The purpose of this article is rather ambitious. I aim to provide a sparknotes version of the Devils Season with regards to advanced statistics. I’m going to give you guys what I consider to be some of the most compelling storylines. I’ll tell you what things are unanimous, contested, and totally unknown. I will also provide plenty of links to data and sources that you can investigate on your own.
I recognize I’m casting a wide net here, but I truly think there will be a little something for everyone. I’ve split this up into Beginner, Intermediate, and Advanced Statistical information. The first two sections use a dataset that is linked below which was genreated from Corsica.
Advanced Stats For Dummies
The chart below does not use advanced stat terminology and is designed for those who have little to no exposure to the field. I am somewhat diluting the actual meaning of some stats, but I believe the integrity of their implications remains mostly intact. If you disagree, say so in the comments. For stat junkies, the actual names of these stats, by row, is Player, Pos, GP, TOI, P1, ixG, iSC, CF%, xGF%, PDO-xPDO, CF%Rel, ZS%Rel.
Depending on your device, that chart may be too small to be visible. If that is the case, you can access all of the tables from this article in their original form with the original raw Corsica data via this link.
Taylor Hall is good no matter how you assess him. This supports the conclusion from our awards that he was the best player on the team this year.
Beau Bennett is the possession leader of the team. According to the data his time on ice has the highest percentage of Devils possession.
Yohann Auvitu was extremely sheltered with massive numbers of starts in the offensive zone, and that led to his more favorable numbers that people often cite and may have also contributed to the perceived good puck luck.
Adam Henrique actually had more expected goals than Kyle Palmieri despite scoring 5 less at even strength. That means Adam Henrique was in a position to score just as often, but Kyle buried them. Kyle also had the best luck of any player.
Devante Smith-Pelley is the worst player on the Devils.
Travis Zajac tied Taylor Hall for the team lead in Primary Points at even strength.
Middling Advanced Stats — A Summary
This one is for those who have at least casual exposure to advanced stats in hockey.
Taylor Hall is really good.
In his half a season here, there is really very little bad to say about Stefan Noesen. In particular, his Scoring Chance rate was second only to Taylor Hall.
Miles Wood is great at the sexy individual stats (in Green). In particular, his expected goal rate was second only to Taylor Hall (Noesen was 3rd). However, the advanced ones paint him as one of the worse players this year with terrible possession numbers and slightly below average expected goal percentages. A possible narrative is that he is great on the puck, and terrible off the puck.
John Moore’s 12 goals this year were kind of ridiculous. Of those, 11 were even strength, but he only had around 4 expected goals.
Zacha is not a top 6 guy yet. After adjusting for zone starts, score effects, and arena (in Red), he was one of the worst possession players on the team. Unlike Wood, though, he didn’t make up for it with his on-the-puck skills. He was below team average (which isn’t high already) on expected goal rate and scoring chance rate.
Advanced Stats and Visualizations
Check out Micah Blake McCurdy’s HockeyViz site for some really interesting visuals. Here’s an example that shows the tendency over the year to switch Severson out of the top pairing in favor of Lovejoy.
WOWYs are a must-have for the budding hockey analyst. It is possibly the single largest piece of contextual information for a players possession possession stats. HockeyAnalysis did it first, but Corsica’s Combos section has a WOWY applet that is pretty great too. You can use it to show stats like this which shows how Palmieri’s numbers were greatly aided by playing with Zajac and Hall.
Some Other Stats Out There
SIT which weighs production for “situation.” This stat likes Noesen, Zacha, and Palmieri more than traditional stats. It also loves John Moore.
XtraHockeyStats has their own expected goals model called EGF. It likes Auvitu and Zacha more than Corsica’s xGF model. It is down on Santini and it is WAY down on Moore.
Hockey-Reference has their more surface-level, but very clean Point Share stat. It labels Palmieri tha MVP among skaters by actually quite a bit over Taylor Hall (2nd). It also has John Moore contributing more offensively than every player other than Hall, Palms, and Rico. He outranks Severson in both stats, and is the highest overall defender.
Watch for Rob Vollman’s 2016-17 Data. It is traditionally very extensive and provides info on just about every stat you could hope for.
It is nearly unanimous that Taylor Hall was the best player on the team. It is pretty close to consensus that Zajac, Palmieri, and Henrique were next. Miles Wood, Stefan Noesen, and Zacha are the most interesting/conflicting to me. Noesen is pretty consistently viewed positively — I couldn’t find any compelling reason to discount his success. Wood is a basic stat darling, but an advanced stat reject. Next year will be very interesting for him. Zacha has some favorable stats, but it’s clear he has not yet made the jump.
Severson, Greene, Merril were our best defenders. Santini is right with them on a per-minute basis. John Moore is all over the freaking place. It all comes down to how much you value his goals and if you think it was him, his situation, or his luck that caused it. Defensively he was probably the worst on the team.
DSP is the most unanimously bad player on the team. Guys like Lovejoy, Fiddler, Lappin, and Kalinin have comparably bad stats, but have tragically low zone starts that likely was not able to be fully accounted for by traditional adjustments. Coleman is also probably not an NHLer.
If you have any questions, comments, criticisms, or other stuff then leave comments below. I hope you guys enjoyed!