For example, runner speed and double plays are still not considered in this task. Therefore, we truly believe that we could keep improving and developing our framework in the future once we get more data and better understand how every potential factors affect **Out Probability** and **Outs Above Average**."
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}