Hockey Graphs’ Ryan Stimson and Corey Sznajder have done an upsetting amount of manual data collection over the last few years. Building off foundational research by Sznajder, Eric Tulsky, Geoffrey Detweiler, Robert Spencer (PDF), and Jen LC, Sznajder and Stimson have been watching games to record information about tactics that just doesn’t exist in the standard NHL data sets that drive so-called “advanced stats” like shot-attempt percentage, expected goals, and PDO.
Stimson has published four years of data in an interactive application, called the Player Traits and Performance Comparison Tool. When I opened the tool, the first thing I did was look up some young Caps defensemen. What I found was fascinating.
First: background upon which we can all agree: the Caps defense in 2017-18 was not great. Here’s a Vollman usage chart that tells the story at a glimpse.
I’ve annotated it a bit. Top left is hard work – lots of faceoffs taken in the defensive zone against good opponents. Bottom right is a cushier gig – lots of opportunity for offense against weaker opponents. The size of the circle indicates how much ice time the player gets, and the color of the circle indicates whether the player saw the team do better in shot attempts (blue) or worse (red) when on the ice.
We see our reliable top-three at top left – Orlov, Niskanen, and Carlson – doing a tough job and doing it well. At bottom right we see our remainders. Djoos did well in sheltered minutes, Bowey got bombed in similarly easy minutes, and Jakub and Kempny did not moving the needle much near the middle.
Oh yeah, and Brooks Orpik, who, well, woof.
Now for Stimson and Sznajder’s data, which won’t tell us about what the team did when the player was on the ice, but rather how the player contributed to the team’s play. A lot of these statistics involve passing (who moved the puck to whom before a shot) and location (where the puck came from before a shot). These are new concepts, so I’ll regurgitate Stimson’s glossary. It’s broken down into five groups, and I’ll explain further below:
Each player is assigned a percentile based on his rank among all players, so higher is better. I’ll demonstrate by picking some random defenseman out of a hat that everyone everywhere agrees is good.
Ah, Nate Schmidt. Good choice. We see Schmidt is in the top 20-30 percentile in most stats, especially in passing, including cross-slot/behind-the-net passes (DZSA60) and expected assists based on pre-shot movement (ixA60). That lines up with our intuition of Schmidt as a play-making defenseman.
Now let’s pick – again at random – a defenseman who isn’t quite as universally acclaimed.
Ooh, Karl Alzner. Nice, random pick. While Alzner has some good showings in his secondary and tertiary passing (BuildUp60), setting up one-timers (1TSA60), and cross-slot/behind-the-net passes, Alzner lags in most of the other metrics, especially in shot creation and his personal involvement in the offense.
Now let’s look at some of those defensemen who struggled with the Caps this season, starting with Madison Bowey, who got demoted to Hershey in March.
This is surprising. While Bowey’s sample is very small (187 minutes compared to several thousand for others), there’s a lot to like here – especially in how involved Bowey gets in the offense (SCB% and xPrP60) and the quality of the shots he helps generate. Bowey ranks brutally in shot suppression (RelTCA60) and shot assists (SA60), but the rest is pretty encouraging. While I don’t think anyone was happy with Bowey’s rookie season, seeing how this data aligns with our previous characterizations of his game has me curious about how he might do next season, given different circumstances.
Now let’s look at the undersized Swede, who, like Bowey, spent a lot of time with Brooks Orpik. Turns out Christian Djoos is kind of a stud.
That’s excellent. Djoos’ sample is also very small (228 minutes), but he’s among the elite in a whole bunch of tactical stats. The only places he really falls short are in overall shot volume (Shots60) and one-timers (1T60).
Here’s where I start to wonder if Djoos and Bowey, who both shared time with the notoriously offense-shy Orpik, might be seeing some distortions given a) how small their samples are, b) how little attack time the Caps had during their shifts, and c) how reluctant their common D partner was to get involved in the offense. Still, with numbers that sterling, I’m very eager to see how they will perform on the blue line next season.
Since we’ve been discussing him so much, here’s Brooks Orpik‘s chart.
I don’t have anything to say about this that you haven’t heard before, but I thought it’d be fun to show two comparisons: Tampa veteran Dan Girardi and Columbus scratch/ex-Cap Taylor Chorney.
One of those players will earn $3 million next season; the other probably won’t get an NHL contract at all. How about that.
I recommend you explore the tool yourself. There’s all kinds of fun things you can learn, and it’s a blast to compare the styles of players involved in a trade.
Corey Sznajder has a Patreon if you’d like to support his tracking project and see his data. You can read Ryan Stimson’s work on The Athletic and Hockey Graphs. Rob Vollman has written a few indispensable books on hockey statistics, including Stat Shot and Hockey Abstract.
Headline photo: Cara Bahniuk
Russian Machine Never Breaks is not associated with the Washington Capitals; Monumental Sports, the NHL, or its properties. Not even a little bit.
All original content on russianmachineneverbreaks.com is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported (CC BY-NC-SA 3.0)– unless otherwise stated or superseded by another license. You are free to share, copy, and remix this content so long as it is attributed, done for noncommercial purposes, and done so under a license similar to this one.