MLB_Algorithm_Optimizer.py

# Author 
  # Brian Schroeder

# Synopsis
  # Get's the stat algorithm that leads to the highest percentage of winning for the projected team.
  
# Description
  # Each stat considered in the algorithm has been selected based off of the most common categories that the Winning Team Lead in after the victory.
  # The MLBWinAnalysis.py script gets all games for a range of dates and gets all the stats the team lead in with victory, which is used in this script to determine the most
  # effective combination of stats to predict the winning team
  
  # For Example, one simple algorithm is Home Batting average vs Away Batting Average and Home ERA vs Away ERA. This script will output the ammount of times 
  # the team has won factoring the simple algorithm provided. This script will do this for all of the possible unique algorithms and output the most effective combination.

from itertools import combinations
import requests
import json
from statistics import mean
from collections import Counter
import datetime
import pandas as pd

#Format for Dates: Day,Month,Year
startDate = "01-04-2019"
endDate = "30-11-2019"

def dateRange(startDate,endDate):
    dates = []
    start = datetime.datetime.strptime(f"{startDate}", "%d-%m-%Y")
    end = datetime.datetime.strptime(f"{endDate}", "%d-%m-%Y")
    date_generated = [start + datetime.timedelta(days=x) for x in range(0, (end-start).days)]
    for date in date_generated:
        dates.append(date.strftime("%m/%d/%Y"))
    return(dates)

gameDates = dateRange(startDate,endDate)

analysis_comparison = []

for gameDate in gameDates:
    # Create Arrays
    mlb_teamStats = []
    mlb_advantages = []
    teamAdvatage = []
    projectedOutcome = []
    game_ids = []

    try:
        request = requests.get(f"http://statsapi.mlb.com/api/v1/schedule/games/?sportId=1&date={gameDate}").text
        request_json = json.loads(request)
        games = (request_json['dates'][0]['games'])
        for game in games:
            game_ids.append(game['gamePk'])
    except:
        continue

    for games in game_ids:
        homeBA = []
        awayBA = []
        homeSLG = []
        awaySLG = []
        homeOBP = []
        awayOBP = []
        homeSO = []
        awaySO = []
        game_analyzer = []

        request = requests.get(f"https://statsapi.mlb.com/api/v1/schedule?gamePk={games}&language=en&hydrate=lineups").text
        games_request_json = json.loads(request)
        try:
            (games_request_json['dates'][0]['games'][0]['lineups']['homePlayers'])
        except:
            continue

        try:
            (games_request_json['dates'][0]['games'][0]['lineups']['awayPlayers'])
        except:
            continue

        homeTeamName = (games_request_json['dates'][0]['games'][0]['teams']['home']['team']['name'])
        awayTeamName = (games_request_json['dates'][0]['games'][0]['teams']['away']['team']['name'])
        homeTeam = (games_request_json['dates'][0]['games'][0]['lineups']['homePlayers'])
        awayTeam = (games_request_json['dates'][0]['games'][0]['lineups']['awayPlayers'])

        # Getting Batters Career Averages
        for player in homeTeam:
            try:
                id = (player['id'])
                request = requests.get(f"http://lookup-service-prod.mlb.com/json/named.sport_career_hitting.bam?league_list_id='mlb'&game_type='R'&player_id='{id}'").text
                request_json = json.loads(request)
                playerStats = (request_json['sport_career_hitting']['queryResults']['row'])
                homeBA.append(float(playerStats['avg']))
                homeSLG.append(float(playerStats['slg']))
                homeOBP.append(float(playerStats['obp']))
                homeSO.append(float(playerStats['so'])/float(playerStats['ab']))
            except:
                continue
        for player in awayTeam:
            try:
                id = (player['id'])
                request = requests.get(f"http://lookup-service-prod.mlb.com/json/named.sport_career_hitting.bam?league_list_id='mlb'&game_type='R'&player_id='{id}'").text
                request_json = json.loads(request)
                playerStats = (request_json['sport_career_hitting']['queryResults']['row'])
                awayBA.append(float(playerStats['avg']))
                awaySLG.append(float(playerStats['slg']))
                awayOBP.append(float(playerStats['obp']))
                awaySO.append(float(playerStats['so'])/float(playerStats['ab']))
            except:
                continue

        # Get Starting Pitcher Stats
        try:
            pitcherRequest = requests.get(f"http://statsapi.mlb.com/api/v1/game/{games}/boxscore").text
            pitcher_json = json.loads(pitcherRequest)

            homePitcher = pitcher_json['teams']['home']['pitchers'][0]
            awayPitcher = pitcher_json['teams']['away']['pitchers'][0]

            homePitcherRequest = requests.get(f"http://lookup-service-prod.mlb.com/json/named.sport_career_pitching.bam?league_list_id='mlb'&game_type='R'&player_id='{homePitcher}'").text
            homepitcherRequest_json = json.loads(homePitcherRequest)
            awayPitcherRequest = requests.get(f"http://lookup-service-prod.mlb.com/json/named.sport_career_pitching.bam?league_list_id='mlb'&game_type='R'&player_id='{awayPitcher}'").text
            awaypitcherRequest_json = json.loads(awayPitcherRequest)

