Breaking down NHL Edge data: How important is shot speed and location?
A few weeks ago, the NHL finally unleashed its long-awaited player tracking data onto the public. Now it’s time to dissect what it all means.
Sometimes, all it means is “that’s cool!” — and that’s OK. Sometimes a fun little tidbit is all a stat has to be. But it’s still worth looking into how meaningful all the new numbers are and what bucket they fall into: relevant info or interesting trivia.
Advertisement
Each new stat poses its own interesting questions worth answering and that’s the goal of this series; diving into the new data to see how much it matters.
Let’s talk about shooting.
Shot speed
Is shot speed repeatable?
Average shot speed is not and should probably be ignored anyway — unless you really believe the 2021-22 New York Islanders, who have Ryan Pulock, had the lowest average shot speed over the last two seasons. We care about repeatability (comparing a stat from the previous year to the following year) because it shows how reliable a stat is. If a stat jumps all over the place from year to year, there’s less signal that what it’s measuring is real.
The other buckets do have some repeatability, though.
Shooting over 90 mph (and 100 mph) is very rare so it’s not surprising that it had the highest year-to-year correlation (rare skills are probably less likely to change teams). There’s an R2 of 0.58 for 90 mph shots per 60, and 0.59 for percentage of shots above 90 mph. An R2 closer to one means a strong relationship, while one closer to zero means the two datasets aren’t closely related.
As for the other two buckets, it looks like it’s better to combine them than to look at them separately. There’s higher year-to-year repeatability that way with an R2 of 0.39 for shots between 70 and 90 mph and an R2 of 0.17 for percentage of shots in that threshold.
Call it “very hard shots” for anything above 90 mph (the Washington Capitals and Tampa Bay Lightning led the way the last two years) and “hard shots” for anything between 70 and 90 mph (the Seattle Kraken and Vancouver Canucks led the charge last season, with the New Jersey Devils, San Jose Sharks and Toronto Maple Leafs having the lowest percentage).
Which shot speed group is most meaningful to offense?
While very hard shots are more repeatable, they seem to have less of an impact on offense.
There’s a much stronger relationship between hard shots and goals (R2 of 0.28) than there is for very hard shots (R2 of 0.10). Further, when it comes to year-to-year change, there is no relationship whatsoever between adding more very hard shots and scoring more goals. For hard shots, there’s a very small signal (R2 of 0.07).
Advertisement
Also worth noting: it’s most important to look at total shots per 60 within those thresholds rather than the percentage of shots within those thresholds. The relationship is much smaller there, which might suggest that the relationship between more hard shots and more goals could just be a product of, well, more shots.
Do teams with harder shots score more goals above expected?
It’s not exactly no relationship, but there’s some small signal here for hard shots (R2 of 0.06) when looking at the percentage of shots above that threshold. Nothing too enlightening, but enough to suggest it would be a helpful feature for an xG model.
Do teams with harder shots shoot more from the outside?
Interestingly, there is almost no relationship at all between teams that take harder shots and teams that take a lot of shots from the outside. That would suggest that teams aren’t prioritizing bombs from the point — they just have guys who can do so.
How does shot speed change in the playoffs?
It doesn’t. Not in any meaningful way. The averages for each threshold are essentially the same from the season to the playoffs. In terms of in-season repeatability, the relationship for hard shots is pretty strong (R2 of 0.34) and a bit weaker for very hard shots (R2 of 0.14). I would surmise that the rarity of very hard shots makes it harder to repeat in a small sample setting like the playoffs.
Shot location
Where do the most shots come from?
First off, I want to point out it’s an absolute shame that the shot locations aren’t separated by zone. Where shots come from will differ depending on manpower situation, as will a team’s finishing ability.
Despite that, it’s still something worth looking into. Yes, we already have xG models that utilize shot location and more to get a shot probability of every shot depending on manpower status. That’s not new and it goes back to 2007-08. Sometimes it’s nice to see it laid out by zone like the way the NHL visualizes it and it’s worth contextualizing that.
Advertisement
Unsurprisingly, the most shots come from right in front of the crease. Teams prioritize those locations because it’s the most dangerous part of the ice and over the last two seasons take 27.2 percent of their shots from there. Next up is the mid-range where 11.3 percent come from the slot and 9-to-10 percent come from the left or right of it. Add the crease (1.8 percent) and teams are taking 59 percent of their shots from the scoring chance area. There’s obviously a very good reason for that.
Where do the most goals come from?
Here’s the breakdown.
