A typical middle needed to land about 14% of the time to break even.
01 · How it works
Two lines, one gap
One book posts a total at 8.5, another at 10.5. Bet the over 8.5 at the first and the under 10.5 at the second. A final of 9 or 10 wins both bets. Anything else wins one and loses the other. The width of the gap is the number of results that win both: here, two.
Worked example · Over 8.5 at +100, under 10.5 at −120, $1,000 in total
- Stake on the over
- $478.26
- Stake on the under
- $521.74
- Result outside the gap
- −$43.48
- Result lands on 9 or 10
- +$913.04
Split so either bet alone returns the same $956.52, a miss costs $43.48 and a landing pays $913.04. So the middle is worth taking only if the chance of a 9 or a 10 is above 4.5%.
02 · What we logged
What 308,396 middles cost and paid
Our scanner first saw 308,396 middles between 2026-09-30 and 2026-10-27 that have since ended. 20,373 of them (6.6%) cost nothing on a miss: an arbitrage with a gap on top, rare and short-lived. The other 288,023 had a real cost. For those, the median cost on a miss was −$122.00 per $1,000 and the median payout on a landing was $756.10.
Typical gap
3.5
Cost on a miss
12.20%
Needs to land
13.9%
| Market | Gap | Estimate | Needs |
|---|---|---|---|
| NBA game totals | 6 | 14.3% | 9.1% |
| NBA player points | 3 | 16.9% | 10.1% |
| NFL game totals | 7 | 23.0% | 20.9% |
| College football totals | 6 | 16.0% | 13.9% |
| NBA player rebounds | 2 | 27.2% | 21.5% |
| NFL receptions | 1 | 28.3% | 16.5% |
| MLB total bases | 1 | 27.1% | 20.9% |
Two things stand out. Low-scoring markets such as MLB total bases and NFL receptions have gaps of one, but one result covers a big share of outcomes, so the chance is high and so is the cost. NBA totals have wide gaps, often six points, because live and alternate lines drift far apart, and the cost of a miss is small. The gap alone tells you little. The chance of landing against the cost is the whole bet.
03 · Reading it honestly
What these numbers do not tell you
- The chance column is our estimate. A middle is only as good as that estimate, and estimates are least reliable on rare scores.
- We have not yet graded how many of these middles actually landed. This study compares prices, not results.
- Most middles lose a little most of the time. You need a lot of them, or a large edge per middle, before results look anything like the averages.
- A middle disappears about as fast as an arb does. The median middle lasted 1m 41s in the same log.
Before you bet
What this means for your next bet
- Work out the break-even chance before you place a middle: cost on a miss divided by cost plus payout.
- Compare it with an honest chance of landing in the gap. If you cannot estimate it, pass.
- Prefer gaps in markets where results bunch (key numbers in football, low counts in baseball).
- Expect to lose small amounts most of the time. Size so a long run of misses does not hurt.
Method
How we measured this
Every middle the scanner first saw between the two dates and later saw end, admitted by the live desk's own rule (no lock above 8%, no leg at a book that stopped taking bets). The cost on a miss and the payout on a landing are per $1,000 of total stake, split so either bet alone returns the same. Break-even chance is each middle's cost divided by its cost plus its payout; the figures are medians across middles. The chance estimate comes from de-vigged alternate lines where at least two books post both lines, and from a measured score model otherwise.
- Sample
- 308,396 middles first seen in the window that have since ended
- Window
- Sep 30, 2026 to Oct 27, 2026
- Measured
- Oct 29, 2026
- Source
- Dynatyze arbitrage ledger
Questions3 questions
What is a middle in sports betting?
A middle is a pair of bets on opposite sides of two different lines, such as over 8.5 at one book and under 10.5 at another. If the result lands between the lines, both bets win; otherwise one wins and one loses.
Is middling profitable?
It can be, when the chance of landing in the gap is higher than the break-even chance. Across 288,023 middles we logged with a cost on a miss, the median break-even chance was 13.9%. Results vary, and most individual middles lose a small amount.
What is the difference between a middle and an arbitrage?
An arbitrage locks a return whatever happens. A middle accepts a small loss on most results in exchange for a large win on a few. Occasionally a pair is both: it locks a small return and also pays big if the result lands in the gap.
Responsible gaming
Dynatyze is analytics and education for people 21 and over. It is not a sportsbook and never takes a bet. Every number on this page describes the past or the arithmetic of a price; none of it promises a result, and results vary.
Set a budget before you bet and keep to it. If gambling stops being fun, help is free and confidential at 1-800-GAMBLER, and our responsible gaming page lists more ways to get support.
Continue reading
Research and education only. Results vary, and nothing here promises a profit. Every number from our own data names its sample and its dates; every worked example is the arithmetic of the prices shown.