研究笔记

Calibrating predictions differently for friendlies vs tournaments

状态: Shipped (Variant 4 — per-tier Platt temperature scaling). Production calibrator uses a hybrid strategy: Platt for the tournament tier (where isotonic collapses to identity at n~70), isotonic for friendlies/qualifiers (where it's more expressive at n~400+). Gate passed. See results below要点 + 全文 · 3,433 字

要点

The shipping ensemble calibrator (scripts/fit_ensemble_calibrator.py) fits per-class isotonic regression curves on the uniform-averaged three-component output (Elo bracket MC + Dixon-Coles + Hierarchical Poisson MAP). The first cut lifted holdout ECE on the 365-day common-subset training pool from 4.62pp uncalibrated → 2.70pp under the pooled-across-tiers fit (5-fold CV, n_train = 939, current artefact at data/wc2026/ensemble_calibrator.json).

A subsequent tier-aware refit (three sets…

全文

Standard Pass

阅读完整研究笔记

Calibrating predictions differently for friendlies vs tournaments 共 3,433 字。Standard Pass 解锁所有研究笔记全文, 以及完整预测和逐队、逐球员评分, 有效期至赛事结束。

获取 Pass — $15

Every forecast graded against the real result, scored on 987 matches since 2014. See the scorecard.

24h money-back, no questions asked·No subscription, no auto-renewal·Access through 31 Dec 2026. See refund policy.