Accumulator Betting Explained: What Our Own Daily Accumulators Show
Accumulators turn a few small prices into one big one. That is exactly what makes them dangerous: every leg adds margin, and every mistake in your estimate multiplies. We tracked our own daily accumulators for five months and compared them with the same bets placed as singles.
How an accumulator works
An accumulator (also called a parlay, acca or combo bet) combines several selections into one bet. Every leg has to win. The odds of all legs are multiplied, so the payout grows quickly, and so does the chance that one of them lets you down.
Three favourites at 1.40 each: 1.40 × 1.40 × 1.40 = 2.74. A 10 unit stake returns 27.40, but only if all three win. A void leg counts as 1.00; a half win or half loss on an Asian line counts half.
The probability works the same way. Three legs that each win 80% of the time win together only 0.80 × 0.80 × 0.80 = 51.2% of the time. That is the part the adverts leave out.
The bookmaker margin stacks with every leg
Every leg carries the bookmaker’s margin, and in an accumulator those margins multiply too. Using the 1X2 margins we measured at Pinnacle (5.4%) and Bet365 (7.9%), this is the share of your expected return the bookmaker keeps:
| Accumulator | At 5.4% per leg | At 7.9% per leg | |
|---|---|---|---|
| Single (1 leg) | −5.1% | −7.3% | |
| 2 legs | −10.0% | −14.1% | |
| 3 legs | −14.6% | −20.4% | |
| 4 legs | −19.0% | −26.2% | |
| 5 legs | −23.1% | −31.6% | |
| 6 legs | −27.1% | −36.6% | |
| 8 legs | −34.3% | −45.6% | |
| 10 legs | −40.9% | −53.2% |
Calculation: 1 − (1 / (1 + margin)) to the power of the number of legs. Margins from our odds guide.
A single at 5.4% margin costs you about 5% of your stake in expectation. A 5-leg accumulator at the same margin costs 23%, and a 10-leg one at a typical bookmaker more than half.
Our accumulator test: 887 picks, five months
Every day since May 2026 we have listed the matches where our model rated the selected outcome at 70% or more, at odds of at least 1.30. We combined each day’s top picks, ranked by model probability, into accumulators of two to five legs and settled every one at the odds shown:
| Accumulator | Combos | Won | Avg. odds | Model expected | Return | Same legs as singles |
|---|---|---|---|---|---|---|
| 2 legs | 101 | 56.4% | 1.92 | 74.7% | +9.2% | +1.3% |
| 3 legs | 97 | 39.2% | 2.71 | 63.3% | +2.8% | −1.8% |
| 4 legs | 94 | 31.9% | 3.87 | 53.0% | +3.0% | −0.7% |
| 5 legs | 92 | 20.7% | 5.51 | 43.8% | +0.9% | −0.2% |
Source: our daily high-probability table, 12 May to 12 October 2026, 1 unit per bet. Model expected = average of the product of the model probabilities.
Our model expected 63.3% of the 3-leg combos to win. Only 39.2% did. Every leg was a little too optimistic, and in an accumulator those small errors multiply into a big one.
Overconfidence multiplies
The reason is visible in the single legs. Grouped by the probability our model gave them:
| Model said | Picks | Model average | Fully won | Return as single |
|---|---|---|---|---|
| 70 to 75% | 99 | 72.8% | 63.6% | −2.9% |
| 75 to 80% | 197 | 77.9% | 67.5% | −0.4% |
| 80 to 85% | 341 | 82.4% | 64.8% | −2.8% |
| 85 to 90% | 204 | 87.2% | 72.1% | +3.3% |
| 90 to 100% | 46 | 91.8% | 60.9% | −10.5% |
On average the model gave these picks 81.9%, while 66.7% won outright (another few were refunds or half results). As singles that costs only a little: −1.3% across all 887. But a 15-point error per leg turns 0.82 × 0.82 × 0.82 = 55% into 0.67 × 0.67 × 0.67 = 30%. The same mistake that is harmless on one bet is fatal on three.
Our own tips combined into a daily accumulator
We also checked what would have happened if you had combined our published main tips of each day into one bet:
| Daily accumulator of our main tips | Days | Won | Avg. odds | Return | Same tips as singles |
|---|---|---|---|---|---|
| 3 tips (treble) | 128 | 34.4% | 3.24 | −10.4% | −5.1% |
| 2 tips (double) | 26 | 61.5% | 2.19 | +25.6% | +5.0% |
The trebles won on 34.4% of days and returned −10.4%, twice the loss of the same tips played as singles. The doubles did well, +25.6% over 26 days, but that is a small and lucky sample. Our full record is on the results page.
Combining our tips did not improve them. Over the 128 treble days it roughly doubled the loss compared with betting the same picks separately.
Why some of our accumulators still made money
The high-probability combos in our test returned between +0.9% and +9.2%, even though the same legs as singles were roughly break-even. That is not a hidden edge. With about 100 accumulators per group, a handful of extra wins at odds of 2 to 5 swings the result by ten points or more in either direction.
The honest reading: singles of these picks lost 1.3%, the model was clearly too optimistic, and the margin per leg is certain. Over a long run, accumulators built from them should be expected to lose more than the singles, not less. We will keep tracking them and publish the numbers either way.
How to use accumulators sensibly
- Treat them as entertainment, not a strategy. Every extra leg adds margin and multiplies your errors.
- Keep them short. At 5.4% per leg, a double costs 10% in expectation, a 5-fold 23%. At 7.9% per leg it is 14% and 32%.
- Check the combined probability, not just the odds. Multiply your probabilities. Three 80% legs are a coin flip, not a banker.
- Be sceptical of high-confidence legs. Our picks rated above 80% won about two thirds of the time. Assume your favourites are less certain than they look.
- Never chase a big payout with money you need. See our bankroll guide, and if betting stops being fun, BeGambleAware offers free, confidential help.
Frequently asked questions
What is an accumulator bet?
One bet that combines several selections, also called a parlay, acca or combo. All legs must win. The odds are multiplied, so the payout is bigger than any single, but so is the chance of losing.
How are accumulator odds calculated?
Multiply the decimal odds of all legs. Three selections at 1.40 give 1.40 x 1.40 x 1.40 = 2.74. A void leg counts as 1.00, so the accumulator continues with the remaining legs.
Why do accumulators usually lose?
Two reasons multiply together. The bookmaker margin applies to every leg, so a 3-leg accumulator at 5.4% per leg costs about 14.6% instead of 5.4%. And small errors in your estimates compound: our model expected 63% of its 3-leg combos to win, and only 39% did.
Is a double or treble better than singles?
Not on average. When we combined our published main tips into a daily treble, it returned -10.4% over 128 days, against -5.1% for the same tips as singles. Combining adds margin and variance without adding edge.
Do you publish accumulator tips?
We publish a daily table of high-probability picks that people often combine, and we track accumulators of those picks openly. We do not recommend accumulators as a way to make money, and our own data shows why.
Get free tips, every day
Daily AI-backed football tips in your inbox. Free, no spam, unsubscribe anytime.
18+ only. Bet responsibly. Gambling can be addictive.