def one_match_simulator(
home_goalkeeper,
home_defenders,
home_midfielders,
home_forwards,
away_goalkeeper,
away_defenders,
away_midfielders,
away_forwards,
model_name='/Users/bgrantham/Documents/Personal/fifa-ratings-predictor/fifa_ratings_predictor'
'/deep-models'
'-all/deep'):
home_defenders = home_defenders + [0] * (6 - len(home_defenders))
away_defenders = away_defenders + [0] * (6 - len(away_defenders))
home_midfielders = home_midfielders + [0] * (7 - len(home_midfielders))
away_midfielders = away_midfielders + [0] * (7 - len(away_midfielders))
home_forwards = home_forwards + [0] * (4 - len(home_forwards))
away_forwards = away_forwards + [0] * (4 - len(away_forwards))
home_feature_vector = home_goalkeeper + home_defenders + home_midfielders + home_forwards
away_feature_vector = away_goalkeeper + away_defenders + away_midfielders + away_forwards
feature_vector = normalise_features(
np.array(home_feature_vector + away_feature_vector)).reshape(1, 36)
net = NeuralNet()
probability = net.predict(feature_vector, model_name)
return probability[0]
def simulate_monte_carlo(self,
number_of_simulations,
verbose=False,
normalise=True):
for k, v in self.predicted_lineups.items():
self.predicted_lineups[k] = normalise_features(v)
probabilities = self.get_match_probabilities(self.fixtures,
verbose=verbose)
results = self.get_match_results_from_probabilities(
probabilities, number_of_simulations)
for i in tqdm(range(number_of_simulations),
desc='running_simulations',
disable=not verbose):
self.run_season(self.fixtures, [x[i] for x in results])
if normalise:
self.normalise_season_values(number_of_simulations)
df = self.convert_to_pandas(write_to_csv=self.write_to_csv)
return df
def one_match_simulator(home_goalkeeper, home_defenders, home_midfielders, home_forwards, away_goalkeeper,
away_defenders, away_midfielders, away_forwards,
model_name='/Users/offera/Desktop/Personal Projects/betPredictor/pl-predictions-using-fifa/fifa_ratings_predictor'
'/models/E0'
'/deep'):
home_defenders = home_defenders + [0] * (6 - len(home_defenders))
away_defenders = away_defenders + [0] * (6 - len(away_defenders))
home_midfielders = home_midfielders + [0] * (7 - len(home_midfielders))
away_midfielders = away_midfielders + [0] * (7 - len(away_midfielders))
home_forwards = home_forwards + [0] * (4 - len(home_forwards))
away_forwards = away_forwards + [0] * (4 - len(away_forwards))
home_feature_vector = home_goalkeeper + home_defenders + home_midfielders + home_forwards
away_feature_vector = away_goalkeeper + away_defenders + away_midfielders + away_forwards
feature_vector = normalise_features(np.array(home_feature_vector + away_feature_vector)).reshape(1, 36)
net = NeuralNet()
probability = net.predict(feature_vector, model_name)
return probability[0]
tf.set_random_seed(8)
np.random.seed(8)
league = 'F1'
inputs = np.vstack(
(np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/feature-vectors-13-14.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/feature-vectors-14-15.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/feature-vectors-15-16.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/feature-vectors-16-17.npy')))
inputs = normalise_features(inputs)
outputs = np.vstack(
(np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/targets-13-14.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/targets-14-15.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/targets-15-16.npy'),
np.load('./data/lineup-data/' + league +
'/processed-numpy-arrays/targets-16-17.npy'))).reshape(-1, 3)
nan_rows = np.where(outputs != outputs)[0]
inputs = np.delete(inputs, nan_rows, axis=0)
outputs = np.delete(outputs, nan_rows, axis=0)
def main():
bet_tracker = BetTracker()
league = 'E0'
match_data = read_match_data(season='2016-2017', league=league)
match_data = assign_odds_to_match(match_data, read_all_football_data(league=league))
player_data = read_player_data(season='2016-2017')
net = NeuralNet()
bank = [100]
all_odds = []
errors = []
cached_players = {}
feature_vectors = []
for match in match_data:
try:
home_players_matched, cached_players = match_lineups_to_fifa_players(match['info']['home lineup names'],
match['info']['home lineup raw names'],
match['info']['home lineup numbers'],
match['info'][
'home lineup nationalities'],
constants.LINEUP_TO_PLAYER_TEAM_MAPPINGS[
'ALL'][
match['info']['home team']],
match['info']['season'],
player_data, cached_players)
away_players_matched, cached_players = match_lineups_to_fifa_players(match['info']['away lineup names'],
match['info']['away lineup raw names'],
match['info']['away lineup numbers'],
match['info'][
'away lineup nationalities'],
constants.LINEUP_TO_PLAYER_TEAM_MAPPINGS[
'ALL'][
match['info']['away team']],
match['info']['season'],
player_data, cached_players)
home_feature_vector = create_feature_vector_from_players(home_players_matched)
away_feature_vector = create_feature_vector_from_players(away_players_matched)
feature_vector = np.array(home_feature_vector + away_feature_vector).reshape(-1, 36)
feature_vectors.append(normalise_features(feature_vector))
except Exception as exception:
print(match['info']['date'], match['info']['home team'], match['info']['away team'])
print(exception)
errors.append(match['match number'])
feature_vectors = np.vstack((x for x in feature_vectors))
probabilities = net.predict(feature_vectors, model_name='./models/' + league + '/deep')
match_data = [match for match in match_data if match['match number'] not in errors]
for match, probability in zip(match_data, probabilities):
# print(match['info']['date'], match['info']['home team'], match['info']['away team'])
pred_home_odds, pred_draw_odds, pred_away_odds = [1 / x for x in probability]
home_odds, draw_odds, away_odds = match['info']['home odds'], match['info']['draw odds'], match['info'][
'away odds']
all_odds.append((pred_home_odds, home_odds))
all_odds.append((pred_away_odds, away_odds))
if pred_home_odds < home_odds < 3.2 and 0.02 <= probability[0] - 1 / home_odds:
stake = calculate_stake(home_odds, probability=1 / pred_home_odds, method='kelly',
constant_profit=20) * bet_tracker.bankroll
profit = stake * home_odds - stake
bet = Bet(true_odds=home_odds, predicted_odds=pred_home_odds, stake=stake, profit=profit, match=match,
type='home')
bet_tracker.make_bet(bet)
if match['info']['home goals'] > match['info']['away goals']:
bet_tracker.bet_won()
else:
bet_tracker.bet_lost()
bank.append(bet_tracker.bankroll)
elif pred_away_odds < away_odds < 3.2 and 0.02 <= probability[2] - 1 / away_odds:
stake = calculate_stake(away_odds, probability=1 / pred_away_odds, method='kelly',
constant_profit=20) * bet_tracker.bankroll
profit = stake * away_odds - stake
bet = Bet(true_odds=away_odds, predicted_odds=pred_away_odds, stake=stake, profit=profit, match=match,
type='away')
bet_tracker.make_bet(bet)
if match['info']['home goals'] < match['info']['away goals']:
bet_tracker.bet_won()
else:
bet_tracker.bet_lost()
bank.append(bet_tracker.bankroll)
return bet_tracker, bank, all_odds