TRANSPARENCY / MODEL REGISTRY

Model Arena

Different models see football differently. We are building a transparent record of how each one performs.

PRE-LAUNCHNo verified historical evaluation has been published. Dashes indicate unavailable metrics, not zero scores.

Evaluation Board

Evaluation protocol
Model registry with unavailable historical metrics
MODELSTATUSMATCHES EVALUATEDBRIER ↓LOG LOSS ↓ACCURACYCALIBRATION
EloIllustrativeNot evaluated
PoissonIllustrativeNot evaluated
Expected GoalsIllustrativeNot evaluated
Recent FormIllustrativeNot evaluated
BayesianPlannedNot evaluated
Machine LearningPlannedNot evaluated
NQ EnsembleIllustrativeNot evaluated

Performance will only be reported for timestamped pre-match predictions evaluated against verified results.

Different lenses on the same game.

01 / TEAM STRENGTH

Elo

Relative team strength, updated after each result and adjusted for home advantage.

Demo v0.1 · Illustrative outputs only.

02 / GOAL DISTRIBUTION

Poisson

A score distribution built from expected scoring rates, assuming independent goal counts.

Demo v0.1 · Illustrative outputs only.

03 / CHANCE QUALITY

Expected Goals

Underlying attacking and defensive performance through the quality of chances.

Demo v0.1 · Illustrative outputs only.

04 / TIME-WEIGHTED

Recent Form

Recent performance with more weight placed on newer matches.

Demo v0.1 · Illustrative outputs only.

05 / UNCERTAINTY

Bayesian

A planned framework for updating team-strength estimates and representing uncertainty.

Planned · No implementation or forecasts published.

06 / FEATURE-BASED

Machine Learning

Planned nonlinear models, evaluated on time-separated holdout data.

Planned · No implementation or forecasts published.

07 / MODEL COMBINATION

NQ Ensemble

A proposed blend of complementary models. Current outputs are illustrative fixtures.

Demo v0.1 · Illustrative outputs only.

A PRINCIPLE, NOT A FOOTNOTE

Probability, not certainty.

A 65% win probability still leaves a 35% chance of a different result. Over many comparable matches, a well-calibrated 65% forecast should win about 65% of the time.

Learn to read probabilities