Playing Possessed: Comparing player usage with possession through 9 games

No, nine games isn't much of a sample. I can't find a fuck to give. These are interesting to look at at any point, when it comes to perceptions of players.

The chart I'm presenting here is automatically generated from war-on-ice.com. On the X-axis we have pure Corsi For percentage. On the Y-axis we have each player's rate of shifts starting in the offensive zone. It's a graph we expect to slope up from bottom left to top right, because if you start in the defensive zone, your Corsi numbers are probably going to be worse; as Steve Burtch found in developing his dCorsi metric, about half of the variance in possession can be explained just as usage.

So with that in mind, guys in the top left of this are starting in offensive zone situations but giving up attempts anyway, while guys to the bottom right are starting in front of their own net and generating shot attempts despite that.

The colors of the bubbles correspond to Corsi quality-of-competition (QoC), which measures how strong your opposition has been in terms of Corsi, and the size of the bubbles correspond to TOI per game. The color doesn't have enough range, but it still shows what we want, which is extremes. QoC isn't the most useful thing ever, and is slightly skewed by the fact the Avs have two of their nine games against possession monsters the Minnesota Wild.

So without further ado, let's see the damn thing already. It's 5 on 5 data.

The Avalanche are a bad possession team, so seeing this many guys squished to the left is not a surprise. Strangely enough they do turn out to be a 50% zone start team. Go figure. So this thing is very squished vertically. Let's talk about some of the standouts here.

Dennis Everberg: Whoa. Okay. We're looking at an extremely limited sample on that guy, 5 games of low minutes, but he absolutely stands alone in a way that definitely should have inspired a longer look, in my opinion at least.

The top line of Matt Duchene, Jarome Iginla, and Jamie McGinn are getting crushed; however, they're starting in their own end more frequently than anyone else on the team and against some of the toughest opponents. If you want that line to score, don't start it back there so often.

Anyone directly above the Duchene line in this chart is therefore, relative to their teammates, getting shit on. Erik Johnson, Brad Stuart, and the forward triumvirate of Cody McLeod, Marc-Andre Cliche, and Maxime Talbot say hello. Especially those forwards, who are up against the weakest competition of the group; McLeod and Cliche actually have positive zone starts which is contrary to what I said on the podcast this week because god dammit war on ice I was looking at all situations rather than 5 on 5. Those five, and especially those last two, are getting steamrolled. Nick Holden is close to that group too.

The top left is generally Not Where You Want To Be, but that's where we find popular whipping boy Nate Guenin, complete with poor competition. Awesome! Also in the top left, Daniel Briere, and Alex Tanguay, though he is playing against the toughest competition of the three. These guys are being given opportunities in the attacking zone and ending up playing defense instead a significant amount of the time. Briere most egregiously. That's about 60% O-zone starts, and less than 45% possession. Dude. Can you veteran-lead the puck away from Varly, by chance?

Nathan MacKinnon sort of fits in here as well, although it's worth mentioning that his competition is tougher and he's been better since that awful line of Tanguay-MacKinnon-Briere was left to rot.

The next group worth mentioning I want to call the 50/50 group, because their zone starts and possession numbers line up pretty closely. John Mitchell, Ryan O'Reilly, Gabriel Landeskog, and Tyson Barrie meet this criteria, and Zach Redmond is fairly close too. It needs to be stated that Barrie's competition has been pretty bad comparatively though. So in case you're counting, that's two defensemen treading water in terms of puck possession, and one of them over a two-game sample that includes a roflstomp.

The Johnson-Stuart pairing was abandoned quickly, but they are in the same sort of usage and results territory here; Nick Holden, Jan Hejda, and half-games of Ryan Wilson have been slightly better--still not good enough--despite being used similarly (maybe easier competition). And Guenin is just hilariously over his head despite heavily sheltered usage. Fun times on the blue line as always.

So anyway my takeaways from this going into tonight's game against a San Jose team the Avalanche frequently struggle to possess the puck against:

• Duchene and Iginla are leading the Avalanche in scoring despite being matched up in their own zone against the toughs--usage we're calling the Stastny Zone. But unlike the aforementioned Stastny, they aren't driving possession really, so If you want more offense, they need to start farther from Varly. There's your new campaign, #FreeRedmonders: #UnburyDuchene.
  
• We knew the defense was bad, but dayum. Seeing it visually still smarts. How long are all these guys signed for again?
  
• Oh.
• If Guenin's bubble stays where it is and he remains un-fired into the sun, we've got a serious issue. At least he's been scratched recently, which is a positive step. McLeod/Cliche/Talbot are next in line.
  
• It's time for Briere's ice time to drop significantly, which means we can expect his contract extension any hour now. Guys in the top left need to have small bubbles to minimize their impact. He doesn't. Sadly the same looks like it might go for Tanguay too, but given he started the season coming off a really lengthy injury, I'm willing to give him a little more time before we start saying, okay, he's not getting that step back. At least we've been seeing Briere's role decrease already.
  
• Fans expected O'Reilly's line to enter the Stastny Zone for a reason. Let's share the load between his line and Duchene's. I know we're trying to shelter MacKinnon. I know. But this team needs to outscore its problems, remember?