[Ed. Note: We're pleased to welcome Ryan Stimson (give him a follow at @RK_Stimp, In Lou We Trust contributor and passing-play tracker extraordinaire, for the first of two guest posts. He and his team of volunteers have been doing excellent work tracking League-wide stats (and be sure to read about it over at ILWT). Today, however, he's turning his focus solely on the Caps - and more specifically, what the defense might look like this upcoming season.]
Why Look at Passes?
In addition to being able to isolate how players contribute rather than relying on on-ice metrics, by examining passes we can quantify which types of shot attempts are the most dangerous. For example, we tracked full seasons for six teams last season (Chicago Blackhawks, Florida Panthers, New Jersey Devils, New York Islanders, New York Rangers, and the Washington Capitals), and found, collectively, these shooting percentages:
Teams that were able to complete at least a single pass prior to shooting saw their shooting percentage rise from 7.2% to 8.0%; teams that completed multiple passes (at least two) prior to shooting saw that 8.0% rise to 9.6%. In other words, teams whose shots were preceded by multiple passes increased their likelihood of scoring a goal by 2.4%.
Now, the first step in looking at passing rates is to identify if they are repeatable and predictive (they are), so if there are players that can boost shooting percentage, this may be a place to start. We won’t know for sure until we get more data in the upcoming season. However, it does show up in the data and it makes logical sense. Any shot that gives the goalie less time to adjust would provide an advantage to the shooter.
I start out by including this data because it will help frame the rest of the piece as I discuss things like primary and secondary passes. You now know the likely increase in shooting percentage based on these sequences, so you can quickly see that not all possession is equal.
Changes on Defense
With the departure of Mike Green on the back end, the Capitals will be looking for added firepower to join their lethal attack up front. Before we get into specific passing metrics, let’s look at how they performed from a possession standpoint.
About halfway through the season, we changed the way we tracked in order to grab on-ice data (44 games of Capitals data in this format). So, the on-ice data isn’t reflective of the entire season like the player data is. There seems to be a renewed debate on how significant the effects of zone starts are, but those are likely to be more pronounced in forty-four games than eighty-two. Just something to keep in mind.
We tracked at least 200 minutes for each Capitals defensemen, with the exception of Nate Schmidt. He was only on the ice for 105 minutes during this forty-four game stretch and, though his numbers looked very good, are inconclusive. In fact, we only had around 250 minutes for Tim Gleason and Jack Hillen each, so keep that in mind as well. The bulk of the minutes came down to Green, Matt Niskanen, Karl Alzner, John Carlson, and Brooks Orpik.
With Niskanen on the ice, the Capitals had the best passing possession numbers of any Capitals defenseman... but you may be thinking, "boy, why is Carlson so bad?" The fact is that sometimes a player’s data is useful in this format, and other times it doesn’t quite tell the whole picture. Looking at how a defenseman plays with each partner would be useful, don’t you think? I did. Let’s have a look. Below is a spreadsheet you can interact with to look at different pairings and their on-ice passing numbers (with a quick glossary below).
SAG%: Shot Attempts Generated Possession, basically Corsi from passes.
SG%: Shots Generated Possession, basically shots from passes.
SAGE w/ and w/o: How efficient the pairings were on the ice together and apart. Efficiency is the proportion of shot attempts that result in shots.
A2%: Shot Attempts Generated from Secondary Passes only, a measure of sustained possession.
SC%: Scoring Chance Possession: Shot Attempts Generated from passes into the home plate area.
RR%: Royal Road Possession.
Now that we have how each defenseman performed with each specific partner, we can look and see why Carlson had low numbers. Of the 820 passing events he was on the ice for, 680 were with Orpik. Away from Orpik, Carlson had 58.8% possession. A huge improvement - although granted it was only on 140 events. Looking at the WOWY Charts for Carlson and Orpik, we see Carlson’s jump in his overall possession stats again and Orpik stay about the same. Clearly Orpik is just a drag on Carlson; not only a drag, but a significant drag. A 10% jump is beyond huge.
This is nothing new to the readers here, I’m sure.
So with Green's departure, who is going to step up on the Capitals' blue line? My suggestion would be Schmidt. I know we only had a little over 100 minutes of on-ice data for him, but it reinforces what his standard WOWY chart says: players generally had better possession numbers alongside Schmidt than apart.
Now, let’s look at some individual passing data. For these numbers, we have data on the full season. Earlier this summer, JP and Muneeb had taken a look at the team’s performance following defensive zone draws. So, I thought it would be relevant to also look at which defensemen are best at linking play from defense to offense and generating shot attempts in transition. First, we’ll look at zone entry assists.
