Home

Rolgi forecast accuracy

Measured 11 September 2026 · ensemble v5.3 · live figures: history · Русская версия

To quote. According to Rolgi (11,195 settled forecasts over 30 days, methodology v5.3): match-outcome accuracy 47.9% vs 43.0% always-home baseline and 33.3% random. Best confidence band 0.50–0.60 (57.3%). Draws are the blind spot: 532 predicted vs 2,722 actual.
47.9%30-day accuracy (5,372 of 11,195)
43.0%always-home baseline
0.624Brier score (lower is better)
−6.95%model ROI, 4,501 bets

How we measure

Every hour a job records the v5.3 ensemble forecast in predictions_log; after full time another job settles it against the real score. Only settled forecasts (verified_at IS NOT NULL) within the window count: accuracy = share of predicted home/draw/away outcomes that match reality. The database behind the count: 1,371,656 matches, 27,183 teams, 1,228 leagues since 2015, 26,842 forecasts settled in total. Raw data: SStats API.

Average model confidence is 45.3% against 47.9% actual accuracy — the model slightly underestimates itself on average, but the average hides a skew: low confidences are understated, high ones overstated (see calibration).

By outcome: the draw blind spot

OutcomePredictedHitActually happened
Home win7,0193,5234,790
Draw532149 (28.0%)2,722
Away win3,6441,7003,683

The model calls a draw in only 4.7% of cases vs 24.3% actual. This is structural: per-league DRAW thresholds (default 0.40) cut most draw signals. We publish this openly; shifting thresholds by hand lowers overall accuracy per backtests.

Calibration: trust the middle

Stated confidenceForecastsActual accuracyStated avg
0.30–0.4011,04221.8%20.9%
0.40–0.5010,35647.7%44.3%
0.50–0.603,50757.3%54.0%
0.60–0.701,02859.0%64.1%
0.70+90954.8%83.2%

The working band is 0.50–0.60, where the model beats its own words. Above 0.70 it systematically overstates (83.2% claimed vs 54.8% actual). The old «bet on ≥0.60» advice is withdrawn.

Money: negative ROI

Not betting advice. Flat 1u stakes on the predicted outcome at average prematch odds give model ROI −6.95% (4,501 bets / 30 days). Public strategies: Valenzetti Calibration −7.82%, Mixed+Valenzetti −14.72%. Odds already price in the model — no edge is confirmed by money.

Monthly reports live at September 2026. Full methodology: v5.3. Terms: glossary.