Photo: Christian Petersen

Editor’s note: The snapshot should be up first thing on Monday morning. Instead, now that the Capitals are twenty games into the season, we’ll be running a series on how they look so far.

Barry Trotz is a huge improvement over Adam Oates. It’s not even close.

But as the season goes on, it’s becoming apparent that Trotz has chosen his favorite players and will place them wherever he wishes in the lineup despite evidence that it’s hurting the team. For all his many, many successes, Trotz is not exercising good evidence-based decision making in his distributing ice time.

Now that we’re twenty games in, this is a good opportunity to understand how Trotz is apportioning time, and how it’s working (or not working). To do that, I’m gonna use a visualization introduced by Tyler Dellow about one year ago that was later adopted a lot of places, including here, back with the top-Heavy Oates!Caps.

It’s not going to be pretty.

Along the left side of the table, I’m going to rank the forwards by their 5v5 ice time. Along the top, I’m going to rank the defenseman by 5v5 ice time.

The top left is where most ice time is– and hopefully the best players. The bottom right is where the least ice time is, and hopefully the weakest players.

In each cell I’ll put the Caps’ shot-attempt percentage when that forward is on the ice with that defenseman. Above 50 means the Caps are outshooting their opponents, but I’ll color code it so you can get the big picture.

Most teams are best when their “best” players (as measured in ice time) are on the ice.

For a functional lineup, we want the top-left to be green (good possession when the top players on) and the bottom-right to be somewhat red (poor possession when the fourth lines and bottom-pairing defense are on).

That is not the case for the Capitals.

Yikes.

Observations

• Before we get into the specifics, take in the big picture. Instead of lots of brilliant green at top left, representing puck domination for top players, we’ve got a stripe of rich red. The Caps’ best possession is at top right, where the top forwards intersect with the bottom-pairing defensemen. That means the Caps aren’t getting the most out of their best players.
• But let’s acknowledge the limitations of the visualization: it doesn’t take into account deployments. John Carlson and Brooks Orpik, we know, face tougher competition than the other defensemen. That should be accounted for when interpreting this chart.
• And then we should admit to ourselves at Orpik and Carlson are not good enough to be a shutdown pairing (WaPo agrees). Their ice time needs to be brought down and their work load shared more evenly with the other two pairings.
• It might also help to understand the difference in ice time on the x-axis. Orpik plays about 90 seconds more than Mike Green per game. Green, who is getting the least amount of ice time since he was 21 in the 2006-2007 season, has played about 60 fewer 5v5 minutes than Orpik, though that includes about twenty minutes in the first game of the season, which Green missed to injury.
• Alex Ovechkin and Nick Backstrom are good enough to get great results even with the weakest D-pairing, Orpik-Carlson. With Green and Nate Schmidt, Ovi and Backstrom are dominating play on a truly elite level. For reasons we do not understand, Barry Trotz has chosen to minimize the duration of that domination and maximize the middling results he gets from putting Ovi and Orpik out together. Trotz should play his top line less with his so-called “shutdown” defense pairing.
• Troy Brouwer sets off the passenger alarm. He gets three drastically different results depending on which D-pairing he is with: clobbered with Orpik-Carlson, even with Alzner-Niskanen, whipping arse with Schmidt-Green.
• Why is Joel Ward getting so much 5v5 ice time? He’s ranked third among forwards in 5v5. He played just three games on the top line (games 9, 10, and 11) and one on the supposed second line (game 13). He’s not getting a lot of special teams time (42:45 shorthanded, 25:49 on the power play), which might explain it somewhat. Whatever the reason, it’s apparent that Joel Ward’s 5v5 ice time would be better apportioned to the strong but underappreciated performers lower in the table.
• Strong but underappreciated: Andre Burakovsky and Eric Fehr. Fehr has been scratched 3 times this season while deserving none of that based on his on-ice play. Burakovsky should be the team’s second-line center, but as you’ll see below, Trotz is not using him that way.
• Jason Chimera seems to be the the forward least deserving of his ice time. He is deep in the red, but look at the actual numbers: 42.7, 42.9, 46.0, 44.0. That’d be bad even on the Oates!Caps. Only the Schmidt-Green pairing gets Jason Chimera on the right side of 50 percent. Chimera’s ice time should be cut by 30 percent (fourth line), or he should be scratched altogether. Using him so much for such poor results is going to cost the Caps goals and wins.
• Jay Beagle is just plain not up to snuff. Yet Trotz started Beagle on the top line in games 15 and 16.
• This important: the third line is actually the second line, and the second line is the third line. Chimera gets more ice time than Burakovsky, and that says it all. Trotz might like the style that Ward and Chimera bring, but they’re not playing hockey in a manner conducive to winning. It’s a waste.
• Tom Wilson‘s colors and position might be misleading. He’s missed some games, and when he has been in the lineup, he’s mostly been on the top line.
• Evgeny Kuznetsov has bounced around the lineup more than anyone else, and that’s probably why his colors are all over the place. I’ll just say that I haven’t seen anything out of him that justifies his hype, so I think it’s best to keep him in the bottom six. Trotz is doing the right thing here in my opinion.
• I don’t think this visualization is particularly useful at identifying on-ice chemistry, though it does raise some questions. Why does Nate Schmidt play so poorly with Burakovsky? And more importantly, why does Michael Latta kill it when with the Orpik-Carlson pairing? Whatever is working there might be instructive in minimizing the harm that fake shutdowners are doing to the team.

That’s all I’ve got, though it’s a lot. Are you seeing anything I’m missing? Are you able to articulate Trotz’s reasons for valuing Orpik, Carlson, Chimera, and Beagle as much as he has despite the plentiful and downright obnoxious evidence that it’s bad? If so, please share.