scienceResearch preview · Frozen before kickoff

The Sirius record — every forecast, graded

Sirius reads each game from measured evidence and publishes three chances — home, draw, away — frozen before kickoff. This page grades all three against what actually happened, misses included, plus the games it honestly sat out.

Updated daily as games finish · last graded game 28 Aug 2026

Games graded
983
2949 claims — every forecast grades all three chances
Typical chance on the result
36.3%
market consensus, same games: 36.9%
Average miss
2.9 pts
distance between said and happened, per claim
No forecast
63
not enough track record — an honest sit-out, kept on the record

When we say 60%, does it happen 60% of the time?

Every claim we’ve published, sorted by the chance we gave it. Matching pairs of bars are the whole point of a probability forecast — the grey bar is the promise, the filled one is what football delivered.

0–10%
121 claims
we said
7.1%
it happened
5.2%15.5%
9.1%
10–20%
474 claims
we said
16.1%
it happened
15.1%22.1%
18.4%
20–30%
1021 claims
we said
25.5%
it happened
25.6%31.1%
28.3%
30–40%
501 claims
we said
34.1%
it happened
28.8%37%
32.7%
40–50%
292 claims
we said
44.4%
it happened
36.3%47.5%
41.8%
50–60%
285 claims
we said
54.8%
it happened
43%54.6%
48.8%
60–70%
130 claims
we said
65%
it happened
53%69.5%
61.5%
70–80%
86 claims
we said
74.5%
it happened
57%76.4%
67.4%
80–90%
38 claims
we said
84.6%
it happened
69.6%92.6%
84.2%
90–100%
1 claims
still collecting
we said
90.1%
it happened
20.7%100%
100%

On each “it happened” bar, the two thin marks bracket where the true rate could sit by chance alone — flip a fair coin 20 times and you can easily get 13 heads, not 10; small samples wobble. A gap between said and happened only starts to matter once “what we said” falls outside the marks, and more claims pull the marks closer together.

Is this better than just guessing?

Four ways to forecast the same 333 graded games, scored with the forecast-error score (statisticians’ Brier): 0 is perfect, 2 is confidently wrong — lower is better.

← betterMarket consensus0.5972the market’s average, margin removedSirius0.6088our model — the market is never an inputLong-run averages0.651always forecast home 44% · draw 28% · away 29%Pure guessing0.6667an even 33% on every result

The colored stretch of the line is the ground we’ve covered — 83% of the distance from pure guessing to the market consensus. Scored only where the consensus was saved at least 30 minutes before kickoff. Across all 983 graded games — market consensus saved or not — we score 0.6066; better, but those games skew easier, so the head-to-head stays on the shared 333. Every term on this page, explained in the glossary. How these forecasts are made — the full methodology.

For what it’s worth on most-likely calls alone: ours landed 157 of 333, the market’s 164. We grade the whole distribution, not the call — that single number hides everything this page is for.

Every forecast

Row tint — where the result ranked in our forecast:our top chanceour secondour third
1 / 43

“We gave it” is the chance the frozen forecast put on the result that then happened — 100 would be clairvoyance, and the thin mark on each bar is this record’s typical 36.3%. A 42% result landing is not a triumph and an 18% one is not a disaster; the bands above are where the honest judgement lives.

Across 2949 graded claims our number misses what happened by 2.9 points on average. The middle of our range is honest — between 20% and 70%, said and happened track within 7 points. The ends run wide: on calls of 70% or more we said 78% on average and 73% happened, and on long shots under 10% we said 7.1% while 9.1% happened.

A fair comparison needs the market consensus saved before kickoff, and we have it for 333 of the 983 graded games. On those, the market has been closer to what happened than us — it misses by 2.1 points on average, we miss by 4.8. We are behind it, and this line will say so for exactly as long as that stays true.

Worth weighing before you lean on any of it: only 55 of the 983 graded games come from the top European leagues Sirius is tuned for. The rest was earned elsewhere — mostly qualifying rounds for the UEFA Europa Conference League and UEFA Champions League, plus Major League Soccer. This caveat keeps count, and retires once the tuned-for leagues carry a real share of the record.

Complete record since 7 July 2026, updated daily as games finish. Every number was written before kickoff and frozen; nothing is removed, nothing is regraded. The record so far leans on 244 UEFA Europa Conference League, 96 Major League Soccer, 85 UEFA Champions League and more, across 9 engine versions.

We use cookies to measure how the site is used so we can improve it. Change your choice anytime via “Manage cookies” in the footer. See our Privacy Policy.