/ MLB

The sample-size goldmine of sports betting

162 games per team, 2,430 regular-season games per league, before playoffs. That is more than twice the NFL and NBA combined. Sharpness comes from sample size. This is where our models see the most data and where CLV compounds fastest.

Markets we cover

Moneyline + run line

Run line is the MLB equivalent of the point spread, typically 1.5 runs. We price both sides using starter quality, bullpen state, and ballpark-adjusted run expectancy.

Game total (over/under)

Full-game runs. Weather and wind direction matter more here than in any other sport. We auto-flag wind-out games over 8 mph in hitters' parks.

F5 (first 5 innings)

Starter-dominant market. Removes bullpen variance. Our highest CLV sport-market pairing historically in back-tests.

Player props

HR, hits, total bases, strikeouts (pitcher), RBIs. Handedness splits, batting order, and recent form drive our prop model.

Alt lines

1.5 runs, 2.5 runs, alternate totals. Useful when you need more payout and have an isolated view on variance rather than direction.

Futures

World Series, pennant, division, win totals. Low-liquidity and slow-moving, but sharp markets exist at the edges (especially at the trade deadline).

What actually moves the needle in MLB

Handedness splits

RHP vs LHB is a different matchup than RHP vs RHB. We compute per-batter platoon splits from the last 3 seasons rolling and overweight sample quality when usage has been low.

Ballpark factors

Coors, Fenway, Wrigley play drastically different than Oracle or Petco. We use a 3-year rolling Statcast-adjusted park factor for runs, HRs, and hit probability.

Weather

Wind direction + speed + temperature. A 14 mph wind blowing out at Wrigley raises the game total expectation by roughly 0.7 runs from a still baseline. We ingest weather at lineup-lock.

Bullpen state

Who pitched yesterday, who pitched back-to-back, who is on the IL, who got DFA'd. The closer being unavailable is often the single most valuable thing to know about a tight late-game total.

Umpire strike zone

Umpire calling data goes into our strikeout props and game-total model. Angel Hernandez is not Pat Hoberg; our model knows this.

Lineup lock

Lineups drop ~2 hours pre-game. Our model refreshes at lineup-lock, not just at first pitch. A star player unexpectedly benched moves totals materially.

How to approach MLB this season

MLB volume is high and vig on sides runs tight. This is a quarter-Kelly sport. The temptation to size up on a 58 percent model read is real; the right answer is usually no. Size per your bankroll plan, let sample size do the work.

F5 first-five-inning markets are our favorite structural angle. They remove bullpen variance and reward clean read on starters. If you can only bet one MLB market consistently, make it F5.

Record every bet through the bot via `/bet` so your CLV is computed automatically. 500+ MLB bets across a season is enough to see whether your process works.

Kelly calculator · CLV calculator · No-vig calculator · Glossary · CLV essay