Rating starts with the ordinary settlement
For a balanced back/lay match using your own money (not a free bet), a useful rating convention is 100% plus the ordinary net profit as a percentage of the back stake. A £100 back stake with a £1 qualifying loss has a 99% rating; a £3 loss gives 97%. A rating of 100% represents a break-even ordinary match before any early-payout benefit.
Rating (%) = 100 × (1 + ordinary net profit ÷ back stake)
Use an unrounded balanced lay for the theoretical rating. Penny rounding can create slightly different actual ordinary results; check both in the calculator. Commission also matters: a pair that looks close at 0% can cost more at your actual rate. A 99% rating is not a 99% chance of winning and does not mean a 99% expected return from the promotion.
EV needs an estimated probability as well
Expected value is a probability-weighted average of the possible net outcomes. The ordinary settlement numbers come from stakes, odds and fees. The chance of an eligible full turnaround is uncertain and needs an estimate. The worked example shows exactly which outcome counts: the backed team goes two goals ahead and then fails to win.
In an idealised balanced match with ordinary net outcome Q, full-turnaround net outcome T, and pre-match full-turnaround probability p:
EV = (1 − p) × Q + p × T
= Q + p × (T − Q)
This is a standard formula, not MatchedBet.app’s own probability model. It assumes every non-turnaround outcome pays Q and all bets settle as intended. A complete assessment must account for any other material outcomes and costs. With rounded stakes, ordinary win and non-win results can differ slightly.
Why the higher rating can have lower EV
Here are two invented teaching scenarios with a £100 back stake. The probabilities are assumptions chosen to show the distinction; they are not historical rates, product estimates or forecasts for any team.
| Scenario | Rating | Ordinary net Q | Turnaround net T | Assumed p | EV per bet |
|---|---|---|---|---|---|
| A | 99% | −£1 | +£199 | 1% | +£1 |
| B | 97% | −£3 | +£197 | 3% | +£3 |
For A, EV is −£1 + 0.01 × £200 = +£1. For B, it is −£3 + 0.03 × £200 = +£3. A has the better rating and loses less on an ordinary settlement. B has the higher EV only if those probability assumptions are justified. Reduce B’s assumed probability to 1% and its EV becomes −£1.
This is why sorting by rating and sorting by estimated EV can produce different orders. A cheap match can still have too little promotional upside. A more expensive match can look better in expectation while producing more costly losing runs.
Use the right probability
The chance of going two ahead is not the same as the chance of going two ahead and failing to win. Nor is the chance of failing to win conditional on having led by two the same as the pre-match chance of the whole sequence. Mixing these quantities can greatly overstate the promotion’s value.
An EV estimate depends on data coverage, assumptions and how well a model represents the particular fixture. It can be wrong even when the back/lay arithmetic is correct. Different estimates can rank the same price pairs differently.
Compare percentages, pounds and capital consistently
When reading EV as a percentage, check its denominator. In the illustration, EV divided by the £100 back stake gives 1% for A and 3% for B. That is not the same as return on the total money committed, because the exchange liability also needs funding. A larger pound EV obtained by increasing the stake also increases exposure; it is not evidence of a better underlying opportunity.
The 2UP Matcher lets you inspect rating alongside EV at your chosen commission. Test how rating and turnaround frequency interact in our 2UP variance simulator, and read why a positive-EV bet can still lose.
Compare back and lay pairs at your exact commission across 40+ UK bookmakers. Start with 7 days free, then £10/mo.
18+ only. Gambling involves risk. Expected value is a long-term estimate, not a guarantee of profit. Prices can move, bets can go unmatched, and promotion eligibility can change.