Created
July 31, 2021 19:48
-
-
Save daveklotz/83f0da28919fcd383d81707255544225 to your computer and use it in GitHub Desktop.
Generate the data and run the ANN so I can get picks for today's games
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| ###### | |
| # NOTE: this includes calculaiton for my first Wins Above Replacement approach, as well as the current DRA- approach | |
| ###### | |
| games = mlbgame.day(2021,7,30) | |
| daily_game_id = 1 | |
| unique_row_id = 1. # just to make sure i have the same format as legacy spreadsheet | |
| location_index = 0 | |
| pix_index = 0 | |
| for game in games: | |
| # First for AWAY team, then do the same thing for HOME team | |
| team = game.away_team | |
| h1v0 = 0 | |
| opp = game.home_team | |
| #API is sending back home starter as visitor, and visitor starter as home... | |
| # and sometimes doesnt have one even for future games | |
| try: | |
| team_starter = game.p_pitcher_home | |
| opp_starter = game.p_pitcher_away | |
| #print(team_starter) | |
| team_starter_war = get_starter_war(pecota_df, team_starter) | |
| opp_starter_war = get_starter_war(pecota_df, opp_starter) | |
| team_starter_dra_minus = get_starter_dra_minus(pecota_df, team_starter) | |
| opp_starter_dra_minus = get_starter_dra_minus(pecota_df, opp_starter) | |
| except AttributeError: | |
| team_starter = "NONE" | |
| opp_starter = "NONE" | |
| team_starter_war = -888 | |
| opp_starter_war = -888 | |
| team_starter_dra_minus = 888 | |
| opp_starter_dra_minus = 888 | |
| money_line = get_money_line(odds_df, team) | |
| implied_win_prob = bu.money_line_to_win_prob(money_line) | |
| team_win_pct = get_win_pct_nickname(teams, injury_df, team, "PECOTA Win Pct 7/5") | |
| opp_win_pct = get_win_pct_nickname(teams, injury_df, opp, "PECOTA Win Pct 7/5") | |
| win_pct_diff = team_win_pct - opp_win_pct | |
| starter_war_diff = team_starter_war - opp_starter_war | |
| starter_dra_minus_diff = team_starter_dra_minus - opp_starter_dra_minus | |
| prediction_warp_based = ann_warp.predict([[h1v0, starter_war_diff, win_pct_diff]]) | |
| predicted_win_prob_warp = prediction_warp_based[0][0] | |
| team_1_predicted_pct_warp = predicted_win_prob_warp | |
| key_vars = [h1v0, win_pct_diff, starter_dra_minus_diff ] | |
| key_vars = dra_minus_scaler.transform([key_vars]) | |
| prediction_dra_minus_based = ann_dra_minus.predict([key_vars]) | |
| predicted_win_prob_dra_minus = prediction_dra_minus_based[0][0] | |
| team_1_predicted_pct_dra_minus = predicted_win_prob_dra_minus | |
| pred_minus_implied_warp = predicted_win_prob_warp - implied_win_prob | |
| pred_over_implied_05_warp = (1 if pred_minus_implied_warp > .05 else 0) | |
| pred_over_implied_10_warp = (1 if pred_minus_implied_warp > .1 else 0) | |
| result_wl = ('?' if pred_over_implied_05_warp == 1 else '') | |
| stake = (100 if money_line > 0 else abs(money_line)) * pred_over_implied_05_warp | |
| pred_minus_implied_dram = predicted_win_prob_dra_minus - implied_win_prob | |
| pred_over_implied_dram = (1 if pred_minus_implied_dram > 0 else 0) | |
| pred_over_implied_05_dram = (1 if pred_minus_implied_dram > .05 else 0) | |
| pred_over_implied_10_dram = (1 if pred_minus_implied_dram > .1 else 0) | |
| result_wl_dram = ('?' if pred_over_implied_dram == 1 else '') | |
| stake_dram = (100 if money_line > 0 else abs(money_line)) * pred_over_implied_dram | |
| daily_df.loc[location_index] = [daily_game_id, team, h1v0, opp, team_starter, opp_starter, | |
