Today's games

Win probabilities from the Poisson matchup engine (matchup__6a856b2b) — team xG-rate ratings → expected goals → a score grid. Every number below was issued before puck drop and is scored honestly against the result. How it works & why it can lose to a logistic.

Slate

2026-01-15

Matchup	Favorite	P(home win)	exp. goals	Elo
MTL @ BUF	BUF	55%	3.22–2.95	60%	final 5–3
PHI @ PIT	PIT	65%	3.23–2.34	58%	final 6–3
SJS @ WSH	WSH	62%	3.37–2.64	70%	final 2–3
VAN @ CBJ	CBJ	71%	3.52–2.21	70%	final 4–1
SEA @ BOS	BOS	63%	3.01–2.24	65%	final 4–2
WPG @ MIN	MIN	63%	3.10–2.35	64%	final 2–6
CGY @ CHI	CGY	47%	2.54–2.71	49%	final 1–3
DAL @ UTA	UTA	56%	3.02–2.69	51%	final 2–1
NYI @ EDM	EDM	59%	3.13–2.61	62%	final 0–1
TOR @ VGK	VGK	65%	3.20–2.33	67%	final 6–5

Click a matchup for its page — an upcoming game shows the predicted shot map and breakdown; a finished game opens on what happened, next to the pregame call. Every slate the engine has ever scored is browsable from the date selector. Or build any matchup in the lab — pick the teams, drag the rest sliders, and score it with the same engine.

Season accuracy tracker

Log-loss (lower is better) over 10 completed games this season, every prediction scored as it was issued. This is the public, unfalsifiable record — the append-only live file means we cannot re-grade a past call.

Engine	log-loss	accuracy	games
Poisson (shipped)	0.632	70.0%	10
Logistic (challenger)	0.632	70.0%	10
Elo (baseline)	0.678	70.0%	10
Home-rate (floor)	0.680	60.0%	10

The shipped engine is Poisson. Its backtest twin, a plain Logistic regression on the same ratings, edges it on winner log-loss by ~0.003 — but only the generative Poisson produces the expected-goals score grid needed to price totals and puck lines, so it stays the headline. Both beat Elo and the home-rate floor. Full walk-forward methodology on the Model page.