            homePitcherStats = (homepitcherRequest_json['sport_career_pitching']['queryResults']['row'])
            awayPitcherStats = (awaypitcherRequest_json['sport_career_pitching']['queryResults']['row'])
        except:
            continue

        advantages = {
            'Home Team': homeTeamName,
            "Away Team": awayTeamName,
            "Home BA": round(mean(homeBA) - mean(awayBA), 3),
            "Home Slugging %": round(mean(homeSLG) - mean(awaySLG), 3),
            "Home OBP %": round(mean(homeOBP) - mean(awayOBP), 3),
            "Home SO %": round(mean(homeSO) - mean(awaySO), 3) * -1,
            "Home ERA": round(float(homePitcherStats['era']) - float(awayPitcherStats['era']), 3) * -1,
            "Home WHIP": round(float(homePitcherStats['whip']) - float(awayPitcherStats['whip']), 3) * -1,
            "Home OBP Against": (round(float(homePitcherStats['obp']), 2) - round(float(awayPitcherStats['obp']), 2)) * -1,
            "Home Homeruns/9 Against": (round(float(homePitcherStats['h9']), 2) - round(float(awayPitcherStats['h9']),2)) * -1,
            "Home BB/9 Against": (round(float(homePitcherStats['bb9']), 2) - round(float(awayPitcherStats['bb9']), 2)) * -1,
            "Away BA": round(mean(awayBA) - mean(homeBA), 3),
            "Away Slugging %": round(mean(awaySLG) - mean(homeSLG), 3),
            "Away OBP %": round(mean(awayOBP) - mean(homeOBP), 3),
            "Away SO %": round(mean(awaySO) - mean(homeSO), 3)  * -1,
            "Away ERA": round(float(awayPitcherStats['era']) - float(homePitcherStats['era']), 3) * -1,
            "Away WHIP": round(float(awayPitcherStats['whip']) - float(homePitcherStats['whip']), 3) * -1,
            "Away OBP Against": (round(float(awayPitcherStats['obp']), 2) - round(float(homePitcherStats['obp']), 2)) * -1,
            "Away Homeruns/9 Against": (round(float(awayPitcherStats['h9']), 2) - round(float(homePitcherStats['h9']),2)) * -1,
            "Away BB/9 Against": (round(float(awayPitcherStats['bb9']), 2) - round(float(homePitcherStats['bb9']), 2)) * -1,
        }

        homeAdvantage = advantages['Home BA'] + advantages['Home SO %'] + advantages['Home Slugging %'] + advantages['Home OBP %'] + advantages['Home ERA'] + advantages['Home WHIP'] + advantages['Home Homeruns/9 Against'] + advantages['Home OBP Against'] + advantages['Home BB/9 Against']
        awayAdvantage = advantages['Away BA'] + advantages['Away SO %'] + advantages['Away Slugging %'] + advantages['Away OBP %'] + advantages['Away ERA'] + advantages['Away WHIP'] + advantages['Away Homeruns/9 Against'] + advantages['Away OBP Against'] + advantages['Away BB/9 Against']


        try:
            if (games_request_json['dates'][0]['games'][0]['teams']['home']['isWinner']) == True:
                homeWinner = 1
            elif (games_request_json['dates'][0]['games'][0]['teams']['home']['isWinner']) == False:
                homeWinner = 0
            else:
                homeWinner = 'TBD'

            if (games_request_json['dates'][0]['games'][0]['teams']['away']['isWinner']) == True:
                awayWinner = 1
            elif (games_request_json['dates'][0]['games'][0]['teams']['away']['isWinner']) == False:
                awayWinner = 0
            else:
                awayWinner = 'TBD'
        except:
            continue

        # Define Stats to Create Algorithm for
        homeInput = "advantages['Home BA'] > advantages['Away BA']", "advantages['Home SO %'] > advantages['Away SO %']", "advantages['Home Slugging %'] > advantages['Away Slugging %']", "advantages['Home OBP %'] > advantages['Away OBP %']", "advantages['Home ERA'] > advantages['Away ERA']", "advantages['Home WHIP'] > advantages['Away WHIP']", "advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against']", "advantages['Home OBP Against'] > advantages['Away OBP Against'] ", "advantages['Home BB/9 Against'] > advantages['Away BB/9 Against']"
        awayInput = "advantages['Away BA'] > advantages['Home BA']", "advantages['Away SO %'] > advantages['Home SO %']", "advantages['Away Slugging %'] > advantages['Home Slugging %']", "advantages['Away OBP %'] > advantages['Home OBP %']", "advantages['Away ERA'] > advantages['Home ERA']", "advantages['Away WHIP'] > advantages['Home WHIP']", "advantages['Away Homeruns/9 Against'] > advantages['Home Homeruns/9 Against']", "advantages['Away OBP Against'] > advantages['Home OBP Against'] ", "advantages['Away BB/9 Against'] > advantages['Home BB/9 Against']"