| money_line, implied_win_prob, | |
| team_win_pct, opp_win_pct, team_starter_war, opp_starter_war, win_pct_diff, | |
| starter_war_diff, unique_row_id, predicted_win_prob_warp, pred_minus_implied_warp, | |
| pred_over_implied_05_warp, pred_over_implied_10_warp, result_wl, stake, | |
| team_starter_dra_minus, opp_starter_dra_minus, starter_dra_minus_diff, | |
| team_1_predicted_pct_dra_minus, pred_minus_implied_dram, | |
| pred_over_implied_dram, | |
| pred_over_implied_05_dram, pred_over_implied_10_dram, result_wl_dram, | |
| stake_dram] | |
| #print(daily_game_id, team, h1v0, opp, team_starter, opp_starter, money_line, implied_win_prob, | |
| # team_win_pct, opp_win_pct, team_starter_war, opp_starter_war, win_pct_diff, | |
| # starter_war_diff, unique_row_id, predicted_win_prob, pred_minus_implied, | |
| # pred_over_implied_05, pred_over_implied_10, result_wl, stake) | |
| if pred_over_implied_05_warp == 1: | |
| if pred_over_implied_10_warp == 1: | |
| over_10 = 1 | |
| else: | |
| over_10 = 0 | |
| if not (team_starter_war == -999 or opp_starter_war == -999): | |
| pix_df_warp.loc[pix_index] = [team, team_starter, opp, opp_starter, money_line, 1, over_10] | |
| pix_index = pix_index + 1 | |
| if pred_over_implied_dram == 1: | |
| if pred_over_implied_05_dram == 1: | |
| over_05_dram = 1 | |
| else: | |
| over_05_dram = 0 | |
| if not (team_starter_dra_minus == 888 or opp_starter_dra_minus == 888): | |
| pix_df_dram.loc[pix_index] = [team, team_starter, opp, opp_starter, money_line, 1, over_05_dram] | |
| pix_index = pix_index + 1 | |
| location_index = location_index + 1 | |
| unique_row_id = unique_row_id + 1 | |
| # then for HOME team | |
| team = game.home_team | |
| h1v0 = 1 | |
| opp = game.away_team | |
| #API is sending back home starter as visitor, and visitor starter as home... | |
| try: | |
| team_starter = game.p_pitcher_away | |
| opp_starter = game.p_pitcher_home | |
| #print(team_starter) | |
| team_starter_war = get_starter_war(pecota_df, team_starter) | |
| opp_starter_war = get_starter_war(pecota_df, opp_starter) | |
| team_starter_dra_minus = get_starter_dra_minus(pecota_df, team_starter) | |
| opp_starter_dra_minus = get_starter_dra_minus(pecota_df, opp_starter) | |
| except AttributeError: | |
| team_starter = "NONE" | |
| opp_starter = "NONE" | |
| team_starter_war = -888 | |
| opp_starter_war = -888 | |
| team_starter_dra_minus = 888 | |
| opp_starter_dra_minus = 888 | |
| money_line = get_money_line(odds_df, team) | |
| implied_win_prob = bu.money_line_to_win_prob(money_line) | |
| team_win_pct = get_win_pct_nickname(teams, injury_df, team, 'PECOTA Win Pct 7/5') | |
| opp_win_pct = get_win_pct_nickname(teams, injury_df, opp, 'PECOTA Win Pct 7/5') | |
| win_pct_diff = team_win_pct - opp_win_pct | |
| starter_war_diff = team_starter_war - opp_starter_war | |
| starter_dra_minus_diff = team_starter_dra_minus - opp_starter_dra_minus | |
| prediction_warp_based = ann_warp.predict([[h1v0, starter_war_diff, win_pct_diff]]) | |
| predicted_win_prob_warp = prediction_warp_based[0][0] | |
| team_2_predicted_pct = predicted_win_prob_warp | |
| key_vars = [h1v0, win_pct_diff, starter_dra_minus_diff ] | |
| key_vars = dra_minus_scaler.transform([key_vars]) | |
| prediction_dra_minus_based = ann_dra_minus.predict([key_vars]) | |
| predicted_win_prob_dra_minus = prediction_dra_minus_based[0][0] | |
| team_2_predicted_pct_dra = team_2_predicted_pct | |
