2026/27 · prospective official record
Season progress
Every matchweek we have tracked, from calls locked before kickoff only. No backtests, experimental forecasts or reconstructed predictions are counted here. A revised call counts once, by its last pre-kickoff version.
Actual against expected
Settled Super Tips only
Table view
| # | Matchweek | Fixture | Locked probability | Result | Actual so far | Expected so far |
|---|---|---|---|---|---|---|
| 1 | 9 | Brentford v Nott'm Forest | 73% | Won | 1 | 0.73 |
| 2 | 9 | Coventry v Sunderland | 63% | Lost | 1 | 1.36 |
| 3 | 9 | Hull v Ipswich | 55% | Lost | 1 | 1.92 |
| 4 | 9 | Man City v Brighton | 58% | Lost | 1 | 2.50 |
| 5 | 9 | Aston Villa v Fulham | 74% | Won | 2 | 3.23 |
| 6 | 9 | Liverpool v Arsenal | 69% | Won | 3 | 3.93 |
Expected wins is the sum of the probabilities we locked for these same 6 settled tips (3.9). It is a yardstick for honest probabilities, not a target or a guarantee. 6 settled tips is a small sample: a gap either way is mostly variance and proves no advantage.
Every matchweek
Open one for its full scorecard and recap
| Matchweek | Status | Coverage | Tips | Passes | Won–lost | Pending | Push / void | Model |
|---|---|---|---|---|---|---|---|---|
| Matchweek 9 Sat 31 Oct | Final | 10/10 | 6 | 4 | 3–3 | 0 | 0 / 0 | tl-poisson-v1.1 |
Model versions
Never pooled silently
| Version | Matchweeks | Calls | Super Tips | Settled | Won | Expected | Average probability |
|---|---|---|---|---|---|---|---|
| tl-poisson-v1.1 | 9 | 10 | 6 | 6 | 3 | 3.9 | 65% |
Not counted, and why
So nothing improves by omission
Missing forecasts (0)
Every kicked-off fixture in a tracked matchweek had a locked call.
Unresolved verification (0)
No open review or played Super Tip awaiting verification.
No profit or return is claimed: that would need odds recorded at lock time, and these figures use probabilities only. Calibration, Brier scores and market breakdowns are on Model accuracy details.