Penguins coach Dan Bylsma’s system validated by advanced stats

Much like in baseball, football and basketball, the hockey community is increasingly turning toward so-called “advanced stats” to help project and evaluate the performance of players and teams.

Most of the popular metrics are based upon the idea that shots, both for and against, are the building blocks of success and failure in the sport. We’ve posted a few articles on this site regarding the Corsi number, which records all shot attempts (on goal, missed or blocked) in an effort to measure puck possession.

The Fenwick metric takes that concept a step further, as it eliminates blocked shots, thus gauging exactly how many pucks have a chance to become goals. There are merits to both Corsi and Fenwick, with the former being more of a catch-all stat and the latter trying to eliminate some of the noise surrounding the signal.

Chris Boyle of the Canadiens blog Habs Eyes on the Prize contends Fenwick is the more predictive metric of the two. He’s put together a wonderful infographic to illustrate that point, using only Fenwick results from close games (click to enlarge):

As Boyle explains on his accompanying post, the better a team’s Fenwick Close number, the better its chance of qualifying for the Stanley Cup playoffs and subsequently making a deep run in them. Each of the team logos above represents a full season, with every team’s results since 2007-08 recorded.

Interestingly, the two Penguins teams on the graph (2007-08 and 2008-09) who finished the regular season with a below-average Fenwick reached the Stanley Cup Final. In fact, those are the only two sub-.500 Fenwick teams since 2007 to make it that far.

There is a caveat, however, one that makes coach Dan Bylsma look good. I’ll use Boyle’s description here:

"When you look at the one outlier who won the Stanley Cup, you also notice that there was a major change that occurred during that season which creates enough noise to skew the result. The 2009 Penguins are the only team to buck the odds as a non +.500 fenwick team, but their .499 was .549 under Dan Bylsma. Mar 9, 2013; Toronto, ON, Canada; Pittsburgh Penguins head coach Dan Bylsma yells out instructions from the bench during their game against the Toronto Maple Leafs at the Air Canada Centre. The Penguins beat the Maple Leafs 5-4 in the shootout. Mandatory Credit: Tom Szczerbowski-USA TODAY Sports"

Bylsma has preached a puck-possession, force-the-issue style of play since arriving in Pittsburgh, with the Fenwick results showing that his teams have implemented that mentality successfully. Although the Penguins haven’t advanced past the second round since winning the Cup, their regular-season Fenwick ratings since 2009-10 then have all been above 53.

Last season’s team topped out at 55.02, one of only eight teams to break 55 in the past five years. This year’s Penguins are right around average in Fenwick Close, placing 13th out of 30 teams with a 50.98 number through 38 games (stats courtesy of BehindTheNet.ca).

Statistics gain accuracy as the sample size grows, which explains why Fenwick Close has been (slightly) better at predicting which teams will make the playoffs as opposed to which will have an extended postseason stint. In a typical season, the Penguins would have 44 games to go before elimination hockey begins, so we can hope they will return to their usual puck-possessing ways over the next few weeks.

Of course, the playoffs are the epitome of a small sample size, so these numbers validate my belief that the Penguins’ recent springtime failings are as much a matter of poor fortune as any flaws in the roster.

In short, Bylsma has been a clear success, but management will need to eventually see his teams’ below-the-surface acumen translate into more playoff wins.

I believe it will, as long as the Penguins can play at their usual Bylsma-led level.