Sunday, October 17, 2010

Last Five Fenwick: Games 1-5

I was planning to do a weekly look at scoring chances, but Neil at RMNB got to that first. So here debuts a new series, looking at the Capitals' Fenwick over the past five games (similar to Japers' Rink's "This Week in Corsi" series). Remember, Fenwick is (shots for + missed shots for) - (shots against + missed shots against). It correlates a bit better with scoring chance ratios than Corsi, which incorporates blocked shots, does, because it's more representative of what we want it to represent--territorial dominance (or lack thereof).

While my numbers should correlate strongly with Neil's over good sample, over short term we could see differences, as Fenwick generally has a bigger sample. We also can't compare scoring chances directly across different teams, as we can do with Fenwick and Corsi numbers.

My hope is that after the regular season ends, we can combine the Fenwick and scoring chances data to see evidence that the Capitals as a team deserve to have high shooting percentage because their average shot is more likely to be a scoring chance. This point is the one made by Caps fans who are adamant that last season's abnormally high 10.9% shooting at evens was a result of skill, not a fluke as the statisticians like Gabriel Desjardins say. I'd also like to see evidence of improvement when I graph the numbers.

Here's a link to my spreadsheet. When Vic Ferrari updates the playershot scripts, I'll add in Fenwick% too. Right now adding those numbers in would make the spreadsheet too messy.

A realistic target for the Capitals I think would be Detroit's totals last season. While Chicago's are probably untouchable, Detroit's totals were pretty similar to Washington's from 2007-2008 an 2008-2009. A 54% Fenwick with the score tied sounds pretty solid to me and would represent a 3% improvement over 2009-2010.

My quick math from the 2009-2010 team Fenwick totals at Timeonice says Chicago out-Fenwick'ed opponents by roughly 10 per game, 51 per 5-game segment. Crazy. Detroit was around 4-21, 09-10 Caps 2-10, 08-09 DC 6-31, and 07-08 Washington 6-32.

I think something in the 20-30 range over 5 games is what we want to see. 10 is just too poor. In other words, to grade on the Caps' Fenwick per 5 games:

30+: A+
25-29: A
20-24: B+
15-19: B
10-14: C+
5-9: C
0-5: D
Under 0: F

To put Fenwick/60 into context, well, last season the league leader was Marian Hossa at 23.01 Corsi/60 (Corsi On). Fenwick would probably be a bit lower, say around 20. Two Caps topped league-wide Corsi/60 in 2008-2009, Sergei Fedorov at 24.83 and Eric Fehr at 23.94 (with tougher competition and weaker teammates than Fedorov...Fehr is a beast). For the top-end Caps (Backstrom, Green, and Ovechkin), a good target would be around 15-20. For others, small positive numbers works, especially those like Gordon and Steckel who play tough minutes. We don't want to see anyone below zero unless Boudreau decides to line-match forwards closely over a 5 game segment.

Results from Games 1-5:

Game 20011: Washington 2 @ Atlanta 4
Game 20017: New Jersey 2 @ Washington 7
Game 20032: Ottawa 2 @ Washington 3 (OT)
Game 20037: New York Islanders 1 @ Washington 2
Game 20060 : Washington 3 @ Nashville 2 (OT)

The shining lights from the first five games were Nicklas Backstrom and Jeff Schultz. The Caps as a whole, though, were fairly inconsistent. Even Backstrom and Schultz used a strong game against Atlanta to buoy their Fenwick/60 to the positive side of the ledger. Brooks Laich and Alexander Semin were the consistent performers, each no higher than 2 and no lower than -2 in any game.

Unfortunately, the Caps as a team didn't do a terribly good job controlling possession. As a team they were to the negative side of the ledger, and quite far from the ~20 target I'd like to set them for every five games.

Another note: clearly Marcus Johansson is not ready for tough minutes. So stop giving them to him, Bruce Boudreau, especially when he's not given great teammates and easy zonestart. That is to say, young player + tough zonestart + tough competition + weak teammates is a recipe for disaster. No wonder he was in the red in scoring chances and Fenwick.

My list says I should grade the team with a F, but the Caps did win four of five. Hockey is all about winning at its core, and I'll give the Caps the benefit of the doubt. Maybe they were controlling something we can't yet quantify. The certainly did decently in scoring chances and the PK was excellent. Nicklas Backstrom was amazing as well, and even John Carlson, for all his...well, let's say his defensive zone coverage wasn't what we're used to...did put up good offensive numbers too. Eric Fehr and Jason Chimera played extremely well for at least three games (more than most other Caps), and our favorite Mr. Nasty Jeff Schultz was the Caps' best player not named Neuvirth. I'm starting to believe that Schultz is a legitimate top-pairing defenseman, skating notwithstanding. He's a great hockey player, and keep in mind he looks even better when you factor in Corsi.

Grade: C


  1. Nice work - one point though: Fehr and Chimera have looked good statistically as you point out, and have been Johansson's most common linemates. Perhaps BB is putting his line out with crap defensive pairings, but it seems like thus far BB, if only inadvertently, has given Johansson strong linemates.

  2. That's a good point. MarJo was even further in the red until the game against Nashville, when his most frequent D pair linemates were Schultz-Alzner. I'll keep a close eye on his D linemates over the next few games.

  3. Yeah, that just looked really strange to me. There's a similar thing going on with Laich/Semin and Flash, but not nearly to the same degree.

    Seeing this has me wondering about line changes and the seams and overlaps that result. My initial hypothesis is that, because these are unsettled events in games, there may be a higher number of Corsi events taking place while lines are discombobulated. I can't really think of what else would explain such a wide discrepancy.