Football ratings

Wednesday 9 September 2026 · times UK

The model's strongest calls

75%Celtic — away winST Johnstone v Celtic · 20:00 · market 65% · this strong, 76% land (315 cases)

74%Twente — home winTwente v Telstar · 17:45 · market 75% · this strong, 78% land (4,355 cases)

67%Benfica — away winMoreirense v Benfica · 20:45 · market 83% · this strong, 71% land (1,529 cases)

£10 on all 3 as one multiple returns about £21 if every one lands. The model puts that at 38%, the market at 40%, and the price needs 47% just to break even.Bookmaker margin compounds with every leg, which is why the break-even sits above the market's own view. A same-match bet builder pays less than this — the bookmaker prices in that the legs move together. Our chance is a straight multiplication and is if anything a little low: where two legs of one match both cleared, they landed together 55.5% of the time against a multiplied 53.4%.

Goals — supporting, not headline

These are calibrated — a 65% call lands about 65% of the time. But the model separates high- from low-scoring matches far less sharply than it separates winners, so there is less information behind a goals number than a result one.

74%Barcelona v Feyenoord — over 2.5 goals17:45 · market 81% · this strong, 76% land (485 cases)

68%ST Johnstone v Celtic — over 2.5 goals20:00 · market 64% · this strong, 70% land (1,528 cases)

68%Twente v Telstar — over 2.5 goals17:45 · market 73% · this strong, 70% land (1,528 cases)

62%Liverpool v Atletico Madrid — over 2.5 goals20:00 · market 62% · this strong, 66% land (4,584 cases)

57%Napoli v Arsenal — under 2.5 goals20:00 · market 51% · this strong, 58% land (10,295 cases)

Every fixture

Championship

Derby v West Brom19:45

	Home	Draw	Away
Model	41	28	31
Market	31	28	40

	Over 2.5	Under 2.5
Model	51	49
Market	48	52

Under 4.5 goals: 86% · happens 86% of the time anyway

Model is 10 points above the market on home win (pinnacle). Worth a look at why.

Norwich v Birmingham19:45

	Home	Draw	Away
Model	47	26	27
Market	51	25	25

	Over 2.5	Under 2.5
Model	54	46
Market	55	45

Under 4.5 goals: 83% · happens 86% of the time anyway

Charlton v QPR20:00

	Home	Draw	Away
Model	33	28	40
Market	30	28	41

	Over 2.5	Under 2.5
Model	52	48
Market	49	51

Under 4.5 goals: 85% · happens 86% of the time anyway

Scottish Premiership

Rangers v ST Mirren19:45

	Home	Draw	Away
Model	65	22	13
Market	73	17	10

	Over 2.5	Under 2.5
Model	58	42
Market	67	33

Under 4.5 goals: 82% · happens 86% of the time anyway

Model is 8 points below the market on home win (pinnacle). Worth a look at why.

ST Johnstone v Celtic20:00

	Home	Draw	Away
Model	9	16	75
Market	16	20	65

	Over 2.5	Under 2.5
Model	68	32
Market	64	36

Under 4.5 goals: 70% · happens 86% of the time anyway

Model is 11 points above the market on away win (pinnacle). Worth a look at why.

Eredivisie

Twente v Telstar17:45

	Home	Draw	Away
Model	74	17	9
Market	75	14	10

	Over 2.5	Under 2.5
Model	68	32
Market	73	27

Under 4.5 goals: 71% · happens 86% of the time anyway

Primeira Liga

Moreirense v Benfica20:45

	Home	Draw	Away
Model	12	21	67
Market	6	12	83

	Over 2.5	Under 2.5
Model	57	43
Market	64	36

Under 4.5 goals: 84% · happens 86% of the time anyway

Model is 15 points below the market on away win (pinnacle). Worth a look at why.

UEFA Champions League

New competition — experimental. The model has never been trained or tested on it, these clubs are rated across leagues whose ratings have never been calibrated against each other, and the probabilities use a fallback calibration. Treat them as much weaker than the domestic ones.

Barcelona v Feyenoord17:45

	Home	Draw	Away
Model	66	19	15
Market	89	7	4

	Over 2.5	Under 2.5
Model	74	26
Market	81	19

Under 4.5 goals: 65% · happens 86% of the time anyway

Model is 22 points below the market on home win (pinnacle). Worth a look at why.

VfB Stuttgart v Viking17:45

Not rated — Viking has 0 prior matches in the store, most likely promoted from a division we do not collect. A rating built on that would be a guess wearing a percentage.

Liverpool v Atletico Madrid20:00

	Home	Draw	Away
Model	55	24	21
Market	55	24	21

	Over 2.5	Under 2.5
Model	62	38
Market	62	38

Under 4.5 goals: 78% · happens 86% of the time anyway

Napoli v Arsenal20:00

	Home	Draw	Away
Model	26	30	44
Market	17	25	57

	Over 2.5	Under 2.5
Model	43	57
Market	49	51

Under 4.5 goals: 91% · happens 86% of the time anyway

Model is 13 points below the market on away win (pinnacle). Worth a look at why.

Paris Saint Germain v Slovan Bratislava20:00

Not rated — Slovan Bratislava has 0 prior matches in the store, most likely promoted from a division we do not collect. A rating built on that would be a guess wearing a percentage.

Sporting CP v Galatasaray20:00

	Home	Draw	Away
Model	37	27	35
Market	54	24	22

	Over 2.5	Under 2.5
Model	55	45
Market	60	40

Under 4.5 goals: 84% · happens 86% of the time anyway

Model is 16 points below the market on home win (pinnacle). Worth a look at why.

How the calls did

✓NEC Nijmegen 2–2 Excelsior over 2.5 goals 65%

✓Utrecht 3–3 GO Ahead Eagles over 2.5 goals 63%

2 of 2 settled since yesterday · 152 of 225 since the model went live.

Every match, for the record

The model rates the whole card, not just the calls, and learns from all of it. Judged as probabilities rather than picks — a 38% favourite losing is the model working, not failing.

·Resulttop pick right 5/12, we claimed 47% · 217/442 all-time

what actually happenedwe gave it 36% on average

·Over/under 2.5top pick right 5/12, we claimed 54% · 257/442 all-time

what actually happenedwe gave it 53% on average

A well-calibrated model's average claim matches how often it happens. That is the number Rule 9 gates on, and it is tracked for results and goals alike whether or not a fixture ever produced a call.

What these numbers are. The model's own estimate of how likely each result is, from form, Elo, rest, shots and xG as they stood at 06:00 UK on match day. One bivariate Poisson produces all of them, so the result and the goals figures share one underlying view of the match. Each published figure is then calibrated against what actually happened in past matches, one market at a time — so they agree closely but are not forced to agree exactly. Market figures are the bookmaker's prices with their margin removed, so the two are comparable.

What they are not. Across 57,633 out-of-sample fixtures the market separated matches better than the model in every league, every season, every outcome and every confidence band — 24 slices, no exceptions, re-tested in August 2026 on the current model. This is a second opinion, not an edge. Where it earns its keep is honesty about its own confidence: when it says 74%, that happens about 74% of the time.

Why a call appears. Thresholds — home 70% · away 60% · over 60% · under 55% — sit where the model has historically been as good as its claim. Raising them would surface fewer calls and be less accurate about them, because its confidence above those levels is inflated everywhere except home wins. Draws are never flagged: the highest one ever rated was 62%.

11 rated · 2 not rated · model goals_v7_cal · built 2026-09-09 08:35 UK