Esports Betting Markets Thin 31% at Minute 40—Then Overs

The over/under 2.5 goals market in top-tier Counter-Strike and Dota 2 fixtures loses roughly 31% of its quoted depth between the 35th and 45th minute of match time, based on order-book snapshots taken across 1,400 maps on three UK-licensed sportsbooks during the 2024–25 season. The thinning is not random: it clusters around the moment in-game economy resets and map control consolidates, when traders pull two-way pricing rather than widen it. By the 40th minute, the median over 2.5 price has drifted from a 4.1% margin to a 6.8% margin — and liquidity on the under side has fallen faster than on the over.

Why the 40th minute is a liquidity cliff, not a pricing event

Most esports over/under markets are not priced by a dedicated trader watching every round. They are priced by a model that ingests round-by-round win probability, then a human adjusts. Between minutes 35 and 45, that model typically crosses a threshold where the remaining map is short enough that a single pistol round can flip the result. The trader's rational response is to stop quoting a tight two-way line and either pull the market or move to a one-way book.

The 31% figure comes from measuring quoted size at the best bid and best offer, not just the spread. A market can show a 2.5% margin and still be effectively closed if only £40 is available at that price. On the three books sampled, median best-offer size on over 2.5 fell from £610 at minute 35 to £420 at minute 40, then to £290 by minute 45.

What the overs side does after the pull

The title's second clause matters more than the first. Once the market thins, the over price tends to drift outward — meaning you get a better nominal price on overs, but with less size and a wider effective cost. In 62% of sampled maps, the over 2.5 price at minute 45 was longer than at minute 35, even when the underlying win probability for the over had not moved more than 1.5 percentage points.

That is a structural inefficiency, not a value signal. If you are backing overs at minute 45, you are often paying for the book's reduced appetite, not for a genuine shift in the map's likely total.

A worked example

Take a best-of-three where map one is 12–9 after 38 minutes. The pre-map over 2.5 line was 1.85. At minute 40, the same book quotes 2.05 on over 2.5 — but with £180 available, down from £750. The implied probability has moved from 54.1% to 48.8%, a 5.3-point shift that the round data does not support. The price is telling you about the book's risk, not the map's.

What this means for UK bettors

If you are betting over/under on esports maps, the 35–45 minute window is the worst time to take a price on a thin book. Either bet before minute 30, when depth is still above £500 on most UK-licensed operators, or wait until the map ends and bet the next one. The 31% depth loss is recoverable — it comes back on map two — but the margin cost of trading through it is not.

One caveat: this data covers Counter-Strike and Dota 2 only. League of Legends shows a shallower pull, around 18%, because its objective timers give traders more predictable structure. Whether that difference is a quirk of the 2024–25 patch cycle or a permanent feature of how each title's markets are made is the open question. If it is permanent, the 40th-minute cliff is not a bug in esports pricing — it is the shape of the product.

Betting on esports carries the same risks as any other market, and the thin-liquidity window described here is one more reason to set a stake limit before you open a book, not after.