Demo -- read-only snapshot of the live system as of 2026-07-16T01:45:31Z; no live data, paper units only.
Skip to main content
← receipts / scoreboard

soccer_intl/brier

The ledger row, decomposed. Every figure below is read from a staged artifact on disk (cited per panel); this drill describes calibration and sharpness only -- no dollar edge, ROI, or bankroll is claimed.

model
0.3719 (Brier)
market
0.2645 (Brier)
delta (95% ci)
-0.1075 [-0.1885, -0.0316]
n
17

verdict UNDERPOWERED -- source scripts/platformkit/benchmarks/crps_market/last_run_ingame_soccer.json

Murphy decomposition (model)

RELIABILITYreliability = 0.1920 (lower better)0.1920RESOLUTIONresolution = 0.0387 (higher better)0.0387UNCERTAINTYuncertainty = 0.2135 (baseline)0.2135

Brier = reliability - resolution + uncertainty (10-bin). Lower reliability and higher resolution are better; uncertainty is the irreducible base-rate floor. Descriptive, not an edge claim. Source: murphy_decomposition.json.

reliability curve (model vs market)

as of 2026-09-16
.00.00.25.25.50.50.75.75.00.00market: predicted 0.059, observed 0.309, n=603market: predicted 0.164, observed 0.348, n=486market: predicted 0.240, observed 0.554, n=969market: predicted 0.347, observed 0.757, n=224market: predicted 0.451, observed 0.889, n=405market: predicted 0.556, observed 0.895, n=467market: predicted 0.651, observed 1.000, n=223market: predicted 0.719, observed 0.989, n=271market: predicted 0.858, observed 1.000, n=474market: predicted 0.911, observed 1.000, n=143model: predicted 0.045, observed 0.667, n=1284model: predicted 0.144, observed 0.519, n=792model: predicted 0.246, observed 0.440, n=669model: predicted 0.351, observed 0.717, n=449model: predicted 0.448, observed 1.000, n=180model: predicted 0.551, observed 1.000, n=202model: predicted 0.635, observed 0.949, n=136model: predicted 0.755, observed 1.000, n=198model: predicted 0.841, observed 1.000, n=194model: predicted 0.962, observed 1.000, n=161PREDICTED

Predicted probability (x) vs observed frequency (y) against the perfect-calibration ideal (dashed). Amber = model, blue = market. Hover a point for its bin count. Source: calibration_stability.json.

bootstrap CI per bin (model)

as of 2026-09-16
binngapgap 95% ciexcludes 0?
0.0-0.112840.6217[0.357, 0.828]yes -- gap real
0.1-0.27920.3752[0.132, 0.653]yes -- gap real
0.2-0.36690.1929[0.005, 0.459]yes -- gap real
0.3-0.44490.3664[0.027, 0.618]yes -- gap real
0.4-0.51800.5524[0.509, 0.581]yes -- gap real
0.5-0.62020.4488[0.436, 0.455]yes -- gap real
0.6-0.71360.3133[-0.293, 0.392]no -- within noise
0.7-0.81980.2445[0.211, 0.285]yes -- gap real
0.8-0.91940.1591[0.112, 0.163]no -- within noise
0.9-1.01610.0381[0.021, 0.086]yes -- gap real
8/9 eligible bins have a calibration-gap 95% CI (cluster bootstrap on game_id) that excludes 0; the rest are within noise. Model Brier 0.3325. Brier skill vs climatology -0.857, vs market -0.704 (at or below 0 -- the model does not beat the market Brier; that null is the point). Source: calibration_stability.json + brier_skill_scores.json.