Same-game accumulators lose 28% more legs when one selection is priced late, according to a 14-month analysis of 41,700 UK same-game acca (SGA) slips settled across three licensed operators. The pattern held across football, racing and tennis markets: when one leg was added more than 90 minutes after the others, 31.4% of slips lost that leg specifically, against 24.5% on slips where all legs were priced within the same window. The finding matters because SGAs now account for roughly a fifth of accumulator stakes at UK-licensed books, and late pricing is the most common way punters build them.

What the data actually shows

The 28% figure is a relative gap, not an absolute one. On matched slips — same sport, same stake band, same number of legs — the late-leg loss rate ran 6.9 percentage points higher. That gap is consistent enough to be worth understanding, but it is not a curse. It reflects what late pricing does to a bet, not what the universe does to it.

The sample was split by sport and by leg count. The effect was strongest on three-leg and four-leg SGAs, where a single late leg carries more weight in the combined price. On five-leg and six-leg slips the gap narrowed to 4.1 points, partly because the late leg is diluted and partly because those slips fail for other reasons anyway.

Leg count	Late-leg loss rate	On-time loss rate	Gap
2	27.8%	23.1%	+4.7
3	31.9%	24.6%	+7.3
4	32.7%	25.2%	+7.5
5–6	29.4%	25.3%	+4.1

Racing showed the widest gap at 9.2 points, driven by non-runners and going changes between the time the first leg was priced and the last. Football sat at 6.4 points, tennis at 5.8.

Why a late leg loses more often

Three mechanisms explain most of the gap. None of them require the bookmaker to be doing anything underhanded, though one of them is worth watching.

Price movement is information

When you add a leg 90 minutes after the rest, the price on that leg has usually already moved. If it has shortened, you are taking a worse price than the market now considers fair — and the market has had time to absorb team news, weather, or money. If it has drifted, you are taking a price the market has moved away from, which on average means you are on the wrong side of a shift. Either way, the late leg is priced against you relative to the earlier snapshot.

The bet is no longer a portfolio

An SGA built in one sitting is a set of correlated views taken at the same moment. A late leg breaks that. You are now combining a view formed at 2pm with a view formed at 4pm, and the second view was formed with knowledge of the first result in some cases — particularly on in-play SGAs, where a late leg added after one selection has already settled is effectively a new bet stapled to a dead one. Those slips failed at 38.6%, the worst cohort in the sample.

Non-runners and line moves

In racing, a late leg is more likely to be affected by a non-runner, which can void that leg and recalculate the whole SGA at reduced odds. In football, a late leg is more likely to sit on a line that has already been adjusted for team news. The bettor is not wrong about the outcome; they are late to the price.

The operator side of the ledger

Books price SGAs with a correlation margin, and that margin is not fixed. It widens when the book is carrying exposure on a popular combination and narrows when it wants volume. A late leg is often added to a slip the book has already priced, which means the operator can re-price the whole combination rather than just the new leg. In practice, the SGA price you see after adding a late leg is frequently worse than the sum of the individual legs would suggest — and worse than the price you would have got had you built the slip in one go.

This is not a scandal. It is a pricing decision, and it is disclosed in the terms. But it does mean the late-leg penalty is partly a margin effect and partly a genuine information effect, and the two are hard to separate from the outside.

What a sensible SGA builder does differently

The practical takeaway is not to avoid SGAs. It is to build them in one window.

• Set a build window. Decide the slip before you start adding legs, and add them all within 15 minutes. If a leg is not priced by then, drop it rather than wait.
• Check the re-price. After adding a late leg, compare the combined price to what the same legs would pay as singles. If the SGA price is more than about 8% below the singles product, the correlation margin is doing more work than the correlation itself.
• Avoid adding legs to in-play SGAs after a leg has settled. This is the cohort with the worst loss rate in the data, and it is the one where the bettor has the least information advantage.
• Treat a shortened late leg as a warning. If the price has moved against you between build and add, the market has already priced the news you are trying to act on.

None of this is a guarantee. A well-built SGA still loses more often than it wins, which is why the stake discipline matters more than the leg selection. The 28% gap is a reason to be deliberate, not a reason to stop.

The question the data cannot answer

The analysis cannot tell us whether late legs lose more because they are late, or because the kind of bettor who adds a late leg is also the kind of bettor who chases. The two are entangled in any observational sample, and no operator will run the controlled test that would separate them. What is clear is that the gap is large enough to be worth building around — and that the cheapest fix, building the slip in one sitting, costs nothing but a little patience. Whether UK books will ever publish late-leg pricing data voluntarily is a separate question, and on current form the answer is no.