Picks backed
by data.
Algorithmic picks for MLB, NBA, and NHL. Every pick is tracked publicly. No cherry-picked wins — full record, every bet, every result.
| Date | Sport | Pick | Type | Odds | Edge | Conf | Result | Units |
|---|---|---|---|---|---|---|---|---|
| Loading history... | ||||||||
Data collection
Every morning the pipeline automatically pulls game schedules, starting pitcher stats, team batting numbers, recent form, and live odds from DraftKings, FanDuel, BetMGM, and Caesars across MLB, NBA, and NHL.
Model training
Separate logistic regression models for each sport and bet type retrain daily on all historical data. The models learn which statistics actually predict outcomes — pitcher ERA, team OPS, rest days, goalie save percentage, and more.
Edge calculation
The model's predicted win probability is compared against the bookmaker's implied probability. When the model sees a team at 62% and the book prices them at 51%, that's an 11% edge — a statistically positive expected value bet.
Pick selection
Only bets with an edge above the minimum threshold (5% for moneylines, 0.5 runs for totals) are flagged. Stake sizes are calculated using the Kelly Criterion — larger edges get larger stakes, within the $10-$25 range.
Line shopping
For every pick, the algorithm identifies which of the four major sportsbooks offers the best available odds. Even small differences in odds compound significantly over a full season.
Transparent tracking
Every pick is logged before games start. Results are graded automatically at 2am after all games finish. Nothing is hidden or adjusted retroactively — wins and losses are all right here.