| pred_minus_implied_warp = predicted_win_prob_warp - implied_win_prob | |
| pred_over_implied_05_warp = (1 if pred_minus_implied_warp > .05 else 0) | |
| pred_over_implied_10_warp = (1 if pred_minus_implied_warp > .1 else 0) | |
| result_wl = ('?' if pred_over_implied_05_warp == 1 else '') | |
| stake = (100 if money_line > 0 else abs(money_line)) * pred_over_implied_05_warp | |
| pred_minus_implied_dram = predicted_win_prob_dra_minus - implied_win_prob | |
| pred_over_implied_dram = (1 if pred_minus_implied_dram > 0 else 0) | |
| pred_over_implied_05_dram = (1 if pred_minus_implied_dram > .05 else 0) | |
| pred_over_implied_10_dram = (1 if pred_minus_implied_dram > .1 else 0) | |
| result_wl_dram = ('?' if pred_over_implied_dram == 1 else '') | |
| stake_dram = (100 if money_line > 0 else abs(money_line)) * pred_over_implied_dram | |
| #let's see if adjusting the win percentages such that the total of the two equals 100% changes anything | |
| #team_total_win_pct = team_1_predicted_pct + team_2_predicted_pct | |
| #team_1_adjusted_win_pct = team_1_predicted_pct / team_total_win_pct | |
| #team_2_adjusted_win_pct = team_2_predicted_pct / team_total_win_pct | |
| # moved to the excel spreadsheet becasuse it was easer to do faster | |
| #daily_df.loc[location_index] = [daily_game_id, team, h1v0, opp, team_starter, opp_starter, money_line, implied_win_prob, | |
| # team_win_pct, opp_win_pct, team_starter_war, opp_starter_war, win_pct_diff, | |
| # starter_war_diff, unique_row_id, predicted_win_prob, pred_minus_implied, | |
| # pred_over_implied_05, pred_over_implied_10, result_wl, stake, ] | |
| daily_df.loc[location_index] = [daily_game_id, team, h1v0, opp, team_starter, opp_starter, | |
| money_line, implied_win_prob, | |
| team_win_pct, opp_win_pct, team_starter_war, opp_starter_war, win_pct_diff, | |
| starter_war_diff, unique_row_id, predicted_win_prob_warp, pred_minus_implied_warp, | |
| pred_over_implied_05_warp, pred_over_implied_10_warp, result_wl, stake, | |
| team_starter_dra_minus, opp_starter_dra_minus, starter_dra_minus_diff, | |
| team_2_predicted_pct_dra, pred_minus_implied_dram, pred_over_implied_dram, | |
| pred_over_implied_05_dram, pred_over_implied_10_dram, result_wl_dram, | |
| stake_dram ] | |
| #print(daily_game_id, team, h1v0, opp, team_starter, opp_starter, money_line, implied_win_prob, | |
| # team_win_pct, opp_win_pct, team_starter_war, opp_starter_war, win_pct_diff, | |
| # starter_war_diff, unique_row_id, predicted_win_prob, pred_minus_implied, | |
| # pred_over_implied_05, pred_over_implied_10, result_wl, stake) | |
| if pred_over_implied_05_warp == 1: | |
| if pred_over_implied_10_warp == 1: | |
| over_10 = 1 | |
| else: | |
| over_10 = 0 | |
| if not (team_starter_war == -999 or opp_starter_war == -999): | |
| pix_df_warp.loc[pix_index] = [team, team_starter, opp, opp_starter, money_line, 1, over_10] | |
| pix_index = pix_index + 1 | |
| if pred_over_implied_dram == 1: | |
| if pred_over_implied_05_dram == 1: | |
| over_05_dram = 1 | |
| else: | |
| over_05_dram = 0 | |
| if not (team_starter_dra_minus == 888 or opp_starter_dra_minus == 888): | |
| pix_df_dram.loc[pix_index] = [team, team_starter, opp, opp_starter, money_line, 1, over_05_dram] | |
| pix_index = pix_index + 1 | |
| location_index = location_index + 1 | |
| unique_row_id = unique_row_id + 1 | |
| daily_game_id = daily_game_id + 1 | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment