The Lost Years: Imputed Fenwick Close % for 2005-06 and 2006-07

One of the more frustrating aspects of advanced statistics in hockey is the difficulty of putting work in historical context. While the availability of data in the so-called “Behind the Net era” has enabled rapid advances in our knowledge of the game, this knowledge only goes back to the 2007-08 season. Unlike, say, baseball (where you can calculate Ty Cobb’s WAR if you’re interested), we don’t know much about advanced metrics in other seasons.

While others are doing God’s work in attempting to address this problem, it occurred to me that I could use a simple statistical model to quickly offer up two valuable sets of information: team Fenwick Close percentages for the 2005-06 and 2006-07 seasons. Fenwick Close, of course, is widely regarded as the best currently-available metric for team puck possession. What you might not know about it is that full-season Fenwick Close correlates very closely with full-season shot differential in all situations. This implies that we don’t need shooting data limited to close-game situations, or even data limited to even strength, to impute full-season Fenwick Close with a high degree of accuracy.

So, I gathered up each team’s Fenwick Close % from stats.hockeyanalysis.com, from 2007-08 through the 2013 season, and supplemented it with each team’s all-situation shots-for % from nhl.com. The correlation between the two is 0.925, or very, very strong. A simple linear regression (Fen Close doesn’t tend to have extreme values, which makes it less necessary to use a logistic model here) of Fenwick Close as a function of SF% has an R-squared of 0.856, and estimates an intercept of -0.057 and a coefficient of 1.113 for SF%. This implies that a team’s SF% can be plugged into the regression equation to yield a pretty accurate estimate of Fenwick Close. Doing this with teams’ SF% from 2005-06 and 2006-07 yielded the following estimates.

imputedFenClo

A few things to note here:

  • I’d be reluctant to go much further back in time than 2005 with an analysis like this. Between changes in the way the game is played, to actual rule changes (e.g., the Bettman Point) affecting teams’ shooting behavior, it’s not clear that the equation would be valid in other eras.
  • Given that the 2007-08 and 2008-09 Red Wings are two of the most dominant possession squads of the Behind the Net era, it probably shouldn’t be surprising that the 2005-06 (estimated Fen Close of 56.9%) and 2006-07 (58.7%) Wings also acquitted themselves well.
  • Another very dominant possession team from these two seasons is the 2005-06 Senators, who won the Eastern Conference that year behind an estimated 55.0% Fenwick Close.
  • Edmonton’s run to the 2006 Finals looks a bit less surprising in light of their possession differential that season.
  • The 2006-07 Toronto Maple Leafs and Florida Panthers: two more strong possession squads undone by sub-0.900 goaltending.
  • The change in puck possession from the Glen Hanlon era to the Bruce Boudreau era is pretty remarkable.
  • The Therrien-era Pittsburgh Penguins: just as awful a possession team as you remember.
  • By these estimates, the 2007-08 Thrashers remain the worst possession team since the 2005 lockout. Though the 2005-06 Blue Jackets gave them a run for their money at 44.3%.

About Nick Emptage

Nicholas Emptage is the blogger behind puckprediction.com. He is an economist by trade and a Sharks fan by choice.
This entry was posted in Original Analysis and tagged , , , . Bookmark the permalink.

One Response to The Lost Years: Imputed Fenwick Close % for 2005-06 and 2006-07

  1. Pingback: Is Fenwick Close a Consistent Predictor of Winning? | Puck Prediction

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s