navigation-background arrow-down-circle Reply Icon Show More Heart Delete Icon wiki-circle wiki-square wiki arrow-up-circle add-circle add-square add arrow-down arrow-left arrow-right arrow-up calendar-circle chat-bubble-2 chat-bubble check-circle check close contact-us credit-card drag menu email embed facebook-circle facebook-square facebook faq-circle faq film gear google-circle google-square google history home instagram-circle instagram-square instagram linkedin-circle linkedin-square linkedin load monitor Video Player Play Icon person pinterest-circle pinterest-square pinterest play readlist remove-circle remove-square remove search share sign-out star trailer trash twitter-circle twitter-square twitter youtube-circle youtube-square youtube

Drilling down on the stats, with Capitals StatTalk

November 6, 2013
Capitals Voice Promo Image

In addition to our regular crew on the Capitals Radio Network this year, Arik Parnass has joined us to provide some in-depth statistical background. Arik is a student at Georgetown, and I've already learned he's a lot better with numbers than I am. If looking at the mathematics behind the pucks is your thing, I hope you'll take a look at Arik's first offering here. You can follow him on Twitter in-game at @arikparnass 


Capitals StatTalk – November 6


91.5% - That’s Washington’s current penalty kill percentage, good for first in the NHL. The Capitals are also first in goals allowed per 60 minutes shorthanded, which is more reliable as a metric because it doesn’t count a goal allowed five seconds into a power play as equivalent to one allowed in the dying seconds, and a major as equivalent to a minor. The Caps allow only 1.81 goals/60 shorthanded, and Colorado, the next best, is significantly behind at 3.17.  The Blackhawks, who bring up the rear, allow 12.66 goals/60, so what Washington’s two units have done is quite impressive.


Last year, the Caps were 21st in the league at goals allowed per 60 minutes shorthanded with largely the same units, so what accounts for the difference? Well last year, Capitals goalies posted a .887 save percentage while shorthanded, which is above league average but not out of the ordinary. This year, however, they’ve got a league high .970. Braden Holtby and Michal Neuvrith are stealing the show, and it’s not only impressive, but it’s also likely unsustainable. The highest shorthanded save percentage last year belonged to the Ottawa Senators, and that was only a .921. Like most stats, save percentage tends to regress towards a mean, and Holtby and Neuvrith are bound to face some shots they can’t stop eventually.


Why is this an issue? Well, the Caps have done a good job of keeping the puck out of the net so far on the PK, but not at keeping the puck away from their goalies.  They have allowed the 7th most shots against per 60 minutes shorthanded in the league, and where shots are allowed, goals tend to follow. If they want to stay at the top of the NHL in penalty killing, limiting shots against will need to be a priority moving forward.


56.8% - That’s the Capitals’ Fenwick For Percentage in the past three games with the score close[i]. Fenwick represents the number of unblocked shot attempts a team records, and over the past three games, the Caps have managed to get 54 shot attempts through unblocked, while allowing only 41[ii]. That number is significant because Fenwick correlates very highly with scoring chances, and teams that finish the season with a high Fenwick percentage tend to be the most successful. The Caps had struggled in the shot attempt department going into the Flyers game, so they need to use the past three performances as a building block in order to develop into true contenders.


40.8% - This is the percentage of John Erskine’s post-whistle non-neutral zone shifts that start in the offensive zone. In other words, Adam Oates calls on the 33-year old defenseman to start his shifts in the defensive zone significantly more than in the offensive zone. Why is this important? Well, a player’s +/- and other goal and shot-based statistics can be influenced by how sheltered that player is zonally. Coaches often designate more offensive-oriented or younger players to start shifts in the offensive zone — bringing with them less responsibility — and more defensive or experienced players to start in the defensive zone. What’s impressive about Erskine’s performance is that despite the tough minutes that he plays, he has put up the fifth-highest Corsi For Percentage among all Caps who have played in at least seven games. The four guys ahead of him (Schmidt, Wilson, Backstrom, and Ovechkin) all get significantly more offensive zone starts.



O/D zone start %

Corsi For Percentage

Nate Schmidt



Tom Wilson



Nicklas Backstrom



Alex Ovechkin



John Erskine





Brooks Laich



Troy Brouwer




Corsi represents all shot attempts, including blocked shots. It correlates well with offensive zone time, and tends to be used over Fenwick when talking about individual players because it represents a larger sample size, and players have far less data to look at relative to entire teams. So what does this all mean? Despite playing tough minutes this year, Erskine has managed to tilt the ice in his team’s favor from a possession standpoint. The Capitals have missed the poise that Erskine has brought when healthy, and although some of the younger blue liners have impressed, they will be looking forward to him getting back in the lineup.

[i] Close means within one goal in the first or second periods or tied in the third period. One uses “score close” metrics to eliminate score effects from the sample. In other words, a team that leads a game by 2+ is more likely to sit back and allow shots, but that shouldn’t be held against their possession stats. The inverse is true for a team trailing by 2+ goals. “Score close” metrics are usually used for teams and not players because with teams there is a larger sample size of data to work with.


[ii] Both Fenwick and Corsi generally refer to 5-on-5 play only unless otherwise indicated