        # Get All Possible Combinations of Stats to be Analyzed
        homeoutput = sum([list(map(list, combinations(homeInput, i))) for i in range(len(homeInput) + 1)], [])
        awayoutput = sum([list(map(list, combinations(awayInput, i))) for i in range(len(awayInput) + 1)], [])


        # Put all Possible Combinations in Operator Format
        for stat_Array in homeoutput:
            algorithm = ''
            for stat in stat_Array:
                if (len(stat_Array)) == 1:
                    algorithm += (stat)
                else:
                    stat = stat + ' and '
                    algorithm += stat
            if 'and' in algorithm:
                algorithm = algorithm[:-5]

            try:
                if eval(algorithm):
                    projectedWinner = {
                        'Projected Winner': advantages['Home Team'],
                        'Winner Advantage (Beta)': homeAdvantage,
                        'Opponent': advantages['Away Team'],
                        'Opponent Deficit': awayAdvantage,
                        'Did Projected Win': homeWinner,
                        'Did Opponent Win': awayWinner,
                        'Score': f"{advantages['Home Team']}: {games_request_json['dates'][0]['games'][0]['teams']['home']['score']}  {advantages['Away Team']}: {games_request_json['dates'][0]['games'][0]['teams']['away']['score']} ",
                        'Algorithm': algorithm
                    }
                    analysis_comparison.append(projectedWinner)
            except:
                continue

        # Put all Possible Combinations in Operator Format
        for stat_Array in awayoutput:
            algorithm = ''
            for stat in stat_Array:
                if (len(stat_Array)) == 1:
                    algorithm += (stat)
                else:
                    stat = stat + ' and '
                    algorithm += stat
            if 'and' in algorithm:
                algorithm = algorithm[:-5]

            try:
                if eval(algorithm):
                    projectedWinner = {
                        'Projected Winner': advantages['Away Team'],
                        'Winner Advantage (Beta)': awayAdvantage,
                        'Opponent': advantages['Home Team'],
                        'Opponent Deficit': homeAdvantage,
                        'Did Projected Win': awayWinner,
                        'Did Opponent Win': homeWinner,
                        'Score': f"{advantages['Away Team']}: {games_request_json['dates'][0]['games'][0]['teams']['away']['score']}  {advantages['Home Team']}: {games_request_json['dates'][0]['games'][0]['teams']['home']['score']} ",
                        'Algorithm': algorithm
                    }
                    analysis_comparison.append(projectedWinner)
            except:
                continue


df = pd.DataFrame(analysis_comparison, columns=['Algorithm', 'Did Projected Win', 'Did Opponent Win'])
df_grouped = df.groupby(['Algorithm']).sum()
df_algorithm_stats = (df_grouped.sort_values(by=['Did Projected Win'], ascending=False))
total_games = df_algorithm_stats["Did Projected Win"] + df_algorithm_stats["Did Opponent Win"]
df_algorithm_stats["Total Games"] = total_games
winPercentage = df_algorithm_stats["Did Projected Win"] / df_algorithm_stats["Total Games"]
df_algorithm_stats["Win Percentage"] = winPercentage
df_algorithm_winpercentage_sorted = (df_algorithm_stats.sort_values(by=['Win Percentage'], ascending=False))
pd.set_option('display.max_rows', None)
df_algorithm_winpercentage_sorted.to_csv(r'test.csv',  header=True)

# Algorithm	Projected Won	Opponent Won	Total Games	Win Percentage
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against'] 	236	92	328	71.95%
# advantages['Home BA'] > advantages['Away BA'] and advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home OBP %'] > advantages['Away OBP %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home WHIP'] > advantages['Away WHIP']	234	92	326	71.78%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home WHIP'] > advantages['Away WHIP'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against'] 	234	92	326	71.78%
# advantages['Home BA'] > advantages['Away BA'] and advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against'] 	210	83	293	71.67%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home WHIP'] > advantages['Away WHIP'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against']  and advantages['Home BB/9 Against'] > advantages['Away BB/9 Against']	182	72	254	71.65%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against']  and advantages['Home BB/9 Against'] > advantages['Away BB/9 Against']	182	72	254	71.65%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home OBP %'] > advantages['Away OBP %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home WHIP'] > advantages['Away WHIP']	250	99	349	71.63%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home WHIP'] > advantages['Away WHIP'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against']  and advantages['Home BB/9 Against'] > advantages['Away BB/9 Against']	194	77	271	71.59%
# advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against']  and advantages['Home BB/9 Against'] > advantages['Away BB/9 Against']	194	77	271	71.59%
# advantages['Home BA'] > advantages['Away BA'] and advantages['Home SO %'] > advantages['Away SO %'] and advantages['Home Slugging %'] > advantages['Away Slugging %'] and advantages['Home ERA'] > advantages['Away ERA'] and advantages['Home WHIP'] > advantages['Away WHIP'] and advantages['Home Homeruns/9 Against'] > advantages['Away Homeruns/9 Against'] and advantages['Home OBP Against'] > advantages['Away OBP Against'] 	208	83	291	71.48%