Once you have a win probability from predict_game_teamwins(), you can pass it to evaluate_bet_teamwins() to get a simple recommendation. To find true value in a bet, compare the model’s probability against the implied probability embedded in the sportsbook’s odds.
This tool is for educational purposes only. It is not financial or gambling advice.
The three-tier evaluation system
evaluate_bet_teamwins(probability) in prediction_ai.py classifies a predicted win probability into one of three tiers:
| Return value | Condition | Meaning |
|---|
"Good Bet" | probability > 0.60 | Model is confident — strong edge over a typical line |
"Slight Edge" | probability > 0.52 | Marginal advantage — worth considering only with favorable odds |
"Avoid" | probability <= 0.52 | Insufficient edge to justify the risk |
from prediction_ai import evaluate_bet_teamwins
evaluate_bet_teamwins(0.65) # → "Good Bet"
evaluate_bet_teamwins(0.55) # → "Slight Edge"
evaluate_bet_teamwins(0.48) # → "Avoid"
Expected value in sports betting
A bet has positive expected value (EV) when your estimated probability of winning is higher than the probability implied by the odds. In other words: the bookmaker is underpricing the outcome relative to what your model predicts.
Example: Your model predicts a 65% chance of winning. The sportsbook offers +120 on that outcome.
from betting_math import OddsConverter
model_probability = 0.65
american_odds = 120
implied_probability = OddsConverter.american_to_probability(american_odds)
print(f"Model probability: {model_probability:.1%}") # 65.0%
print(f"Implied probability: {implied_probability:.1%}") # 45.5%
Full workflow
The recommended pattern is: predict → evaluate → verify against odds.
from prediction_ai import predict_game_teamwins, evaluate_bet_teamwins
from betting_math import OddsConverter
# Step 1: Get a win probability from the model
prob = predict_game_teamwins(
points_diff=4.2,
team_reb_roll=44.1,
opponent_reb_roll=41.8,
team_ast_roll=25.3,
opponent_ast_roll=23.0,
home=1
)
print(f"Predicted win probability: {prob:.1%}")
# Step 2: Get a tier recommendation
recommendation = evaluate_bet_teamwins(prob)
print(f"Bet evaluation: {recommendation}")
# Step 3: Compare against the sportsbook's implied probability
american_odds = 120 # The line offered by the sportsbook
implied_prob = OddsConverter.american_to_probability(american_odds)
print(f"Sportsbook implied probability: {implied_prob:.1%}")
# Step 4: Decide
if prob > implied_prob:
print("Model has an edge over the market.")
else:
print("No edge detected — the market has priced this outcome correctly or better.")
Predicted win probability: 65.0%
Bet evaluation: Good Bet
Sportsbook implied probability: 45.5%
Model has an edge over the market.
Always convert the sportsbook’s American odds to implied probability with OddsConverter.american_to_probability() and compare it directly against the model’s output. A "Good Bet" rating alone does not guarantee value — the odds on offer matter just as much as the predicted probability.
Reference
| Function | Module | Description |
|---|
predict_game_teamwins(...) | prediction_ai.py | Returns a float win probability (0–1) |
evaluate_bet_teamwins(probability) | prediction_ai.py | Returns "Good Bet", "Slight Edge", or "Avoid" |
OddsConverter.american_to_probability(american) | betting_math.py | Converts a moneyline to implied probability |
OddsConverter API.