Expected goals (xG) is the probability that a given shot attempt becomes a goal, estimated from where and how it was taken — distance, angle, shot type, whether it followed a rebound or a rush, the strength state, and more. The model is deliberately talent-blind: it scores the chance for a league-average shooter against a league-average goalie, so that a player's finishing (goals minus xG) measures their shooting talent against that neutral bar.

That choice is the whole design. A model that already knows Auston Matthews is shooting cannot tell you how good Auston Matthews is at shooting — the talent gets absorbed into the prediction and disappears from the residual. Keeping identity out is what keeps the residual meaningful, and it is still true of the model scoring this site today.

Updated 20 July 2026 — which model scores the tables. Until now it was the v1 model described on this page. The replacement was the identity-blind structural half of Model v2's joint fit: the same talent-blind contract, but with the structure re-estimated net of player talent instead of absorbing it. That matters for exactly the reason this page is about — a baseline fit without identity quietly soaks up whatever talent correlates with its features (skilled shooters over-represented in good locations inflate the location terms, which then under-credit those shooters), and the joint fit measures the two in tension so it does not. The scoreboard is still talent-blind; v2's player-aware composite remains out of the tables, because you cannot measure a shooter against a yardstick that already knows who they are.

Updated 25 July 2026 — the cascade came back. The canonical model is now nested_structure__nested_structure_xg__d2ec2662, the nested structural export of the same joint fit: three chained questions — does the attempt get through? → is it on net? → does it go in? — with each stage's structure re-estimated net of that stage's own people terms (shooters, goalies, and the blockers in front of them), then composed with every player term left out. Same talent-blind contract, better chassis: each attempt is scored by the stages that actually decide it, and the per-stage breakdown returns to the charts below. The fitted artifact was rebuilt once since (August 2026) on corrected blocked-shot origin imputation — a data fix, not a design change. Model v3 is the full account: why the cascade won, and the decomposition that inverted the reason it was built. Everything below this line — the training discipline, the segregated regimes, the held-out scorecard — describes the model now in use.

How it is trained

• Same-era, grouped by game. Evaluation uses K-fold cross-validation split by whole game (a shot and its rebound never straddle the split), with train and test drawn from the same era — the game changes over time, so we never train on old seasons to score new ones.
• Two prediction columns, never conflated. Out-of-fold predictions are the honest source for performance; the in-sample final model produces the canonical xg used for aggregation (uniform scoring across every shot). The tables on this site are the in-sample canonical column — labeled as such, not a held-out score.
• Segregated regimes. Empty-net and penalty shots are scored by dedicated models, so their potent goal environments never contaminate defended-net predictions.

Honest limitations

• xG is a model of chance quality, not a verdict — it is calibrated in aggregate, but any single shot's number carries uncertainty.
• Shooter/goalie identity is not in this model (by design — that is what makes finishing measurable). The talent-aware variant is its own post.
• Coordinates are arena-adjusted but imperfect; blocked-shot origins are imputed.

This build

• Model id: nested_structure__nested_structure_xg__d2ec2662 (nested_structure)
• Season shown: 2025-26
• Displayed xg: in-sample canonical column (aggregation)

Every number in this section is out-of-fold — each shot scored by a model that never saw its game in training — over 2,344,674 shot attempts (112,235 goals) across the full corpus. This is the honest held-out report, not the in-sample column the tables use.

The matchup engine (matchup__5184e7b0) predicts game winners from the xG pipeline. How it works, in one breath: every team carries ratings — opponent-adjusted, league-centered expected-goals rates per 60 (5v5 attack & defense, special teams, discipline, finishing & saving residuals), updated game by game with a 40-game half-life and shrunk ⅓ toward average each summer. A Poisson regression turns tonight's two rating vectors + rest/schedule features into expected goals for each side (λ), anchored to an as-of league scoring-environment offset so the level tracks the era rather than lagging it; the finishing-and-saving signal is reinforced within the season by the player-identity model's own monthly-refreshed conversion ratings. A score grid over all possible finals converts the λs into win, tie, and total-goals probabilities — with a fitted Dixon–Coles adjustment correcting the independent-Poisson grid's undercount of ties — and an OT/shootout layer settles the tie mass.

Every number below is from the walk-forward backtest: to score season N, the engine is fit only on seasons before N — it never peeks. Seasons through 2018-19 were the development window (used for tuning); the shaded holdout was never used to make a modeling choice. The Games page tracker is the third, strictest tier: predictions published before puck drop this season, scored as issued.