● METHODOLOGY

How the AI Model Works

Every pick on Bookie Bullies comes from a documented, reproducible AI model. No black box, no tout hype. Here's exactly what goes in, what comes out, and how picks are graded.

Updated 2026-06-07

MLB + NBA coverage

Positive-EV filter

The Model in One Paragraph

For every game on the slate, the model ingests sport-specific inputs, simulates the matchup, and produces a win probability for each side. That probability is compared to the market moneyline. If the model's probability is higher than the market implies, the pick has positive expected value. Only positive-EV picks make the card. Every side also gets a confidence rating and a Kelly-fraction staking suggestion.

MLB Inputs

Starting pitchers: ERA, xFIP, stuff+, pitch mix, recent form, batter-handedness splits.
Bullpen: reliever strength, rest days, high-leverage usage, recent workload.
Offense: team wOBA splits vs. RHP/LHP, recent-form adjustment, lineup order.
Park factors: run-scoring environment, HR-factor, foul-territory adjustments.
Weather: wind direction and speed, temperature, precipitation risk.
Umpire: strike zone tendencies (K% delta vs. league average).
Lineups: live lineup card once posted, fallback to projected lineup.
Travel + rest: back-to-back travel, time zone crossings.

NBA Inputs

Team ratings: offensive and defensive rating, adjusted for recent form.
Pace: projected possessions based on both teams' pace.
Home court: team-specific home court advantage points.
Rest: days of rest, back-to-back status for both teams.
Travel + altitude: miles traveled, time zone crossings, altitude at venue.
Injuries: VORP lost from inactive players (replacement-level model).
Line movement: spread, total, and moneyline deltas since open; steam flag.

WNBA Inputs (Different Algorithm Than NBA)

The WNBA model is its own algorithm, not an NBA copy. Key differences: 40-game seasons require heavier sample-size regression than the NBA's 82-game model, smaller rosters mean star availability swings spreads harder, and pace varies wider across the 12 WNBA teams than across NBA's 30.

Net Rating (regressed): raw point differential per game shrunk toward 0 based on games-played sample size. Catches early-season variance.
Pace factor: estimated from total scoring vs league average. Pace × efficiency = total. Aces play fast, Lynx slow — that 12-possession gap drives totals.
Form trend: last-10 win % vs season win %. Positive = team is heating up vs season baseline.
Per-team HFA: computed from home/road records when available. Falls back to 2.3 league avg (smaller than NBA because WNBA crowds are smaller).
Sample-size shrinkage: at 10 games played, 50% shrinkage. At 30 games, 18%. Prevents 5-0 teams from getting +20 net rating treatment.

NHL Inputs (Different Algorithm Than NBA/WNBA)

Hockey is low-scoring and HFA muted. We use Pythagorean win expectancy to flag luck-driven records and regress lucky teams down to their goal-differential baseline.

Expected goals (regressed): blend of own offense and opponent defense, shrunk toward league avg for small sample size.
Pythagorean win expectancy: goals scored^2.05 / (goals scored^2.05 + goals allowed^2.05). Independent of actual record.
Luck factor: actual win % minus Pythagorean. Positive luck = team will regress down. Adjusts margin by ±0.3 goals.
Form trend: last-10 W% vs season W%. Adjusts margin by ±0.4 goals.
Per-team home-ice: from home/road records when available. Fallback to 0.18 goals league avg (small vs NBA's 3 pts).

Why Each Sport Has a Different Algorithm

Sports betting models that apply the same equation to every sport leave money on the table because the underlying score distributions differ:

MLB uses the Skellam distribution for run margins (integer-run differences) and Negative Binomial for run totals (overdispersed Poisson). Pitcher quality (FIP/xFIP) drives most of the margin signal.
NBA uses Gaussian distribution for point margins (continuous, large-sample). Pace × efficiency drives totals.
WNBA uses Gaussian with heavier sample-size regression than NBA (40-game season vs 82). Wider pace variance forces pace-adjusted totals.
NHL uses Gaussian for goal margins but adds Pythagorean luck regression to flag teams winning more than goal differential predicts. Goalie variance dominates but lacks game-by-game data.

How Probability Becomes a Pick

Raw model output is a margin or total estimate. That gets converted to a win probability via a normal CDF with sport-specific standard deviations. The raw model probability is then blended into a final probability using 50% model output, 30% market-implied probability, 10% Statcast ensemble, and 10% closing-line implied probability. The blend shrinks overconfident predictions and avoids over-betting noisy inputs. The final blended probability is compared to the market moneyline. Expected value is calculated as (prob × decimal odds) minus 1. Only picks with positive EV make the card.

Confidence Tiers

Confidence is independent of expected value. It measures signal strength, not edge size.

Lock: Inputs strongly agree, low variance, multiple corroborating signals.
Strong: Clear lean with one or two light counter-signals.
Lean: Model edge exists but inputs disagree or sample is thin.
Pass: No edge, no pick issued. These don't appear on the card.

Staking Suggestion

Every pick carries a unit-size suggestion based on a fractional-Kelly calculation (quarter-Kelly by default) against the blended probability. The suggestion is informational. Users set their own bankroll rules. Bookie Bullies does not provide money management advice and is not a licensed financial adviser.

Public Grading

Every pick gets logged the moment it's published. Graded outcomes (W / L / P for push) are posted by 8 AM Pacific the next morning. The running record is visible on the archive page. Nothing is hidden, edited, or retroactively removed.

Caveats + Limits

Models are probabilistic. Even positive-EV picks lose, often in streaks. Variance is real and the sample size required to confirm edge is in the thousands of picks, not dozens. Models also miss real-world context: last-minute injury news, umpire substitutions, weather that shifts late. Treat the output as a starting point, not a command. Wager only what you can afford to lose. Gambling problem? Call 1-800-GAMBLER.