Free scale checker · No upload

Brier score scale & normalization checker

Convert the four conventions most likely to make two football forecast reports look different—even when the underlying squared error is identical.

Canonical publication record

Abstract

A deterministic browser tool for converting summed, class-averaged, half-scaled and binary Brier conventions while preserving displayed precision and enforcing outcome-space, row and aggregation evidence gates.

Author and publisher
Football Proof AI
Technical report
football-brier-scale-compatibility/1.0.0
Published
Last modified
Release status
Current release
Review status
Editorial technical note; not externally peer reviewed
Version history
  1. football-brier-scale-compatibility/1.0.0 : Initial public release.

Interactive scale passport · Browser local

Put two Brier scores on the same mathematical scale

The default example is deliberate: 0.150 class-averaged and 0.450 summed are the same three-class error. Conversion removes notation ambiguity; the evidence checks decide whether a performance comparison is even admissible.

Score ADeclare its published convention
ΣK original
0.45
÷K class average
0.15
÷2 half-scaled
0.225
Uniform reference
0.666667 on ΣK
Score BDeclare its published convention
ΣK original
0.45
÷K class average
0.15
÷2 half-scaled
0.225
Uniform reference
0.666667 on ΣK

Formula crosswalk

One forecast error can carry three multiclass numbers

Let K be the number of mutually exclusive outcomes. The original multicategory form averages the sum of K squared class errors over N rows. Other conventions divide that same value by K or by two. The conversion is arithmetic, not a performance claim.

Original summed conventionSΣ = N⁻¹ Σᵢ Σ꜀ (pᵢ꜀ − yᵢ꜀)²

Exactly one y value is 1 for each settled row; all other class indicators are 0.

Declared conventionRelationship to SΣTheoretical rangeSame 1X2 error
Summed full vector0 to 20.450
Class-averagedSΣ ÷ K0 to 2/K0.150
Half-scaledSΣ ÷ 20 to 10.225
Binary event MSESΣ ÷ 2 only when K = 20 to 1Not a 1X2 convention

For uniform K-class probabilities, the summed reference is 1 − 1/K. In 1X2 that is 0.6667 summed, 0.2222 class-averaged or 0.3333 half-scaled.

Search question

Is a Brier score of 0.15 better than 0.45?

Not enough information.

If 0.15 is a class-averaged three-outcome score and 0.45 is the original summed score on the same rows, they are identical. If both use the same convention, 0.15 is lower—but “lower” becomes an admissible model comparison only after matching the exact targets, rows, exclusions, weights and aggregation.

Interpretation boundary

What is a good Brier score for football predictions?

There is no context-free threshold. Lower is better only within a fixed scoring convention and outcome task. A useful report compares the model with a frozen baseline on identical rows, publishes the sample size and period, and inspects calibration instead of promoting one aggregate number.

  1. 01

    Fix the task

    1X2, BTTS and binary match events have different class structures and base rates.

  2. 02

    Fix the scale

    Name the formula and K. “Brier 0.20” alone is not reproducible.

  3. 03

    Fix the rows

    Use the same settled matches, exclusions, time cut-off and row weights.

  4. 04

    Fix the baseline

    Compare against a declared frozen reference rather than an invented universal pass mark.

Do not auto-convert

Four labels that need more than a multiplier

Reference-dependent

Brier Skill Score

BSS = 1 − BS/BSref. Without the exact reference score and matched rows, it cannot be reversed into a raw Brier score.

Unit error

Percentage-point squares

Probabilities must be divided by 100 before applying the standard formula. Squaring percentage points produces values 10,000 times the decimal-space score.

Different metric

RPS, log loss or ECE

These answer different questions. None can be recovered from an aggregate Brier score by a fixed conversion.

Missing evidence

Unknown convention

A plausible-looking number is not enough to infer a formula. Ask the publisher for the equation, K, sample and exclusions.

Precision-aware comparison

0.150 is a rounded interval, not an infinitely precise value

The checker preserves the typed decimal places. A displayed value of 0.150 represents the rounding interval from 0.1495 to 0.1505 before scale conversion, assuming ordinary rounding to the nearest displayed unit. Two converted point estimates may differ while their declared precision intervals still overlap. The JSON passport records that distinction.

Interval overlap means arithmetic compatibility at the displayed precision. If a publisher truncated instead of rounded, its rule must be checked separately. Compatibility does not establish calibration, statistical significance, independence, profitability or historical publication.

Primary and authoritative sources

Definitions anchored to the scoring-rule literature