The Success Of Getting Shots Through

Often times you'll listen to hockey coaches, to players after games, to talking-heads, and they'll say a lot of the same things about defensemen in hockey today. One key component to playing D right now is being able to get a shot past the first-line of defenders. If your shot from the blueline gets blocked by a forward, it's a wasted shot. More often than not, that puck will bounce out of the offensive zone, causing the team to reset and re-enter the zone, or perhaps even turn into odd-man rush the other way.

The question then becomes, is getting shots through the first line of defense or net a skill? Is it a repeatable skill? Well, let's try to answer that.

My methodology here was pretty basic considering the resources (including my own knowledge) that I currently have on hand. I opted to look at success rates. First, the percent of shot attempts that were unblocked, and then the percent of unblocked attempts that were on net. 

All data provided by War-On-Ice.com, 5v5 play, defensemen who played in more than 30% of games in a given season from 2007-2008 through 2014-2015.

To see if this was a repeatable skill, I went to the ever trust-worthy boxplot! I wanted to use players that had enough data, so I narrowed it down to players who had played at least four seasons in the above time frame, and totaled more than 500 shot attempts. 

In order to avoid such a noisy graph, I first charted the top 10 (well, I went for 11 here to include Keith Yandle. Call it Rangers bias) players by success of attempts that are unblocked. Then, I went for the bottom 10 to see if they were consistently low. Figuring that may introduce some bias, I also ran a chart for a random sample of 20 players from the data pool. Here are the boxplots:

Top Shooters

Top Shooters

Top Shooters by Accuracy

Top Shooters by Accuracy

Bottom Shooters

Bottom Shooters

Bottom Shooters by Accuracy

Bottom Shooters by Accuracy

Random 20, unblocked success

Random 20, unblocked success

Same random 20, shots on net

Same random 20, shots on net

To be honest, I was definitely a bit surprised at some of the ranges a few of these players are operating at. You'd expect at both sides of the spectrum for the top players to be a little tighter, as well as the bottom players to be a little tighter as well. Of course, some of them are, proving that yes of course some players are skilled at this.

Another option I took to try and find consistency among d-men was to look at the standard deviation of the player's data. This is represented in the frequency chart below:

Surprisingly, most of the NHL d-men who met the criteria fall at or under a 5% standard deviation among their successes in shot attempts that are unblocked. Not so surprisingly however is that our tail leans heavily to the right, as we get to the few players who well, are a little more lucky than skilled (or just purely unskilled). 

This does seem to be a skill set for some players. The question now becomes, well, does it matter? Our top ten players in terms of consistency, among D who have played in four or more seasons of this data feature names like Matt Gilroy, Karlis Skrastins, Steve Eminger and Nick Boynton; but also including players like Wade Redden, TJ Brodie, Chara, Christian Ehrhoff, and Oliver Ekman-Larsson.

Meanwhile, the players with the highest standard deviations boast players like Derek Meach, Mark Eaton, Aaron Johnson, Cody Franson, and Mark Fayne, with no one who I'd call particularly great (though you could argue Franson) being featured in the bottom 15, outside of all-star [oh NHL, why?] John Scott.

The final test here was to see if there was a relationship between shot successes and primary points (goals plus primary assists). If there is a skillset to be had here, is it important?

No relation to speak of for either having a shot attempt unblocked, or having an unblocked attempt hit the net. I opted for a color variable of individual shot attempts to see if more data would be a sufficient way to see if these things even out over time. But really, what the color variable showed me is that the more shot attempts a d-man has, the more primary points they are likely to have, which is charted below:

A pretty clear positive correlation between the amount of shot attempts a d-man has, and the amount of primary points a d-man has. Which really has this conversation come full circle to the fact that this whole shot attempts (read: Corsi) thing is kind of important. 

It's not so much consistently getting your shots through the D, but firing enough shots that even if you don't get them through, you are creating offense.