Do NHL goalies get hot in the playoffs? A multilevel logistic regression analysis

TL;DR

This study tests the hot-hand theory in NHL playoff goaltenders using multilevel logistic regression, finding no evidence of hot streaks and some evidence of negative effects of recent success on save probability.

Contribution

It applies a multilevel logistic regression approach to NHL playoff data to rigorously test the hot-hand hypothesis, controlling for various shot and game factors.

Findings

No evidence of hot-hand effects in playoff goaltenders.

Recent good performance may negatively impact save probability.

Permutation tests rule out regression to the mean as an explanation.

Abstract

The hot-hand theory posits that an athlete who has performed well in the recent past performs better in the present. We use multilevel logistic regression to test this theory for National Hockey League playoff goaltenders, controlling for a variety of shot-related and game-related characteristics. Our data consists of 48,431 shots for 93 goaltenders in the 2008-2016 playoffs. Using a wide range of shot-based windows to quantify recent save performance, we find no evidence for the hot-hand theory, and some evidence that good recent save performance negatively impacts the next-shot save probability. We use a permutation test to rule out a regression to the mean explanation for our findings.