What is a value bet?
By ProbaPredict Data Desk · Last reviewed
A value bet is a bet where the odds on offer are higher than the true probability of the outcome justifies. If you estimate a 50% chance and the price is 2.20, the edge is 0.50 × 2.20 − 1 = +10%. Value depends entirely on how accurate your probability is, and a value bet still loses often.
What does "value" mean in betting?
Value is about price, not likelihood. A heavy favourite can be poor value if the odds are too short, and an outsider can be good value if the odds are long enough. The question is always: are the odds higher than the chance of the outcome justifies? If an outcome happens 50% of the time, any price above 2.00 is value; any price below is not.
How do you calculate the edge?
Edge (expected value per unit staked) = probability × decimal odds − 1. Positive means the price is higher than your estimate justifies; negative means lower. Equivalently, compare your probability with the price's implied probability: if your 50% is higher than the 45.5% implied by 2.20, the price looks generous to you.
Worked example 1: a home win
Our model gives the home side a 52% chance. A bookmaker offers 2.10. Edge = 0.52 × 2.10 − 1 = +9.2%. Implied probability of 2.10 is 47.6%, below our 52%. Over many such bets, if our 52% were exactly right, you would expect to get back 1.092 for every 1 staked. But in any single match the home team fails to win 48% of the time.
Worked example 2: Under 2.5 goals
Our model gives Under 2.5 a 44% chance. The bookmaker offers 2.05. Edge = 0.44 × 2.05 − 1 = −9.8%. The price implies 48.8%, above our 44%, so by our estimate this price is too short. Our value bet calculator runs these numbers for any fixture we cover, or with your own probability.
Why is finding value so hard?
Because the edge is only as good as the probability. Bookmakers' closing odds are among the best forecasts in sport. In our own back-test, margin-free closing odds are slightly more accurate than our model on both the match result and Over/Under 2.5. We publish that comparison rather than hide it:
| Market | Graded picks | Hit rate | Brier score |
|---|---|---|---|
| 1X2 match result | 0 | – | – |
| Over/Under 1.5 goals | 0 | – | – |
| Over/Under 2.5 goals | 0 | – | – |
| Over/Under 3.5 goals | 0 | – | – |
| Both teams to score (BTTS) | 0 | – | – |
| Correct score | 0 | – | – |
| All markets | 0 | – | – |
Source: ProbaPredict graded predictions (0 graded picks), updated Fri, 2 Oct 2026. Each pick is our most likely outcome in that market; Brier score is per pick (0 is perfect).
| Measure (lower is better) | Our model | Closing odds |
|---|---|---|
| 1X2 Brier score | 0.609 | 0.592 |
| 1X2 log loss | 1.016 | 0.991 |
| Over/Under 2.5 Brier score | 0.246 | 0.239 |
Source: ProbaPredict analysis of 18,807 matches, updated Fri, 2 Oct 2026. Back-test, not live predictions: walk-forward out-of-sample predictions with the model refitted weekly on earlier data only. Closing odds made margin-free with the proportional method.
That means a gap between our probability and a closing price is more often a small model error than a genuine mispricing. Prices earlier in the week can differ more from the eventual closing odds, but those differences cut both ways.
How should you read a positive edge?
As information, not a prediction. A +5% edge on a 30% outcome still loses seven times in ten. Results over a handful of bets mostly reflect luck. That's why we never recommend stakes and keep the line "A positive edge means the price is higher than our model's estimate, not that the bet will win" under every calculator result.
UK and US terminology
UK: "value bet", "edge", "overlay". US: "+EV bet" (positive expected value). Same idea.
Common mistakes with value betting
- Treating any model as truth. Your edge estimate has its own error.
- Judging by short-term results. A run of losses doesn't disprove value, and a run of wins doesn't prove it.
- Confusing value with likelihood. Value bets on outsiders lose most of the time.
- Using staking systems to "maximise" edge. We don't recommend any; set a fixed budget instead. See betting budgets and responsible gambling.
Why don't we recommend stakes?
Staking plans, including formulas that size bets by the estimated edge, depend entirely on the probability being right, and small errors in the estimate can lead to large over-betting. We publish probabilities and fair odds as information, and leave decisions about whether and how much to bet to you. Our firm recommendation is a fixed budget you can afford to lose.
What about closing line value?
Some bettors judge themselves by closing line value: whether the price they took was longer than the closing price. Because closing odds are such strong forecasts, consistently beating the close is widely seen as a better sign of skill than short-term profit. It still offers no certainty, and it requires careful record-keeping over hundreds of bets.
We mention it because it explains why we benchmark against closing odds and why we're open that they are slightly more accurate than our model. A value calculation using our probabilities is a comparison with one transparent model, not a verdict on the bet.
Frequently asked questions
- How do I calculate the edge on a bet?
- Edge = probability × decimal odds − 1. At 40% and odds of 2.75: 0.40 × 2.75 − 1 = +10%.
- Does a value bet mean it will win?
- No. A positive edge means the price is higher than the estimated probability justifies, not that the bet will win. A 40% outcome still loses 60% of the time.
- Where does the probability come from?
- From your own estimate or a model. You can use our published probabilities, but check our accuracy record first: in our back-test, margin-free closing odds are slightly more accurate than our model.
- Do you recommend stakes?
- No. We do not recommend stakes or staking systems.
Related pages
- Value bet calculator
- Our accuracy record
- Fair odds explained
- How accurate football predictions are
- What Is an Accumulator (Acca) Bet?
- What Is a Parlay in Soccer Betting?
- All football betting guides
18+. Gambling can be addictive; only bet what you can afford to lose. Free help: BeGambleAware.org · GamStop.