A complete guide to betting MLB, moneyline, run line, totals, parlays, and F5. How the markets work, where value tends to hide, and how Bookie Bullies' AI-assisted model finds edges across 35+ factors.
This guide walks through the four main ways to bet MLB games — moneyline, run line, totals, and parlays — plus the First 5 Innings (F5) market that's grown rapidly the last few seasons. Each section explains the bet, when it has value, and what the model looks for.
First 5 Innings (F5) bets resolve based on the score after 5 innings only. Bullpen variance is stripped out — only the starting pitchers matter. F5 markets are softer than full-game markets because public attention focuses on full-game outcomes. We have a dedicated F5 picks page updated daily.
Value comes from probability — specifically, when your estimate of a team's win probability is HIGHER than the de-vigged market implied probability. The market sets a price (e.g., -120 = ~52.4% market implied); a +EV bet is one where your probability estimate exceeds that.
Bookie Bullies' AI-assisted model produces probabilities for every side of every game. We blend the model with the market closing line at 50/30/10/10 (model/market/ensemble/close). When the blended probability exceeds the de-vigged market by 4%+, we call it a Lock — our highest conviction tier.
Stake sizing matters as much as picking. We use 1/8-Kelly with variance adjustment via a Bayesian uncertainty band. Real Kelly assumes the probability is exact; ours has noise, so 1/8-Kelly protects ROI against estimation error. Most picks land at 0-3 units (1u = 1% of bankroll). The Bankroll Calculator at the top of the daily card converts units to dollar stakes for your bankroll.
MLB has the longest schedule (162 games) which means rich data per team but also frequent injuries, rotation churn, and lineup changes. Starting pitcher quality dominates run distribution — far more than any single player matters in NBA or NFL. Park factors and weather create wide variance in totals. The model adjusts for all of this; humans tend to undervalue weather and park-by-handedness effects.
Books charge ~5% vig (vigorish) on most lines. -110 means betting $110 to win $100; the difference is the book's cut. Over a season of -110 bets, you need to hit ~52.4% just to break even. Beating the line — measured by Closing Line Value — is the single best predictor of long-term profitability.
Every pick is made by MarcDuck with the assistance of AI software. The AI weighs up to 35 factors per game (FIP, xFIP, bullpen quality, park-by-handedness, weather, lineup splits, line movement, recent form, etc.). MarcDuck reviews edge cases and formulates the final pick. The full methodology is documented at /methodology.html; the public track record is at /track-record.html.