slickbet.model

Betting model for calculating win/loss probabilities.

Implements a weighted scoring system over form, league position, home advantage, H2H, bookmaker odds, goals, venue form, defense, momentum, match statistics, expected goals (xG), and discipline/reliability. Produces BetPrediction objects including draw risk and double-chance probabilities.

Classes

BetOutcome	Possible betting outcomes.
BetPrediction	Represents a betting prediction for a match.
BettingModel	Model for calculating betting probabilities based on match statistics.
DoubleChancePrediction	Represents double chance betting predictions for a match.

Module Contents

class slickbet.model.BetOutcome

Bases: enum.Enum

Possible betting outcomes.

AWAY_OR_DRAW = 'away_or_draw'

AWAY_WIN = 'away_win'

DRAW = 'draw'

HOME_OR_AWAY = 'home_or_away'

HOME_OR_DRAW = 'home_or_draw'

HOME_WIN = 'home_win'

class slickbet.model.BetPrediction

Represents a betting prediction for a match.

confidence: float

defense_score: float = 0.0

double_chance: DoubleChancePrediction | None = None

draw_risk: float = 0.0

property expected_value: float

Calculate expected value assuming fair 2.0 odds.

EV = (probability * potential_win) - (1 - probability) * stake

Returns:

float – Expected value per unit stake at even money.

form_score: float

goal_score: float = 0.0

h2h_score: float

home_advantage_score: float

match: slickbet.api.Match

match_stats_score: float = 0.0

property match_stats_used: bool

Check if match statistics were used in this prediction.

Returns:

bool – True if match statistics contributed to the prediction

momentum_score: float = 0.0

odds_score: float = 0.0

position_score: float

probability: float

reasoning: list[str] = None

recommended_outcome: BetOutcome

reliability_score: float = 0.0

venue_form_score: float = 0.0

xg_score: float = 0.0

class slickbet.model.BettingModel(form_weight: float | None = None, position_weight: float | None = None, home_weight: float | None = None, h2h_weight: float | None = None, odds_weight: float | None = None, goals_weight: float | None = None, venue_form_weight: float | None = None, defense_weight: float | None = None, momentum_weight: float | None = None, match_stats_weight: float | None = None, xg_weight: float | None = None, reliability_weight: float | None = None, min_confidence: float = 0.55)

Model for calculating betting probabilities based on match statistics.

The model uses a weighted scoring system combining multiple factors.

ENHANCED VERSION with 12 factors: - Core: Form, Position, Home Advantage, H2H, Odds - Goals: Attack/defense strength, Venue Form, Defense (clean sheets) - Momentum: First-half performance and consistency - Match Stats: Historical averages (possession, attacks, shots on target) - xG: Expected Goals differential (quality of chances created) - Reliability: Team discipline based on card counts (fewer cards = more reliable)

Weights optimized based on 120-day backtest across 5 major leagues (704 matches): - Bundesliga: 81.3% accuracy (excl. draws) - Premier League: 78.2% accuracy (excl. draws) - Serie A: 74.8% accuracy (excl. draws) - Ligue 1: 70.9% accuracy (excl. draws) - La Liga: 68.9% accuracy (excl. draws)

DRAW_RISK_THRESHOLD = 0.15

HOME_ADVANTAGE = 0.12

MIN_CONFIDENCE = 0.55

WEIGHTS

filter_confident_predictions(predictions: list[BetPrediction], min_confidence: float | None = None) → list[BetPrediction]

Filter predictions by minimum confidence threshold.

Parameters:

• predictions (list[BetPrediction]) – List of predictions to filter

• min_confidence (float or None, optional) – Minimum confidence (defaults to model’s threshold)

Returns:

list[BetPrediction] – Filtered list of predictions

min_confidence = 0.55

predict(match: slickbet.api.Match) → BetPrediction

Generate a betting prediction for a match.

Parameters:

match (Match) – Match object with statistics

Returns:

BetPrediction – BetPrediction with probabilities and recommendation

weights

class slickbet.model.DoubleChancePrediction

Represents double chance betting predictions for a match. Double chance allows betting on two outcomes at once (safer but lower odds).

away_or_draw_prob: float = 0.0

property best_double_chance: tuple[str, float]

Return the best double chance bet and its probability.

Returns:

tuple[str, float] – (label, probability) for the highest double-chance option.

home_or_draw_prob: float = 0.0

no_draw_prob: float = 0.0