Every profitable bettor I have ever met — every single one — thinks in expected value. Not hunches, not streaks, not “this team feels due.” Expected value, abbreviated EV, is the mathematical answer to the only question that matters: does this bet make me money over time? I ignored EV for my first two years of MLB betting and spent those years roughly breaking even. Once I learned to calculate it, my results changed permanently.
The concept is borrowed from probability theory, but the application is brutally practical. EV tells you how much you can expect to gain or lose on a bet if you placed it an infinite number of times under identical conditions. A positive EV bet (+EV) is one where the expected return exceeds the stake. A negative EV bet (-EV) costs you money in the long run regardless of what happens tonight. The entire discipline of analytical sports betting reduces to one objective: find and place +EV bets repeatedly.
The Expected Value Formula Applied to MLB Markets
The formula itself is simple enough to calculate on your phone between innings. Here it is:
EV = (Probability of Winning x Profit if You Win) – (Probability of Losing x Amount Lost if You Lose)
Suppose you are looking at a moneyline bet on a team priced at 2.20 decimal. You estimate the team’s true probability of winning at 50%. Here is the calculation: Profit if you win on a GBP 10 stake is GBP 12.00 (10 x 2.20 = 22.00 total return minus 10.00 stake). Probability of winning is 0.50. Probability of losing is 0.50. Amount lost if you lose is GBP 10.
EV = (0.50 x 12.00) – (0.50 x 10.00) = 6.00 – 5.00 = +1.00
That +GBP 1.00 means that for every GBP 10 bet placed at these odds when your 50% probability estimate is correct, you expect to profit one pound on average. Across 100 such bets, you would expect roughly GBP 100 in profit. The word “expect” is doing heavy lifting — any individual bet can lose, and short runs of 10 or 20 bets can go either way. EV is a long-run concept, and baseball’s 162-game season provides the volume needed to let long-run expectations materialise.
Historically, moneyline favourites in MLB win approximately 58-62% of the time. But that average conceals massive variation. A -300 favourite might have a true probability above 70%, while a -120 favourite sits closer to 55%. The EV calculation tells you whether the price you are being offered compensates for the actual probability, and that answer changes with every game and every line.
Estimating True Probability: Model-Based vs Consensus-Based
The formula is the easy part. The hard part — and the part that separates profitable bettors from everyone else — is estimating the “true probability” input.
There are two broad approaches. The model-based approach builds a quantitative model using pitcher metrics, team offence, park factors, and other variables to generate a win probability for each team. This is what I use, and it took years to refine. The inputs include FIP, xERA, wOBA, bullpen quality indices, and park-adjusted run projections. The model outputs a probability for each side, and I compare that probability to the bookmaker’s implied probability to find +EV situations.
The consensus-based approach is simpler and accessible to anyone. Instead of building a proprietary model, you aggregate closing lines from multiple sharp sportsbooks and treat the average closing line as the market’s best estimate of true probability. The logic: if five sharp books all close with a team at roughly 1.85, the implied probability (54.1%) is a reasonable proxy for the true probability because these books have been sharpened by professional action. You then compare this consensus-implied probability to the odds at your own bookmaker. If your book offers 1.95 on the same team, the gap between 54.1% and 51.3% (implied by 1.95) suggests +EV.
Neither approach is perfect. Models can be miscalibrated, and consensus lines still contain margin. But both are vastly superior to betting based on team names, records, or gut feelings. I recommend the consensus approach for bettors who are new to EV-based thinking — it requires no model-building and provides a solid baseline for identifying value.
Worked Example: Calculating EV on a Moneyline Bet
Let me walk through a full example from a recent slate. Team A’s starting pitcher has a 3.10 FIP and 28% K rate. Team B’s starter has a 4.40 FIP and 19% K rate. My model assigns Team A a 59% win probability. The bookmaker offers Team A at 1.72 decimal.
Step one: calculate implied probability from the bookmaker’s odds. 1 / 1.72 = 58.1%. Step two: compare to my model. My model says 59%, the bookmaker implies 58.1%. The gap is +0.9 percentage points in my favour. Step three: calculate EV. On a GBP 10 bet, profit if I win is GBP 7.20 (10 x 1.72 = 17.20 – 10.00). EV = (0.59 x 7.20) – (0.41 x 10.00) = 4.248 – 4.10 = +0.148.
That is a marginal +EV of about GBP 0.15 per GBP 10 staked. Is it worth taking? Individually, it is a thin edge. But over 500 bets at similar margins, it compounds to roughly GBP 75 in expected profit. MLB’s volume — up to 15 games per day, six months of season — provides exactly the kind of sample size where thin edges become meaningful returns.
The critical discipline is walking away from bets where the EV is zero or negative. A game might be fascinating to watch, the matchup might be compelling on paper, but if the bookmaker’s line already reflects or exceeds your estimated probability, there is no mathematical reason to bet. Sitting out is the hardest skill in sports betting and the most important one. Every EV calculation is ultimately an exercise in deciding when not to bet, which is why it belongs at the centre of any MLB beat bets framework.