Skip to main content
    Skip to main contentSkip to navigationSkip to footer
    Data & Analytics

    Brier Score

    Also known as:
    Brier Score
    Mean Squared Probability Score
    Updated: 2/12/2026

    A metric measuring the quality of probabilistic predictions – MSE on probabilities (0=perfect).

    Quick Summary

    Brier Score = MSE on probabilities – evaluates calibration and discrimination simultaneously.

    Explanation

    Brier Score = 1/n × Σ(p - y)². Evaluates both calibration and discrimination.

    Marketing Relevance

    Ideal for risk models in medicine, finance, and lead scoring.

    Common Pitfalls

    Only for binary classification. Hard to interpret without baseline.

    Origin & History

    Glenn W. Brier introduced the score in 1950 for weather forecasting. Standard in medical risk research.

    Comparisons & Differences

    Brier Score vs. Log Loss

    Log Loss penalizes errors logarithmically (stronger); Brier quadratically (milder).

    Related Services

    Related Terms

    👋Questions? Chat with us!