Recently, as we try to get our NHL model put together in time for the playoffs (which don’t have a start date but could have a start date next week?), we’ve been looking at various relationships between shots and goals between the 2015-16 and 2019-20 NHL seasons. Five seasons worth of data, and so far what we’ve found is that the number of shots a team takes in a game doesn’t correlate to the number of goals a team scores in a game, but that the number of shots a team takes over a season does loosely correlate to the number of goals a team scores over a season. This is helpful for us to know, because what we’re really trying to do is put a number or numbers on how good a team is, and to put a number or numbers on how high-scoring a team’s games are, and knowing these relationships helps us determine whether to use shots, goals, or a combination in determining those numbers, and whether to split things by defense and offense or just go with one big number.
Today, we looked at the relationship between a team’s shots and goals and their opponents’ shots and goals, all over the course of each season (we’ll likely go game-by-game next, then try to build the system and hope we can actually get it to work this time, unlike our frenzied attempt in January that got real messed up). What did we find? Well…
| Relationship | R-Squared |
| Shots v Goals | 0.373814 |
| Shots v Opponent Shots | 0.202677 |
| Shots v Opponent Goals | 0.068634 |
| Goals v Opponent Shots | 0.074577 |
| Goals v Opponent Goals | 0.001691 |
| Opponent Shots v Opponent Goals | 0.462255 |
We found that we didn’t find much. Over the course of the season, a team’s goal-scoring and shot-taking tendencies don’t have much bearing on those of their opponents, with the slight exception of shots and opponent shots. Which goes to say that shot volume and scoring in the NHL probably don’t exactly work like, say, tempo in college basketball, in which what one team does directly impacts the other. At least, over the course of the season. Again, we’ll probably look at individual games next, and those R-Squared values aren’t zero, so we aren’t closing the door on that relationship entirely. We’re just recognizing the absence of an obvious, direct relationship.
For what it’s worth, a team’s opponents’ shot volume appears to be more directly tied to the opponents’ goals scored than is true of the team’s own shot volume and goals, but that might be unique to this five-year sample. The values aren’t far off.
