Guide

Expected value betting, explained properly

By Adam · Updated 16 July 2026 · 6 min read

Strip away the tipsters, the streaks and the gut feelings, and betting reduces to one question: is this price bigger than the probability deserves? Expected value is that question written as arithmetic. It is also, for most people, the least intuitive idea in gambling, because it asks you to judge decisions by something other than whether they won.

One coin, two prices

Imagine a fair coin, and a bookmaker willing to pay decimal odds of 2.10 on heads. You know the probability precisely: 50%. Fair odds would be 2.00. At 2.10, every £10 staked returns £21 half the time and nothing the other half, which averages out to £10.50 back per £10 in. You lose the toss just as often as anyone else does. You are also, slowly and invisibly, getting richer.

Now price heads at 1.90 instead. Same coin, same 50%, and the average return drops to £9.50 per £10. Whoever takes that bet all evening is quietly funding the operation, whatever their results happen to look like by closing time.

The example is artificial on purpose. Nobody in it is better at predicting coins. The only difference between the winning bettor and the losing one is the price they accepted. Real betting buries this under injury news and form tables, but the structure underneath never changes.

The arithmetic

EV = your probability × decimal odds − 1.

A player you rate 40% to commit two or more fouls, priced at 2.75, works out to 0.40 × 2.75 − 1 = +10%. The same player at 2.30 is −8%, and how much you like him doesn't change it. The formula takes ten seconds to learn. The probability that goes into it is the entire discipline, and it has to be yours: use the bookmaker's own implied probability and the equation just returns their margin with the sign flipped.

A price is an opinion with a fee attached

It helps to know what you're actually looking at. A bookmaker's price bundles together their estimate of the probability, their margin, and adjustments for where the money is going — if one side is taking heavy support, the price moves to manage liability, not because the truth changed. On heavily traded markets, Premier League match odds above all, this process is ruthless and the final number is very sharp indeed.

Player markets get less of everything: less liquidity, less attention, less modelling effort. Lines are generated in bulk and corrected late. That is the honest reason a disciplined model can still find arguments with their prices, and it comes with an equally honest catch: soft markets have low limits, and bookmakers restrict accounts that beat them. The same softness that creates the value also caps how much of it you can take.

Where an edge can come from

Three places, broadly. Information the market doesn't have yet — team news minutes old, a lineup leak. Better processing of information everyone has, which is what a statistical model is: four seasons of shot and foul data, weighed properly, will out-argue a vibe about form. And timing, taking early prices before the market has finished thinking.

Far more things get mistaken for an edge than qualify. Confidence is not an edge. A story about a striker being due is not an edge. A run of five winning weekends is not, on its own, evidence of one; give enough people coin flips and some of them will book five winning weekends. An edge is a specific claim — this probability is higher than this price implies — that survives being checked against a lot of results.

The part that feels wrong

Betting +EV means volunteering for losses. At odds around 3.00, a bet with a genuine 8% edge still loses roughly two times in three. Strings of losses arrive that are long enough to make a sensible person question everything. None of this is the process failing. Variance runs on its own schedule, and it does not care that you did the maths right.

Which is why judging yourself on short-run profit is a trap. Results over a weekend contain almost no information; over a month, barely more. The measure that speaks sooner is whether the prices you took beat the closing price — closing line value, which has its own guide. Profit is what you want. CLV is what tells you whether you've earned the right to expect it.

Frequently asked

Keep going