Live vs Prematch: Where Surebets Last Longer
In Dozor's archive, the median prematch surebet lasts 41 seconds vs 2 in live; 30% vs 1.8% last over a minute. What that changes in how you work each mode.
A prematch surebet lives twenty times longer than a live one. In Dozor's surebet archive for September 6, 2026 (17:21–21:08 UTC, 14,645 surebets with a known mode), the median lifetime was 41.1 seconds in prematch and 2.1 seconds in live. 29.7% of prematch surebets lasted longer than a minute, versus 1.8% in live. At the same time, prematch produced 1,197 surebets versus 13,448 in live: the mode is calmer, but also thinner.
Methodology
The Dozor server writes every surebet it finds to an archive: the books, market, odds, and the time of the first and last cycle on which it held. Lifetime is the difference between those cycles; a surebet found on a single cycle gets 0 seconds. The mode is determined by the match status at the first cycle: started means live, not started means prematch. For 10,008 of the 24,653 surebets in the sample, the match status was unknown (the book doesn't provide a start time, or the match wasn't linked to the schedule), and they were left out of the mode comparison.
How long live and prematch surebets last
| Metric | Live | Prematch |
|---|---|---|
| Surebets in sample | 13,448 | 1,197 |
| Median lifetime | 2.1 s | 41.1 s |
| Over a minute | 1.8% (≈242) | 29.7% (≈356) |
The key row is the last one. In absolute numbers, prematch produced more surebets lasting over a minute than live did (≈356 vs ≈242), on one-eleventh of the volume. If you only bet on what a person with an alert can catch, prematch is no poorer than live; what differs is the density. In live you have to dig those surebets out of thousands of instant discrepancies; in prematch they make up a third of the flow.
Distribution across all 24,653 surebets in the sample, including those with an unknown mode:
| Lifetime | Surebets | Share |
|---|---|---|
| under 2 s | 9,302 | 37.7% |
| 2–10 s | 8,663 | 35.1% |
| 10–30 s | 4,148 | 16.8% |
| 30–60 s | 1,555 | 6.3% |
| 1–5 min | 809 | 3.3% |
| over 5 min | 176 | 0.7% |
Almost three quarters of surebets close within ten seconds. That's live: one book updated its line after a round, and the other caught up a moment later.
Why prematch lasts longer
Before a match, lines are moved not by events but by money and information. A bookmaker changes its odds after a large bet, lineup news or a trader's reassessment, and it does so rarely: hours between adjustments are normal. A gap between two books in prematch is more often a different assessment of team strength than a delay, and it holds until one of the books changes its mind.
In live, an algorithm moves the line after every round, kill or tower. Two books react with different delays, and in those fractions of a second a surebet forms between them. It closes on its own as soon as the second book updates its price — hence the two-second median. More on the mechanics on the Live surebets and Prematch surebets pages.
Where to look for long-lasting surebets
The difference between modes shows up within markets, too. The median lifetime on handicaps, totals and match results is 2.0–2.1 seconds: these are live markets with algorithmic lines. Total kills lasts 14.3 seconds, because a trader adjusts kill lines by hand, and less often. The same picture holds by game: CS2 and Valorant sit at a median of about 2 seconds, League of Legends at 10.2, Rainbow Six at 41.3. The fewer books that price a game in live, and the less often they update the line, the longer a discrepancy holds.
Hence a practical list of places where a prematch approach works in live as well:
- kill totals and kill handicaps in Dota 2 and LoL — markets with a manual line;
- second-tier games and leagues: Rainbow Six, LoL, regional CS2 leagues;
- crossed totals (middles) in prematch, where both bets can win — see the term middle.
What this changes in practice
- Prematch is the mode for checking. Forty seconds is enough to open both books, compare the settlement rules and work out the stake split in the calculator. In live you don't have that time: the calculation has to be ready in the card before you open it.
- Live is the mode for volume, but with a filter. 13,448 surebets in four hours is a flow in which anyone without a filter on lifetime and price age will only see surebets that have already closed. Dozor shows the age of every price and lets you filter out surebets with a stale leg.
- Prematch gets you limited faster. A long-lasting surebet isn't visible only to you: a bet against a line the book has held for hours is a clear sign of an arber. More in the article “Why Bookmakers Limit Arbitrage Bettors and What to Do”.
- Exchanges are a special case. On Polymarket and Kalshi, prices are moved by trades, and discrepancies with bookmakers last longer in both modes; the numbers are on the Prediction markets page.
FAQ
Why are there so few prematch surebets in the sample?
On a weekday evening, most of the volume comes from matches in progress. Prematch lines for tomorrow's matches are posted by fewer books and rarely change, so there aren't many discrepancies between them over an evening. On weekends and before major tournaments, the prematch share is higher.
Does a 41-second median mean a prematch surebet is guaranteed to survive until the second bet?
No. Half of prematch surebets closed in under 41 seconds, and what closes them isn't an event in the match but the second book noticing bets against its line. You still need to check price age here.
Why is the mode unknown for 10,008 surebets?
Some books don't provide the match start time, and the scanner can't link some matches to the schedule when books spell team names differently. These surebets are in the archive but don't make it into the breakdown by mode. The share will shrink as the matching rules fill out.
Which mode should a beginner pick?
Start with prematch and markets with a manual line: there's time to check the rules and limits, and a mistake costs you one bet, not a series. Live makes sense once the math and keeping the books' tabs open have become habit.