Net front: 48.1 percent of all goals
Slot: 13.9 percent
Left slot: 8.7 percent
Right slot: 8.1 percent
Crease: 4.3 percent
Only one other area is above 2.5 percent and that’s the center point (3.1 percent). There’s a reason this is universally known as the scoring chance area: it’s where 83 percent of goals have been scored over the last two seasons.
That’s not new information, but it is valuable to know the exact amount. It also pays to know how often shots go in from those zones and unsurprisingly those are the most high-danger areas.
Net front: 17.7 percent
Slot: 12.4 percent
Left slot: 9.1 percent
Right slot: 9.0 percent
Crease: 23.8 percent
Here’s what the two NHL shot maps look like with shots and shooting percentage mapped out with the percentage of shots and goals from each area.
And here’s which teams had the most shots from the most dangerous areas over the last two seasons. And which teams prioritized the most dangerous areas (SCP%, scoring chance priority percentage).
Is goal scoring by certain zones repeatable?
Rather than look at repeatability for 16 different shot zones, I decided to look at scoring chances and outside shots.
Scoring goals from the scoring chance area (R2 of 0.17) is more repeatable than scoring goals from the outside (R2 of 0.11). It’s even more telling when looking at shot selection weighted by the likelihood of the puck going in. Expected goals from scoring chances (R2 of 0.62) is significantly more repeatable than expected goals from outside shots (R2 of 0.30). Finally, the ability to score above expected is more repeatable for scoring chance shots (R2 of 0.18) than outside shots (R2 of 0.11).
Advertisement
Not surprising in the slightest, but it does show that getting to the scoring chance area is a repeatable skill for teams, and that they’ll have more luck finishing those chances than from the outside. Shots from there are less reliable and less likely to be finished with the same consistency as the year prior.
That brings us to scoring chance priority — the percentage of shots from the scoring chance area. With an R2 of 0.23 there’s some repeatability there, but a decent amount of variability as well. It means teams don’t have as much power to get to the most dangerous areas of the ice year-over-year.
Interestingly, the repeatability of high-danger shots (R2 of 0.35) and slot shots (R2 of 0.32) is much higher than when combining the two, for whatever reason.
How much does the side of the ice matter?
It seems to matter a little, but it could be just noise. Over the last two seasons, there were nearly 800 more shots taken from the left side (0.2 more per 60), but 107 more goals from the right side thanks to a slightly higher shooting percentage (6.2 percent to 5.8 percent).
That could be just random variation, but could also mean goalies are a bit more vulnerable to their left where the glove side is the short side. Seeing this kind of data broken down by shooter handedness would be neat.
How important is the middle of the ice?
The middle of the ice is key real estate, clearly and it’s where half the shots come from. It’s also where 70 percent of the goals come from.
The importance is best felt when looking at scoring from the slot and the point. It’s the same general distance, but there’s a clear increase in finishing ability. From the slot, moving from the left or right to the middle sees a shooting percentage jump of nine percent to 12.4 percent, a 38 percent increase. From the point, it’s 2.7 percent to 4.1 percent, a 52 percent increase.
Advertisement
Again, it’s not rocket science knowing the center of the ice is where goals happen and finishing ability rises, but it is nice to know how much.
How does shot and goal location change in the playoffs?
Games get tighter in the playoffs so I wanted to look at how shot locations showed that.
Net front: -0.9 percent
Slot: -1.6 percent
Left slot: -0.1 percent
Right slot: -1.0 percent
Crease: Same
So 3.6 percentage points of fewer shots from the most dangerous areas of the ice. Getting to those areas of the ice is more difficult, but it’s also worth noting that teams make it count more when they do get those opportunities.
Here’s the change in shooting percentage.
Net front: +0.7 percent
Slot: -0.3 percent
Left slot: +1.2 percent
Right slot: +0.4 percent
Crease: +0.7 percent
That leads to the same number of goals being scored in those areas, just on fewer shots.
With that in mind, it’s worth looking at the previous team chart above, with special attention to what changed come playoff time. Which teams were able to get to the most dangerous places on the ice to the same degree as they did during the regular season?
The Vegas Golden Knights may not have earned as many chances as usual, but they made them count. Same with the Colorado Avalanche the year prior who added chances all over the ice.
As for what teams can control most, the strongest relationship between season and playoffs comes from scoring chance expected goals (R2 of 0.37) followed by shot priority (R2 of 0.34). What actually goes in the net has the biggest influence on results, and that’s much less stable (R2 of 0.13).
Data via NHL Edge
(Top photo of Alex Ovechkin taking a shot: Scott Taetsch / Getty Images)
