Getting Shots Through

Number one is not missing the net. I think with a lot of these scorers, like the Heatleys and the Kovalchuks, they don’t miss the net that often. Most pucks cross the goal line at 65 miles per hour, so there ends the myth about the hard shot. How hard the shot is isn’t as important as the puck on the tape and off the tape and on the net, not trying to find holes.

– Tim Burke, San Jose Sharks Director of Scouting, as quoted by Shane Malloy in The Art of Scouting

I started reading Malloy’s book last night (I’ll have more thoughts on it further down the line) and one of the quotes that really jumped out at me was Burke’s comment above. Obviously, it’s important to get shots on net, but is getting shots through really vital for a goal-scorer? Couldn’t a goal-scorer simply compensate by firing more pucks, despite a high number of misses and blocks? I decided to double check whether there was a strong relationship at the NHL level between goal-scoring and getting shots on net.

Before I get to the numbers, I want to make a couple of points: first, Burke’s a very bright hockey man and second that he isn’t alone in this sort of thinking. On the latter point, here for instance is what legendary scout Marshall Johnston said for the same book:

“The biggest difference in goal scorers is from continually missing the net to getting shot after shot on net (Mike Bossy never missed the net – including in warm-ups).”

As for Burke, he’s been the top amateur scout in San Jose for 14 seasons, and the Sharks drafting record during that span has been very good. Prior to that, he was the top professional scout for the same organization, he coached for both the Devils organization and Princeton University, and before that he was a professional and college player. It’s impossible not to respect his track record and credentials.

Still, what I found is that while there is a relationship between goal-scoring and getting the puck through to the net, it’s a rather weak one.

My methodology was simple: I took every NHL regular who missed minimal time this season (I set my cut off at 75 games played, which left me 187 players to use) and ranked them by shots per game, goals per game, shooting percentage, and percentage of shots on net [shots on net/(shots on net + missed shots + blocked shots)] and then ran correlations comparing the rankings. A perfect relationship is equal to 1.00

Correlation between shots per game and goals per game: 0.816
Correlation between percentage of shots on net and shots per game: 0.631
Correlation between percentage of shots on net and goals per game: 0.448

It’s an interesting trio of numbers. There’s a very, very strong connection between shots per game and goals per game, showing what is perhaps obvious: players that get a lot of shots generally get a lot of goals. Of the league’s 20 best players in terms of shots per game, none has less than 20 goals (the worst total of the group belongs to Florida’s David Booth, with 23).

There’s a reasonable connection between percentage of shots on net and shots per game, which again makes sense: players that get most of their shots through are going to end up with more shots than those who don’t. Still, it’s a weaker connection than the one between shots on net and total goals, and several of the league’s best players this season at getting their shots through (notably Henrik Sedin and Shawn Thornton) had rather low shot totals.

Finally, there’s a relatively weak connection between percentage of shots on net and goals per game, and some notable players were very bad at the former and very good at the latter. Three of the league’ five best goal-scorers this season – Corey Perry, Steven Stamkos and Ryan Getzlaf – were not especially good at getting shots through. Stamkos was well below average, while Ryan Kesler was in the bottom quarter of the league in that category.

There is one last correlation I decided to run, since it seemed an obvious leap. Shooting percentage is generally regarded as a decent measure of shot accuracy, and in theory at least players with good accuracy should be the ones who get a high percentage of their shots through to the net. Yet, the correlation between percentage of shots through and shooting percentage was practically nil: 0.038.

It’s worth noting that both of the quotes above could be read as arguing the importance of getting a lot of shots – and that’s a conclusion I can wholeheartedly agree with, based on the data from this season (and more esoteric reasoning, such as the basic common sense idea that more shots = more goals). But I would suggest, based on this data, that keeping the number of missed and blocked shots to a minimum is of relatively low importance for the average NHL shooter.

Comments (5)

  1. Interesting read. One thing that jumps to mind though, is that maybe at a lower level (notably juniors and international), the correlation would be greater, and despite the flawed use of heatley and kovy as examples, his true intent was to refer to the amateurs that he scouts on a day to day basis. Maybe to jump to the higher levels as a goal scorer it is vital that at the junior level they are getting an abundance of shots, and getting a large percentage through to the net.

    Regardless, interesting read. Thanks

  2. @ Joe:

    My inclination is to think that things would be the same in the NHL and at the junior level, but it’s certainly possible that isn’t the case.

    I was actually a little surprised to see this data at a crossraods with Burke’s comments.

  3. This is interesting. Where did you get the attempts/blocked statistic? I know they’re counted on the NHL Event Summaries but they aren’t in the Real-Time category on NHL.com.

  4. “Most pucks cross the goal line at 65 miles per hour, ”

    Given the amount of garbage goals, dekes, and backhanders, my guess is that average is at least 20-30 MPH higher than the real figure.

  5. Nice article! I’d love to see this methodology applied to more than one season. A larger sample size would lend a lot of credence to your premise.

Leave a Reply

Your email address will not be published. Required fields are marked *