Season 26/27Exclusive AccessSign inOpen the boardSeason 26/27Exclusive AccessSign inOpen the boardEvery match of the day by hour, which of them we called, and how those calls finished. Open an hour for its fixtures.
1,891 of 2,920 in-play calls landed. Every call is logged when it is made and settled from the official result, misses included.
Three separate records, never blended into one figure. The thin mark on each bar is 50%.
4,660 in-play calls are left out of this record because the market was already decided when the call was made, for example over 1.5 goals at 2-0, or normal time was over. Nobody could have backed them. A further 392 in-play calls are left out because the call was made before half-time, when the half-time position it names was not yet known. A further 2 in-play calls are left out because the same call on the same match had already been recorded. From 23 September 2026 this record counts only calls timed on the match clock, the day the engine began reading it; the calls made before are shown separately as history and are never added to it.
4,882 single picks are left out of this record because the market was already decided when the pick was published, for example over 1.5 goals at 2-0, or a pre-match pick went out after kick-off. Each pick is counted once, on the day it was first published.
5,788 accumulators are left out of this record because a selection was already decided when the slip was published, for example over 1.5 goals added at 2-0, or a match was already under way. A slip published more than once is counted once.
Every day is published, the weak ones included. Open a date to see its calls.
Pale bars are days with fewer than 20 settled calls; read them as counts, not rates. Today is still running. A strike rate is not a profit.
| Market | Settled | Won | Hit-rate | |
|---|---|---|---|---|
| Over 2.5 goals | 22,977 | 12,502 | 54.4% | |
| Both teams to score | 22,836 | 11,724 | 51.3% | |
| Over 1.5 goals | 21,949 | 16,607 | 75.7% | |
| Home or draw | 16,615 | 12,266 | 73.8% | |
| Home win | 16,615 | 8,427 | 50.7% | |
| Away win | 8,046 | 3,724 | 46.3% | |
| Away or draw | 8,046 | 5,554 | 69.0% |
Markets clear at very different rates, an over-0.5 call and a correct-score call are not comparable, which is why they are never averaged together.
A call is stored the instant it is made, so it cannot be revised afterwards, and it is settled from the official result. In-play calls settle the moment the outcome is beyond reversal, a cleared goals line stays cleared, while results that can still change, like a lead, wait for full time.
Every settled call carries the score it was settled from. A daily integrity pass re-derives each result from that stored evidence and counts any disagreement, currently 0 across 2920 calls.
Expected rates are measured on 76,000 past matches, not assumed. Voids are excluded. Withdrawn strategies keep their losing history in this record, removing them would raise the headline by deleting losers. Predictions are information only and not betting advice. Past performance is not a guarantee of future results. 18+.
Free shows you how we have done. A pass shows you what we are doing right now.
Model picks and in-play calls for 24 hours. No subscription.
Get 24 hours, £4.99Everything stays unlocked, including global live games, mixed and live accumulators, and deeper filters.
Get Pro, £29.99 a monthAlready have an account? Sign in
Every settled call stays in the public record, including the losses. A strike rate is not a profit.
18+. Information only. Gambling involves risk.
Our in-play strategy calls have settled at 64.8% across 2920 calls, and our single picks at 60.5% across 117084 settled picks. Wins and losses are both counted. Past performance is not a guarantee of future results.
We publish a full, chronological, loss-inclusive record with the denominator shown beside every percentage, broken down by day and by market. That is what separates a real record from a headline accuracy claim. 18+, information only.
Each call is recorded when it is made and settled from the official match result. Wins and losses are both counted and voids are excluded. Calls made after their market was already decided, such as over 1.5 goals at 2-0, are left out, and each pick is counted once. Expected rates are measured on 76,000 past matches, not assumed.
Yes. The log is loss-inclusive and every day is published, including the weak ones. Hiding losses would make any hit-rate meaningless.
Because they measure different things. In-play strategy calls, single picks and accumulators are separate populations with different sample sizes, so blending them into one number would hide more than it showed.
