Middle bet calculator: stakes, outcome table and expected value

The middle bet calculator

Enter both lines and the decimal odds from the bet slips, the total stake and your estimate of the hit rate; a comma or a full stop both work as the decimal mark. The example values are the Over 2.5 / Under 3.5 middle worked through below.

The inputs

A middle is two bets on opposite sides of two different lines, for example Over 2.5 goals at one bookmaker and Under 3.5 goals at another. If the match ends with exactly three goals, both bets win. The theory, including why middles appear at all, is in the middles guide. To calculate one you need:

• Line and odds of leg A, the lower line (here Over 2.5).
• Line and odds of leg B, the higher line (here Under 3.5).
• Total stake across both legs.
• Your estimate of the hit rate: how often the result lands in the window between the lines.

The stake split and the outcome table

The usual way to split a middle is the same as a sure bet: stakes in proportion to 1/odds, so that a miss on either side costs the same.

S = 1/odds A + 1/odds B
stake A = total × (1/odds A) / S
stake B = total − stake A

Each winning leg then returns about total / S. That gives three outcomes:

below the window: stake B × odds B − total
above the window: stake A × odds A − total
in the window (both win): stake A × odds A + stake B × odds B − total

If S is above 1, a miss costs money: a negative middle. If S is below 1, a miss still leaves a small profit: an arb-middle, a sure bet with an extra chance of a large win. Most middles you will find are negative.

Worked example: Over 2.5 and Under 3.5

The break-even hit rate

The question that decides everything: how often must the window hit for the middle to break even?

break-even hit rate = miss loss / (hit profit + miss loss)

For the example: 28.45 / (143.10 + 28.45) = 28.45 / 171.55 ≈ 16.6%. If the match ends with exactly three goals more often than about one time in six, the middle has positive expected value; if less often, it doesn't.

Estimating the hit rate

The hit rate is an estimate, never a known number. Three ways to get one, from rough to better:

1. Historical frequency. How often did matches in this league end with exactly three goals? Easy to look up, but it ignores that this match may be expected to be high- or low-scoring.
2. A goal model. A simple Poisson model with the expected number of goals implied by the market. In football matches priced around 2.5 to 3 expected goals, such a model typically puts exactly three goals at roughly 20 to 23%. A Poisson model is a simplification; real goal distributions deviate from it. The calculator above uses this model when you choose "Goal model" and enter the expected goals.
3. Market prices themselves. If a fair "exactly 3 goals" price or Asian lines around 3.0 are available, they contain the market's own estimate after the margin is removed.

The window width matters more than anything else. One whole goal in football is a big window, because low-scoring sports concentrate results on a few numbers. In basketball a one- or two-point window lands only a few percent of the time. In American football, windows that include the key margins of 3 and 7 points hit noticeably more often than others.

Expected value of the middle

EV = hit rate × hit profit − (1 − hit rate) × miss loss

For the example, at three different hit-rate estimates:

These are illustrations, not forecasts. The table shows how sensitive the result is to the estimate: a few percentage points turn a middle from worth placing into one to skip. EV is also an average over many bets. Any single middle ends either at −€28.45 or at +€143.10, and a run of ten misses in a row is entirely normal at these hit rates. The expected value guide explains the spread in detail.

Whole and quarter lines

With whole and quarter lines the outcome table needs a fourth column, because part of a stake can be returned (a push).

• Whole line. Over 2.5 at 2.10 and Under 3.0 at 1.65, €200: stakes €88.00 and €112.00, each winning leg returns €184.80. A miss costs €15.20. At exactly three goals the Over wins and the Under 3.0 is a push, so its €112.00 comes back: +€96.80. Break-even: 15.20 / 112.00 ≈ 13.6%.
• Quarter line. Under 3.25 is half on Under 3.0 and half on Under 3.5. Over 2.5 at 2.10 and Under 3.25 at 1.55, €200: stakes €84.93 and €115.07. A miss costs about €21.65. At exactly three goals the Over returns €178.35, half the Under stake is pushed (€57.54 back) and the other half wins at 1.55 (€89.18): +€125.07. Break-even ≈ 14.8%.

For the 2.5/3.5 window, the "middle" is a single number of goals. With whole and quarter lines you trade a smaller hit profit for a smaller miss loss. Which is better depends on the prices; the calculation above is the only honest way to compare them. Quarter lines are explained in the Asian handicap guide.

Checking a middle in five steps

1. Confirm both legs: the same match, the same market type and the same settlement period (see overtime betting rules), opposite sides of two different lines.
2. Split the stakes with the formula above and write down every outcome, including pushes.
3. Work out the break-even hit rate.
4. Compare it with a realistic hit-rate estimate. If your estimate is not clearly above break-even, skip it.
5. Place the leg whose price is most likely to move first, then recalculate the second leg with the price you actually got.

Betting fees change the effective odds. If a bookmaker passes on Austria's 5% betting fee, use the odds after the fee in every step; the arbitrage calculator page shows how.

Middles in ArbVantage

ArbVantage lists middles separately from sure bets, with the stakes, every outcome and an estimated hit rate and expected value from a statistical model of how results fall around the market line. They are estimates: judge them over hundreds of middles, not one. You decide which to place.

Frequently asked questions

How likely is a middle to hit?

It depends on the sport and the width of the window. In football, a window of exactly three goals in a match priced around 2.5 to 3 expected goals hits roughly 20 to 23% of the time in a simple goal model. In basketball a one- or two-point window hits only a few percent of the time. Treat every hit rate as an estimate.

Is a negative middle worth placing?

Only if your realistic hit-rate estimate is clearly above the break-even hit rate, which is the miss loss divided by the hit profit plus the miss loss. A negative middle at poor prices has negative expected value even with a good window.

Which sports are best for middles?

Sports with low scores or key numbers: football goal lines, ice hockey goal lines, and American football spreads around 3 and 7 points. High-scoring sports such as basketball offer more lines but each point is a small window, so the prices need to be much better.

Why does the calculator split stakes by 1/odds?

So that a miss costs the same whichever side of the window the result lands on. You can split differently, for example to make one side break even, but then the other side's loss grows.

Sources

• Egidi, Pauli, Torelli: Combining historical data and bookmakers' odds in modelling football scores (arXiv, 2018)
• Štrumbelj (2014): On determining probability forecasts from betting odds, International Journal of Forecasting 30(4)
• Wikipedia: Poisson distribution
• RIS: Gebührengesetz 1957, § 33 TP 17 (Austrian betting fee) (in German)

As of October 2026. Bookmakers' rules, odds and offers change; their current terms and betting rules always apply.

Middles are bets with a real chance of losing on each one: a miss is the normal outcome. Only use money you can do without. Our responsible gambling page lists free help in Austria and Germany.

Information, not betting, legal or tax advice. 18+.