The Penguins and shot quality: a look at the individual advanced stats

The Trib had a couple interesting quotes on shot quality:

A more contemporary statistic Sullivan is less hesitant to put weight into is shots. On Saturday, the Penguins were even with the Blackhawks in shots on net at 31. That marked the first time in 11 games they had not outshot an opponent.

“We look a number of different statistics,” Sullivan said. “The shot (total) is one of them. But there are a number of statistics that, together, tell the clearest story of how the game is played. The shot (total) is an indication of where the game is played. So if teams are racking up a lot of shots, usually it’s as a result of them controlling territory.

“And we’re the same way. We’re looking at shots in conjunction with scoring chances. Then we break down specifics, Grade A chances versus Grade B chances. So there are a lot of other things that we look at, not just the shot (totals). It does tell you something but not the whole story.”

The story through 17 games this season has the Penguins overwhelming their opponent in the shot battle 13 times. They have underwhelmed in that area three times. Saturday’s contest was the first occasion they simply “whelmed” the opposition.

“Shots can be a little deceiving, sometimes,” said forward Bryan Rust, who has averaged two shots per game in six contests this season. “But not all the time. Usually, if you have more shots, you’re usually controlling the play. But a lot of teams might shoot the puck from the outside and things like that. We like to measure the game by scoring chances and kind of where the game is being played. If we have more (offensive) zone time, that will indicate we played pretty well.”

Sullivan has spoken more of this before as far as “Grade A” chances versus other ones. With Grade A being the prime, high danger chances with a good chance to score on. The Penguins have been great, as mentioned, at overall controlling the territory of where games have been played. Their 53.51% Corsi For% at 5v5 so far this season ranks second in the league. That only looks at pure volume of shots though, and would rank a goal the same event as a soft point shot that gets blocked, so it doesn’t much speak to quality.

If you look at xGF, expected goals, (and a refresher on how that’s configured can be found here), Pittsburgh leads the league at 5v5 with a 55.45% of expected goals, according to Natural Stat Trick. Their scoring chance for% is fourth in the NHL at 53.52%. Their high danger scoring chances are first at 56.74%. This is what Sullivan is talking about when he’s citing the context of “Grade A” opportunities the team is liking tracking, though every NHL team is unique and fairly hush-hush about their exact process.

Still, the quality chances that are being commonly tracked are showing that so far the Pens have been a dominant 5v5 team early in the 2019-20 season.

Which players are looking the best individually? Let’s take a quick peak:

Corsi For

This is tracking the raw rate of Corsi Events FOR the Pens every 60 minutes. So while Sidney Crosby is on the ice, the Pens are have 68.31 shot attempts per hour. With Erik Gudbranson, it was only 44.34 shot attempts. That sort of makes sense, as does this list — the best offensive players (Crosby, Evgeni Malkin, Kris Letang, Jake Guentzel) tend to drive the most offensive opportunities. Weaker players that spend more time playing defense (fourth liners, third pair players) don’t generate the volume of offensive events. Pretty logical stuff.

Corsi Against

Same chart as above, just showing the opposite — the number of shot attempts that opponents have per 60 minutes. So here we see that a player like Malkin is still seeing a lot of wide open action while he’s on the ice; the Pens are shooting a lot, but so too are the opposition. Fourth line players like Zach Aston-Reese, Teddy Blueger and Brandon Tanev have done a really good job at suppressing the opposition. For this metric at least, the Marcus Pettersson / Justin Schultz pair has been playing pretty loose and seeing a lot of shots fly at their net. The Brian Dumoulin / Kris Letang group has been the stingiest pair per/60.

Corsi Differential

I created this look to simply subtract the CF/60 minus the CA/60 to show which players are spending time in which zone, generally speaking. Since the Pens are the No. 2 Corsi team in the league, this is showing what should be expected: pretty much all the players are in the positive here, which is a good thing. Even better, most of the negative players are the AHL call-ups who have already been returned.

