Every soccer 1X2 number we published, scored against what actually happened — including the places it is measurably off.
1,937 fixtures27 Jun 2026 – 29 Oct 2026captured ~20 min before kick-offmarket 1X2
Headline
55.1%Called rightHighest-rated outcome matched the result
0.5630Brier scoreWhole-forecast error. Lower is better
0.9497Log lossPenalises confident misses. Lower is better
6.8%Margin removedStripped before the number is published
SkipOdds called right55.1%
Always taking the home side42.8%
Same 1,937 fixtures, both figures.
Soccer is a low-signal sport — the ceiling is lower than most people expect.
Any number claiming to call much past the mid-fifties across mixed competitions is worth reading twice.
Calibration
Each dot is one confidence band, sized by how many outcomes fall in it. A perfectly calibrated number sits on the dashed line; the grey stem is the miss.
Band
Outcomes
SkipOdds said
Happened
Gap (pts)
0–5%
53
3.3%
1.9%
-1.4
5–10%
186
7.7%
3.8%
-3.9
10–20%
864
15.8%
14.6%
-1.2
20–30%
2,180
25.3%
25.0%
-0.3
30–40%
860
34.3%
33.5%
-0.8
40–50%
670
44.6%
46.7%
+2.1
50–65%
606
57.0%
57.8%
+0.8
65%+
392
74.8%
78.3%
+3.5
5,811 outcomes across 1,937 fixtures.
Where it is off, stated plainly
Widest two bands: 5–10% — stated 7.7%, happened 3.8% over 186 outcomes; 65%+ — stated 74.8%, happened 78.3% over 392 outcomes.
Long shots come in under their stated chance; strong favourites over it — the classic favourite–long-shot pattern.
Three alternative ways of removing the margin were tested. All three flatten those tails as theory predicts.
None improved overall accuracy on the sample tested — 856 fixtures, paired t of 0.3.
The pattern is real, the obvious fix does not pay, and nothing was changed on the strength of it.
By competition
Accuracy is not one number — it is a different number in every competition.
Below: every competition with at least 25 scored fixtures in the window.
Competition
Fixtures
Called right
Brier
Major League Soccer
226
47.3%
0.6320
EFL Championship
144
46.5%
0.6346
FA Cup
114
56.1%
0.5936
UEFA Nations League
104
58.7%
0.5163
Liga MX
100
45.0%
0.6421
La Liga
99
55.6%
0.5415
UEFA Europa Conference League
88
58.0%
0.5348
Serie A
88
61.4%
0.5276
Eredivisie
87
50.6%
0.5971
Saudi Pro League
86
53.5%
0.5388
Brasileirão
86
62.8%
0.5344
Premier League
80
48.8%
0.6075
UEFA Europa League
79
62.0%
0.5216
Ligue 1
72
52.8%
0.5825
Primeira Liga
71
49.3%
0.5964
UEFA Champions League
67
65.7%
0.4785
Bundesliga
63
55.6%
0.5955
Süper Lig
59
61.0%
0.5204
DFB-Pokal
46
76.1%
0.3711
FIFA World Cup 2026
38
73.7%
0.4191
CAF Champions League
32
68.8%
0.3830
Türkiye Kupası
32
62.5%
0.5089
What this does not yet prove
Coverage is 44%
44%
56%
1,937 of 4,403 finished fixtures carried a closing number.
The rest were never covered by enough sources to publish one.
This describes the fixtures SkipOdds covers, not the whole calendar.
The aggregation itself is untested
SkipOdds averages the prices it collects from up to 79 sources, converts them and normalises.
Whether that is the best way to combine them is untested — against a median, against weighting each source by its own record, against discarding placeholder quotes.
It could not be answered from what was stored: individual source prices were overwritten on every refresh.
Since 14 September 2026 they are frozen alongside each closing number.
The answer will be published here whichever way it falls.
Method
Every figure is recomputed from the frozen pre-kick-off snapshot — the number as published, roughly twenty minutes before the start.
Nothing is scored from prices captured after kick-off.
Called right — share of fixtures where the highest-rated outcome matched the result.
Brier — mean squared error across all three outcomes. A number honest about a close fixture scores well even when its top pick loses; one emphatic about a coin-flip scores badly even when it wins.
Margin removed — the average overround stripped before publication.
Recomputed 2026-10-29 14:00 UTC. Past accuracy does not predict future accuracy.