Close games
Winning tight games feels like a signature skill. The repeatability checks ask whether it lasts.
How it works
Separate team quality from tight finishes
Across 2022-23 to 2025-26, one-goal, overtime and shootout win percentage repeated much less than blowout win percentage or goal differential. After accounting for team quality, the close-game relationship was essentially absent.
Big moments are not all the same
Win probability added by a player's goals, per 60 minutes (late tying and go-ahead goals count most), repeats strongly, mostly because good scorers score. Chance share in big minutes repeats much less. A repeatable scoring contribution does not make every close-game record a lasting team trait.
The Ducks example
The Ducks went 24-12 in close games in 2025-26, about 18 percentage points of win percentage above what their play predicted. That describes the gap; it does not establish that the same edge will return.
The formula
The formula
Close-game win % = wins ÷ games in the one-goal/OT/SO sample. Compare that percentage year to year, then check what remains beyond team quality.
Goal WPA per 60 adds up how much each of a player's goals moved his team's win chance, per 60 minutes. Chance share asks who gets the chances in those minutes.
How we tested it
| Check | Finding |
|---|---|
| One-goal / OT / SO win % | r = 0.13 |
| Blowout win % | r = 0.59 |
| Goal differential | r = 0.60 |
| Close games beyond team quality | r = −0.04 |
| Win probability added by his goals, per 60 | r = 0.78 |
| Chance share in big minutes | r = 0.12 |
The strong WPA result mostly follows scoring ability. The close-game record itself provides little evidence of an extra, lasting edge beyond team quality.
What does correlation mean?
Correlation, written as r, describes how closely two sets of results move together. A stronger positive relationship means better results in one tend to go with better results in the other. A value near zero means little relationship. It is not a percentage of predictions that were correct.
What it can't tell you
These results cannot explain every close win or erase what happened. They tell us to be careful about treating a run of tight wins as a forecast.