Entering the zone with possession has been shown to increase the rate at which a team will generate shots. So, identifying which players are generating more offense in transition and entering the zone is certainly worthwhile. For these numbers, I simply added together the primary and secondary passes a player made that led to a shot attempt. The entries are limited to only those that led to shot attempts because we weren't tracking all entries, so this is a quick and dirty way to take a look at which defensemen were assisting in this phase of the game.
Lo and behold we find Schmidt atop the chart. Carlson shows that he more than doubles the production from his partner, Orpik, followed by Green and Niskanen. These were the big four producers in terms of linking play between defense and attach. Based on this area of the game, as well as his solid possession numbers, Schmidt seems like the likely candidate to move up and fill the void left by Greene. The problem is, will Barry Trotz move Orpik to the third pairing?
Let’s look at a few examples of why Schmidt could be in the top four.
We’ll start with this March 21st tilt against the Winnipeg Jets. Schmidt has the puck under heavy pressure in the neutral zone.
Schmidt will eventually attract a second Jets forward, but he’ll backhand a pass into the neutral zone towards a waiting Jason Chimera. Chimera will take this puck in on goal and force a save. He gets a little lucky with this pass, but not giving up the puck under heavy pressure and helping to assist going the other way is a good play to make in this situation.
Here, the San Jose Sharks have just thrown the puck in the Capitals end. Schmidt gathers the puck and makes a tape-to-tape pass to Nicklas Backstrom. The key here is that Backstrom doesn’t have to settle it down or waste time – the puck is right on his stick, so he can continue on into the zone.
This whole sequence comes from Schmidt getting the puck out on a defensive zone draw, then making the pass to lead the transition and result in Backstrom getting the shot on goal. With so many of these sequences each game, having a player who make simple, successful plays on a consistent basis is always a luxury.
Here, Schmidt collects the puck after the Rangers dump it in on a change. He looks up and manages to complete a stretch pass to Evgeny Kuznetsov at the far blue line.
And here is Kuznetsov. He manages a shot on goal because Schmidt was able to recognize the situation and make a quick stretch pass.
Those are a few examples of him contributing offensively. Let’s take a quick look defensively.
This first one is not very kind. It’s a lengthy sequence, so I felt the video would be better. First, Schmidt recovers a loose puck in the neutral zone. He tries to send it into the Rangers zone, but fails to do so. Then Carl Hagelin comes over to take the puck and skate in. Schmidt is behind the play, but he fails to identify Dan Boyle coming in behind him. Boyle gets off a shot that allows the sequence to continue, which ultimately results in Hagelin scoring.
Schmidt had two opportunities here that would have prevented the goal: 1) successfully getting the puck deep and 2) doing a better job of marking Boyle in the slot to prevent that shot from happening.
This next one isn’t kind either. The Rangers send a pass up the boards.
Schmidt again loses a board battle and lets JT Miller go past him. The sequence would end with a shot on goal. On another sequence earlier in this game, Schmidt lost a battle behind the net and conceded a wrap-around chance. It could have just been an off-game, but individual battles might be an area of his game that need improvement.
In reviewing a lot of these events, even with this opportunity for growth, there’s plenty to like about Schmidt’s game. He is a strong possession player, likely owing to a good sense of positional play to make up for occasionally being overpowered on the puck. He is an accurate and efficient passer, evident by the high number of passes he makes in transition that lead to shot attempts. He also had the highest rate of one-timers generated from his passes. This implies an accurate pass right into the "wheelhouse" for shooters. A clumsy pass will not allow the recipient to fire it right away. A one-timer, when directed on goal, was far more likely to to result in a goal than non one-timer shots (11.7% to 6.2%).
The question again is, will Trotz pair him with Carlson in a top-four role? Or, will he play in a third-pairing role? That role wouldn't necessarily be a bad thing. The Devils signed John Moore this off-season and I am excited about it because he’s someone who has feasted on lower competition. There’s nothing wrong with a player dominating against inferior competition, and if Schmidt plays third-pairing minutes, he’ll probably continue to look like a star. Time will tell, but, hopefully, what you’ll come away with here is that Schmidt looks like a solid up-and-coming player.
Lastly, if you use the visualization in that Moore piece I linked (or you access it from Spencer Mann’s Tableau page as well, you’ll see that, across the board in the passing game, Schmidt is a like-for-like offensive replacement for Green.
Offensively, Schmidt should be able to help offset the loss of Green. Who his partner is will likely have a greater hand in how he performs overall. If Carlson and Alzner were a pair again, Schmidt and Niskanen would make for an exciting and offensive-minded pairing.
Finally, a big thanks to Jesse Severe (@jessesevere) for tracking the majority of the Capitals games last season. Thank him on Twitter. Without him, I wouldn't have been able to write this.
Thoughts on this data? Complaints? Suggestions? Anything unclear? Sound off below and I'll circle back and try and answer your questions/comments.