Despite the fantastic start, I’ve been curious about what I perceive to be a troubling inability for the Islanders simply to put their shots on the net. (And judging from Howie’s commentary and some comments in recent threads, I’m not the only one wondering about this.) Sure they’re scoring at an excellent rate, but obviously they’d be doing even better if more of their shots were forcing saves and generating rebounds.
But are the Isles really as bad as I have made them out to be in my head? Or is my hyper-critical homerism shielding me from the fact that every team is similarly saddled with this same apparent affliction? We’re all psyched about the Isles’ possession metrics so far this season, but from this fancystat-neophyte’s perspective, all the Corsi and Fenwick in the world won’t help if most of their shot attempts whiz harmlessly past the cage. (I recognize that many shot attempts miss the net intentionally, of course, either intending to be redirected or hoping for a nice ounce off the boards. But since there’s no way to analyze intent, we’ll just go with what the shot stats tell us.)
A disclaimer: I am by no means a statistician and have only recently dipped my toe into waters of modern analysis. It’s very possible that others have done this same sort of study elsewhere. But not having seen shot accuracy studies elsewhere, I decided to see if I could draw some conclusions on my own. And now that we’ve reached the season’s quarter pole, this seemed like as good a time as any to share the results. (I’ll update the numbers after each successive quarter.)
Oh – to NDRE and garik and the other analysis experts out there: be gentle… this is my first go at this. I welcome any and all feedback.
Do the Isles Really Suck at Getting Shots on Net?
I define shot accuracy as Shots on Goal (SOG) divided by Fenwick For (FF). I didn’t use Corsi because Corsi numbers include blocked shots and you really can’t gauge the accuracy of a blocked shot. Please note that this analysis has nothing to do with whether the puck actually goes in the net – I’m simply focusing on how often they’re putting the puck on the net. I used the stats from the hockeyanalysis.com site, which depicts only 5-on-5, 5-on-4, and 4-on-5 situations (sadly, no 4-on-4).
Anyway, the results were, for the most part, just what I suspected. As you’ll see below, the Isles really do suck at getting shots at the net.
But does it really matter? Let’s explore that together.
We’ll look at 5-on-5 situations first:
The Isles rank next-to-last at 69.3%, about 4.15% worse than the league average. In other words, they can’t hit the broad side of The Barn. But the one team shooting worse? The defending Stanley Cup champions (who obviously had a bit of a sub-par first quarter in standings-wise).
It’s interesting to see the degree to which a team’s shooting accuracy correlates – or doesn’t – to wins and losses. The current top seven teams by points (Montreal, Tampa Bay, Anaheim, Nashville, NYI, Pittsburgh, and Calgary) are actually scattered up and down the rankings, with only one (Tampa Bay) in the top 10.
An anomaly? Too small a sample size? Or is shot accuracy simply not that meaningful? Well get back to this question in a moment.
Next: same metric, this time showing how teams fare on the power play.
You can see that the Isles’ shot accuracy improves with the man advantage to 71.3%, which is right in the middle of the pack and just a hair better than the league average (which also rounds to 71.3%).
A close look at the two graphs above show something worth noting – league-wide, shot accuracy actually decreases on the power play (71.3% compared to 72.6% in 5 on 5 situations). I could give you a couple gut feelings on why this might be, but the truth is I’m really not sure.
Finally, we’ll total the data (5 on 5 plus 5 on 4 and 4 on 5 situations):
Clearly, the Isles’ improved shooting accuracy on the power play wasn’t enough to affect their overall relative position. In all, they’re finding the net on 69.3% of all their shot attempts, which is 3.25% below the league average. (Again, this does not include 4 on 4, 5 on 3, or 3 on 5 data.)
Ok, They Suck at This. So What?
Fair question. They’re currently second in the NHL in goals scored and third in goals per 60 minutes. So poor shooting accuracy apparently has little impact on goal scoring. Right?
I wanted to see just how impactful inaccurate shooting might be to a team's expected goal output. To do so, I extrapolated the Isles current FF per game (44.5) to an 82 game season. In this case, the Isles would take about 3,650 unblocked shots over the course of the season, resulting in 2,530 SOG given their current 69.3% shooting accuracy percentage.
Now, if their shooting accuracy matched the league average of 72.6%, they would end up with 119 more SOG over 82 games. And given the current league average of one goal for every 11.5 SOG, an Isles team with average shooting accuracy would expect to net 10 more goals over the course of the season.
And while there might not be a magic formula for translating goals into standings points, 10 more goals a season could have a huge impact in a division that appears to be as congested as the Metro.
Comparison to Last Year
In 2013-2014, the bottom-dwelling Islanders actually had a respectable shooting accuracy of 73.1% (including all three manpower situations), which was about 1.1% above the league average and good for eight in the NHL.
The three most accurate teams in the league last year were teams who had excellent seasons – the Avalanche, Black Hawks, and Penguins. But before you start drawing full-season parallels between accuracy and success, take a look at who finished dead last: Again, the Los Angeles Kings. (And by a substantial margin, too.)
Some Individual Player Metrics
I then looked at how each individual player fared in the shooting accuracy department. A few surprises emerged as well as a few… well, take a look for yourself:
For this table, I used 5 on 5 only, as 5 on 4 and 4 on 5 situations had too little data to draw comparisons. The forwards are shaded blue, the defensemen brown.
The first thing that jumps out is that the blueliners in general have far worse accuracy than forwards… which makes intuitive sense given than they’re generally taking longer and harder shots. But Thomas Hickey is the outlier here, with 15 of his 19 unblocked shot attempts being on target. Meanwhile, to nobody’s surprise Brian Strait lags behind everyone (though the sample size is extremely small).
As for the forwards, the red flags seem to be Tavares and Grabovski – let’s hope Grabbo can get his shot straightened out because his goals-per-SOG ratio is excellent thus far at 19%. On the good side, two-thirds of the kid line are proving themselves to be snipers, while the sample-size-challenged Josh Bailey isn’t far behind and has the team’s best goals-per-SOG ratio at over 21%.
See You After Game 41
So any thoughts on this? Does anyone out there see things in the numbers that my untrained eye does not?
I’ll run the numbers again after the Isles reach the midway point. I have no idea what constitutes a decent sample size in this department, so I’m very curious how dramatic the changes might be from quarter to quarter.