How to Measure Your Hockey Games

Did I have a good or bad game?

Nov 21, 2021

When looking at game-to-game performance, ‘big picture’ stats like goals & assists can be misleading. While they are (obviously) important, they do not paint the full picture. There are a few reasons for that, including sample size, luck, and trusting results over process.

The Toronto Maple Leafs start to the 2021-22 season is a great case study for this dilemma. They started the season 2-4-1 start, but were leading the NHL in many key areas on October 27th, including:

Shot attempts
Slot shots
Cycle chances
Rebound chances

And on an individual level, key players such as William Nylander & Mitch Marner were at the top of the dangerous chances created-for and other key micro-stat charts. The Leafs ‘regressed back to the mean’ in a good way, and ripped off a dominant five-game winning streak.

Process Focus

Scott Gomez was on the Real Kyper & Bourne podcast and was talking about his time with legendary GM Lou Lamoriello and the New Jersey Devils.

Greg Revak, CFP® @CoachRevak

Scott Gomez from interview with @RealKyper & @jtbourne “Sooner or later it’s going to go in… Even If we didn’t get a point that night, we always go by our chances “It’s when you aren’t getting the chances, that’s when you start getting nervous All about a player’s PROCESS

He discussed the process to attaining consistently great results. Sometimes you do the right things and bad luck gets in the way. But as we all intuitively know, doing the right things over and over lead to favorable results (e.g. law of averages, regression to the mean, etc.).

Further reading - Evaluating your process rather than your results

Micro Statistics

Micro statistics are a much better way of measuring game performance and process. They can also do a much better job at showcasing which players are likely to continue to produce fruitful results in the future, which are likely to break out of a slump, and which players are due for some regression.

Some micro-stat examples:

Chances for/against
Battles won vs lost
Controlled Zone Exits
Denied Zone Entries

Game score

From these micro-stats, we can get a better idea of how players performed in given games. This is particularly illuminating for players that may not be large point scorers.

These game scores can be transformed into something intricate, like expected goals (xG), or broken down into simpler metrics, like turnover ratios.

Over in Sweden, @BetterThanAMonkey does a game score for SHL games. Each category is scored, weighted, then totaled to give a picture of how each player performed.