This is again showing that Pettersson (who is playing almost 17 5v5 minutes a night this year, up from 15:30 last season) might be playing a bit too much against a bit too strong of competition right now. And while it is nice that Jack Johnson, who doesn’t often look good in advanced metrics, is showing a slightly positive Corsi differential, it’s still an indictment on his level of play that he’s playing just under 15 5v5 minutes this year. It was 16:06 last season.

The efforts this season have been to use Johnson more carefully and less, and the results have helped mask him and not allowed him to hurt the team too much. Having a partner who can move the puck in rookie John Marino often carry him helps too, had Gudbranson still been around the third pair would be facing disastrous results.

***

With the above charts being fairly basic and not showing the concept of shot quality, let’s take a peak into the other perspectives that try to cast a little bit more detail.

xGF Differential

I’ve cropped things down and combined the above three Corsi charts and put them all in one view for xGF in the interests of brevity. This look really shows why the Pens like the Aston-Reese-Blueger-Tanev line and have gone to great lengths to keep them together, even as other injuries or healthy scratch decisions change the other lines. Jared McCann also has been driving great results so far this season. And it’s interesting to see Alex Galchenyuk (who hasn’t been rewarded yet with a real goal) look about as good per/60 as Guentzel has. One would hope that means a payoff is soon coming for Galchenyuk.

Scoring Chance differential

It’s a bit surprising that Bryan Rust (who has scored 5 goals and 7 points in 6 games) is towards the bottom of the heap here, but that’s a good reminder that this look, especially rate based, is only showing one piece of a bigger puzzle to consider.

Usage again matters a lot, as we see the third pair defenders a bit better than the Pettersson/Schultz group who is playing more minutes and against better competition but paying for it by providing worse results. John Marino will be a fascinating individual to watch — will he hit a “rookie wall” and have his performance drop? Or can he continue to grow even stronger (his 5v5 minutes already suggest he’s become the No. 5 defenseman on the team to pass Johnson) and help take some of the load off the second pair? Johnson’s results here, to be fair, are good to see. Considering contract and experience that’s still not a great investment but JJ was at 27.61 Scoring Chances Against/60 last year. He’s cut that by over 7 scoring chances against per hour so far in the early sample of the young season. Keeping that up would obviously be very helpful.

(Schultz, btw, was at 28.48 SCA/60 last year, and even in Schultz’s career 2016-17 season he was one of the worst on the team at suppressing scoring chances at 25.81 against per 60. It can be debated how big of a factor he is in this, but again this year he’s at 26.8 SCA/60, showing that the other team is going to get chances while he’s on the ice. That’s another interesting story to keep track of, especially with Letang out injured at the moment.)

High danger scoring chance differential

Another interesting look, I’d be curious if the Pens’ internal numbers showed Sam Lafferty as the worst here like Natural Stat Trick does (which, I’m guessing they don’t). Overall though, this is a pretty picture for the Pens with pretty much all their regular players as being on the ice for more positive things happening than negative at 5v5. Again, that’s kind of logical when you remember in general so far this year the Pens have been strong at 5v5 so far.

And these numbers aren’t the be-all end-all, no one stat is. The team probably doesn’t care that Rust is in the negative here when he personally is finishing a lot of those HD chances for and turning them into goals. It’s probably another debate about how sustainable that can be for him to keep converting, and certainly things will normalize soon and hopefully for him the amount of chances he concedes will also regress downward.

Final takeaways

The great thing about this data, there’s a lot to digest. Here’s a reader’s digest version of what we’re seeing

• The Pens overall are playing very well at 5v5
• Crosby-Guentzel + Dumoulin-Letang grouping has been especially dominant and driven tremendous results, which they have been doing for years
• There should probably be a bit of worry at the under-lying numbers of the Pettersson/Schultz pair, particularly an on-going trend for Schultz not being able to suppress a lot
• John Marino is certainly one of the most important players on the team and has salvaged a third pair. The improvement of Marino over Gudbranson (the expected third pair defenseman) can’t be under-stated.
• Jared McCann has some very impressive numbers too, as do Zach Aston-Reese and Brandon Tanev. These recent additions over the last 12 months to help remake the bottom-six forward group have been paying off big time.