A month ago, I published an article on Drive4Five about a statistic I created to measure an NHL team’s performance over a season: average game intensity.

The metric is comprised of three parts:

1. Average win intensity (AWI): This statistic is calculated by taking the games that a team wins over a season and summing the goal differential in those games. Afterward, divide the goal differential by the total number of wins during the season. On April 29, 2021, the Islanders had a +62 goal differential in their 29 wins for an AWI of 2.1379
2. Average loss intensity (ALI): This statistic is identical to AWI, except that it involves the games that the team lost. To date, the Islanders have a -40 goal differential in 20 total losses (including overtime and shootout) for an ALI of -2.
3. Average game intensity (AGI): This statistic aggregates AWI and ALI into a statistic that gives a better picture of a season’s performance. Subtract the absolute value of ALI from AWI to get this measurement. Thus the Islanders have a .1379 AGI this season.

As this statistic takes every game into account, it most likely a very accurate measure of performance. The higher the AGI, the better the team should be. That represents a team that dominates their opponents in wins and keeps it close in losses.

While the previous article delved into the Islanders’ historical performance on this metric, I wondered if I could use this statistic in other ways. Again, hockey-reference.com was the sole contributor of data for this project.

Here’s what I found

From the, albeit small, sample that I used, average game intensity was a better predictor of NHL playoff performance than regular-season win percentage. But, not in the way that I expected.

AGI and playoff performance

I found that NHL teams with better AWI, ALI and AGI numbers do well in the playoffs. When using historical AWI, ALI, and AGI data from the six Metropolitan Division teams that are currently in the MassMutual East Division, AGI had the lowest p-value with playoff wins. A lower p-value indicates a higher likelihood of correlation.

For a statistically significant relationship between two variables, a p-value below .05 would be necessary. AGI’s p-value with playoff performance was .00039446.

Playoff Performance and Regular Season Point Percentage

I also looked at the correlation between regular season point percentage, which typically indicates playoff seeding. Regular season point percentage’s p-value with playoff performance was .001020559.

Both of these numbers would indicate that their relationships with playoff performance are legitimate.

AGI and Playoff Wins

But something in the data is strange, average game intensity had an inverse relationship with the number of wins an NHL team would typically get in the playoffs.

So, while having a fantastic significance in the data, it was inversely related to playoff wins. A team that maximizes its victories and minimizes losses during the regular season did worse in the playoffs.

How can that be?

This could very well be due to a sample size issue. The data was collected over only 13 different playoff runs between six teams over the course of four years. One hundred samples would have elicited more accurate results.

But there are some more fun theories on how this negative correlation is possible with the high p-value.

A Hard-Working Regular Season Tires a Team Out

A team that is maximizing its AGI is both winning by a lot and losing by a little. This lends itself to a coaching style of never letting up. The team that’s already winning by three will try to score five. When they’re behind, they keep clawing for that comeback.

While this is indicative of a gritty team, this style of play may just wear them out. If a team pushes as hard as it can for 82 games, won’t they play below its potential come playoff time?

High Intensity Means Little Prep

If a team has a high AGI value, it may have much less to do with the coaching styles or individuals on their team. It may just be an issue of schedule strength. A high AGI could be indicative of a team that just had an easy schedule, leading to a lot of success in the regular season. Though, once they hit the playoffs, they’d get squashed by teams that had harder schedules and thus lower AGI numbers.

The data is a little confusing, but there seems to be something here. If you want to reach out to me about the project, you can reach out to me at [email protected].

[sendtonews key=”SlALljLt66-1293748-9119″ type=”single”]

John Smark