import math import random import itertools from data_loader import teams, team_map def elo_probability(teamA, teamB): teamA = team_map.get(teamA.upper(), teamA.upper()) teamB = team_map.get(teamB.upper(), teamB.upper()) teamA_row = teams[teams['team'] == teamA] teamB_row = teams[teams['team'] == teamB] if teamA_row.empty: return f"Error: Team {teamA} not found" if teamB_row.empty: return f"Error: Team {teamB} not found" eloA = teamA_row['elo_rating'].values[0] eloB = teamB_row['elo_rating'].values[0] probA = 1 / (1 + 10 ** ((eloB - eloA) / 550)) return { "teamA": teamA, "teamB": teamB, "probA": probA, "probB": 1 - probA } def simulate_match(teamA, teamB): probs = elo_probability(teamA, teamB) r = random.random() if r < probs["probA"]: return teamA else: return teamB def monte_carlo_group(group_name, simulations=3000): group_teams = teams[teams["group"] == group_name]["team"].tolist() qualify_counts = {team:0 for team in group_teams} for _ in range(simulations): results = {team:0 for team in group_teams} for a,b in itertools.combinations(group_teams,2): winner = simulate_match(a,b) results[winner] += 2 standings = sorted(results.items(), key=lambda x:x[1], reverse=True) qualify_counts[standings[0][0]] += 1 qualify_counts[standings[1][0]] += 1 output = {} for team in qualify_counts: output[team] = qualify_counts[team]/simulations return output