Expected Value in MLB Betting: How to Calculate and Apply EV to Props
Available in US
The Single Number That Defines Whether a Bet Is Worth Taking
Early in my betting career, I measured success by whether I won today’s bets. It took two losing seasons to understand that daily results are noise, and the only metric that matters over time is expected value. EV is the average amount you win or lose per bet if you could place the same wager thousands of times. A positive EV bet does not guarantee a win tonight — it guarantees that the math is on your side over a large enough sample. That distinction changed everything about how I approach MLB props, and it should change how you approach them too.
The average sportsbook win rate across sports betting reached a record 9.7% in 2025. That number represents the aggregate edge the house holds over all bettors, on all bet types, across all sports. Your job as a bettor is not to overcome the house edge on every bet — that is impossible. Your job is to find the specific bets where your estimate of the true probability differs enough from the sportsbook’s implied probability that the expected value turns positive. Those bets exist. They are just not as common as the sportsbook’s marketing department would have you believe.
The EV Formula and How to Run It for Any MLB Prop
I am going to walk through the formula once, concretely, because every guide I have read either makes this too abstract or too complicated. Expected value is calculated as:
EV = (Win Probability x Profit if Win) – (Loss Probability x Amount Lost)
Suppose a sportsbook offers a pitcher’s strikeout over at 6.5, priced at -120. At -120, a £94.8 bet returns £79 in profit. The implied probability the book assigns to the over is roughly 54.5%. Now suppose your model — based on the pitcher’s recent form, the opposing lineup’s strikeout rate, the umpire’s zone, and the park — estimates the true probability of the over at 60%.
Plug in: EV = (0.60 x £79) – (0.40 x £94.8) = £47.4 – £37.9 = +£9.48.
That £9.48 is your expected profit per £94.8 wagered, or about 10% return on stake. In betting terms, that is a strong positive-EV opportunity. You will not win every time you take this bet — 40% of the time the pitcher falls short of 7 strikeouts and you lose £94.8. But over fifty or a hundred repetitions of bets with similar profiles, the positive EV compounds into real profit.
The difficult part is not the formula. It is the input: your estimate of the true probability. If your 60% estimate is wrong — if the real probability is 53% — the EV flips negative, and you are the one funding the sportsbook’s record win rate. Every EV bet is only as good as the model behind it, which is why I spend far more time refining my probability estimates than I do shopping for lines.
Applied Example: Evaluating a Strikeout Over Line
Let me take a real-world scenario I evaluated last season. A starting pitcher was lined at 5.5 strikeouts with the over priced at -115. His season strikeout rate was 27%, his recent five-start average was 29%, and the opposing lineup struck out at a 25% clip as a team. The game was at a neutral park with a home-plate umpire who had a wider-than-average zone — a factor that tends to increase called strikes and, by extension, strikeout opportunities.
I estimated the pitcher would face roughly 24 batters. At a 28% strikeout rate — blending his season rate, recent form, and opponent tendency — that projects to 6.7 strikeouts. The probability of recording 6 or more strikeouts, using a Poisson distribution centered on 6.7, came out to approximately 62%. The sportsbook’s -115 line implied about 53.5%. The gap — 62% versus 53.5% — translated to a positive EV of roughly 8 pence per pound wagered.
The bet hit. The pitcher recorded 8 strikeouts. But I want to be clear: the outcome was irrelevant to the process. The EV was positive before the first pitch was thrown. If the pitcher had recorded 5 strikeouts and the bet had lost, the analysis would still have been correct. One of the hardest things in betting is separating the quality of a decision from the quality of a result. EV thinking forces you to make that separation every single time.
Sportsbook Hold Rate and What It Costs the Average Bettor
That 9.7% average hold rate is not evenly distributed across all markets. Moneylines on high-profile games carry slimmer margins — often 3% to 4% in implied probability. Player props, particularly on less popular games, can carry margins of 6% to 10% or more. The more niche the market, the wider the sportsbook’s edge, because fewer sharp bettors are competing to push the line to fair value.
This matters enormously for EV analysis. A prop with a 7% built-in margin requires your probability estimate to be at least 7 percentage points better than the book’s implied probability just to break even. On a moneyline with a 3% margin, the bar is half as high. That is why I always start my evaluation by calculating the implied margin on the line I am considering. If the margin is too wide, the bet needs to clear a higher hurdle, and my confidence in my estimate needs to be correspondingly stronger.
The 60/1 problem compounds this. Research on NFL bettors showed that 60% of bettors generate only 1% of sportsbook revenue — they bet small, lose small, and churn. The remaining 40% generate the vast majority of the house’s income, often through high-frequency prop betting and parlays where margins are widest. If you are betting props without calculating EV, you are statistically likely to be in that 60% — not because you are unlucky, but because the margin silently erodes your bankroll one bet at a time.
The antidote is straightforward but demanding: calculate EV on every prop bet before you place it. Develop a model, even a simple one, that generates a probability estimate for each outcome. Compare that estimate to the sportsbook’s implied probability. Only bet when the gap is positive and meaningful — I use a threshold of 5% or more in implied probability, though some sharper bettors set the bar at 3%. Below that threshold, the noise in your model is likely larger than the edge, and you are trading on uncertainty rather than information. Building this discipline ties directly into the broader analytical approach I outline in the player props strategy framework.
What is a positive expected value bet in MLB?
A positive expected value bet is one where your estimated probability of winning exceeds the sportsbook’s implied probability by enough to overcome the built-in margin. If the sportsbook implies a 55% chance of an outcome and your analysis estimates 62%, the difference creates positive EV — meaning the bet is mathematically profitable over a large sample of similar wagers, regardless of whether any individual bet wins or loses.
How does the sportsbook hold rate affect my long-term results?
The hold rate represents the percentage of total wagered money that sportsbooks retain after paying winners. The average UK hold rate reached 9.7% in 2025. For individual bettors, the hold rate manifests as the vig or margin built into every line. Higher-margin markets like player props require more accurate probability estimates to overcome than lower-margin markets like moneylines. Without a systematic edge that exceeds the hold rate, long-term losses are a mathematical certainty.
Elaborado por el equipo de «mlb Players Betting».
