As Pat wrote earlier, how the Washington Capitals line up against the New York Islanders will factor largely in which team escapes the first round. Generating a little more offense and shutting down John Tavares during 5v5 are key.

With that in mind, I thought we’d revisit a fun visualization of how the Caps are most and least effective. I promise it’s pretty.

I first applied this idea, devised by Tyler Dellow, on the woefully top-heavy Oates!Caps. Then, twenty games into this season, I took a look at Trotz’s Caps, who were still pretty amazing at the time. Things have calmed down since then, but it’s still a great exercise.

Here’s how it works. You put your defense along the top and your forwards along the left. You rank each by ice time (I used all situations). Where each forward and defenseman meet, you show the percentage of the total shot attempts the Caps take when those two are on the ice during 5v5. Then you color-code them– from a luscious dark green of possession domination down to the putrid tomato red of getting shelled.

On a good, well-organized team, the most-used players (at top left) get strong possession and the least-used players (at bottom right), if they’re outshot at all, don’t hurt the team too badly. So that wonderful forest green should be at top left, and hopefully the bottom right isn’t too red.

Here’s how the Capitals look after 82 games. My observations follow.

Before I bust out the bullets, a note on reading this table: I suppose you could use it to look for individual chemistry– which two guys do well together. For me, I like to generalize a bit, and let my eyes glaze over a bit and look for the patterns.

Okay, let’s dance.

• The prettiest dark greens are not at top left, where most of the ice time is. Instead it’s at the top right– where the top forwards meet the third-pairing defense, and bottom left– where fourth liners and shutdown defense meet. Perhaps those players are getting easy shifts in the offensive zone against weaker competition– or perhaps those players are undervalued and underutilized. I think it’s both, but I suspect the latter is a big factor.
• Alex Ovechkin spends about 38 percent of his 5v5 time with Orpik and Carlson, during which he does pretty well– though not remotely as good as he does with Alzner and Niskanen or Mike Green. When the Caps get an offensive zone faceoff against the Islanders, they’d do well to put Ovi out with Green.
• This isn’t an “Orpik is bad” thing. That’s reductive and not really accurate and generally snide. One thing I find fascinating is the players who do better with the top defensive pairing (and presumably against the toughest competition). That would be Brooks Laich, who has done pretty terribly away from Orpik, and one of my faves, Michael Latta, whom I bet will be the heart of this team’s middle six in the next two years.
• If fourth liners like Latta and Wilson do so great with Orpik and Carlson (despite increased defensive-zone starts), but Ovechkin and Backstrom do better apart with the other defensive pairings, doesn’t this sort of suggest its own resolution?
• I think Mike Green is a tremendous player, but this visualization might offer criticism of him. When Green plays with checking-line players like Laich and Jason Chimera, presumably in a defensive-assignment role, the Capitals perform atrociously. He should be optimized for the attack, not defensive match-ups.
• It’s good to acknowledge the limitations of an exercise like this. For Curtis Glencross and Tim Gleason, we’re dealing with relatively small sample sizes. As a result, their numbers are a bit scattershot.
• Further, I wish I could find a way to combine the information up there with the amount (or percentage) of ice time the players share or the nature of their deployments together. It’d be neat to have a GIF or slideshow that switches from one dataset to another.
• I’m not sure how this works exactly, but I noticed that forwards tend to do better with the heavier ice time defensemen of each pair: possession is better with Carlson than Orpik, Niskanen than Alzner, and Green than Gleason. Maybe that is hinting at who’s carrying the water for each pairing.
• Andre Burakovsky‘s performance is a bit colored by favorable zone starts and sampling bias, but he looks to me like a decent top-six forward (and, one day, a true top liner). I’ll be sad if he doesn’t see any action against the Islanders. He makes the Caps better.
• I’ve been playing this tune for 18 months now, but I think these data provide even more evidence that Tom Wilson should be off the bottom line. And also off the top. Wilson does best apart from the extremes– with the neutrally used Alzner-Niskanen pairing. He also outperforms most of his non-Latta fellow grinders. With him, Latta, and Burakovsky, the Capitals might be able to surprise the Isles with youth.

Good times. I know I’ve been harping on this a lot, but the Caps have the potential to be even better than they are. In a few years, player development and some expiring contracts will see to that. But the Caps need to be better now if they wanna challenge for a Cup. Those are some of my ideas on how. What do you see?

Data by Hockey Analysis