Since Cam Janssen was re-signed and we're stuck with Eric Boulton for another season, I've been thinking a lot about whether or not there is an advanced statistic that can demarcate the line between a hired goon, such as Cam Janssen, and physical players who spend a lot of time in the box like David Clarkson. I plan on using this analysis to kick off my summer re-launch of my blog (don't click the link in my signature since I haven't updated the blog since around the all-star break). After the jump I'll look at two different potential advanced statistics for this.
Penalty minutes alone are not enough to really differentiate between the two types of players. The top five leaders in PIM this season were Derek Dorsett (CBJ, 235), Zac Rinaldo (PHI, 232), Zenon Konopka (OTT, 193), Chris Neill (OTT, 178), and Cody McLeod (COL, 164). There is an obvious distinction between a player like Dorsett who had 20 points this season and Rinaldo who had 9. In Ottawa even, Konopka only had 5 points while Neill had 28. Maybe the ability to produce is the line between a goon and a physical player. After all, if the goon can't score points, which are needed to win games, why is he playing in the NHL at all?
This leads to the first proposed advanced statistic which is the player's ratio of PIM to Points. For a guy like Cam Janssen, with 75 PIM and 1 Assist on the season, this number is a fairly high 75. David Clarkson, on the other hand, with 138 PIM lead the Devils, but his 46 points brings his PIM:P (yeah, I went there) rating down to 3. The higher your rating, the more likely it is that you're paid to bully people. Eric Boulton's PIMP rating for 2012-13 is infinity (with a lifetime rating of 18.9).
A rating of zero is useless without the context of points. Antti Miettinen had 13 points in 45 games with 0 PIM. That's pretty great for him, but not impressive at all. Kyle Wellwood also had a 0.09 rating with 47 points and 4 PIM.
Obviously what you really want to know is who the highest (non-infinite) PIM:P ratings in the NHL belong to. I haven't crunched every player yet, so this isn't 100% guaranteed, but here are your likely top 3, all belonging to players who had 1 point on the season. These guys were all very close to joining Eric Boulton with a score of infinity:
I wasn't really happy with this metric entirely though. It's a great measure, I think, of uselessness with respect to point production, but perhaps players have other intangibles that make them worth playing in the NHL level. Maybe they bring high energy to the team, or are able to shut down the opposition in terms of shots against or even goals against. If these were the case, then surely these guys would be getting a respectable amount of time on ice, right?
This brought me to advanced statistic number two: Time On Ice Less Penalty Minutes per Game. The idea is that if you're doing good things for your team, your coach will reward you with more minutes. Calculating this number is easy from the NHL.com statistics. I'll provide Cam Janssen as an example. Cam had 75 penalty minutes in 48 games played in the 2011-12 season. This means that, on average, he was taking 1:34 in penalties every game. His time on ice was a miserable 4:40. The difference gives us 3:06. Only three players who played more than 10 games this season managed a worse number than Janssen in this rating:
Tom Sestito - 0:57
Brandon Bolig - 2:40
Darcy Hordichuk - 2:52
Cam comes in at a (dis)respectable 4th place behind these three goons. Brad Staubitz, who lead the league in non-infinite PIM:P rating, came in at 16th with a 4:33. Eric Boulton was 13th (4:20) and Brad Mills came in at 25th (5:59).
Neither of these ratings is perfect, but the numbers do a good job of showing us which players are a useless waste of roster space versus which players are producing at an NHL level with a physical edge to their game. I will work on expanding my analysis to include all NHL skaters for this season over on my blog and then give some more in-depth analysis on both ratings. In addition, I will look to see if there's a way to combine both statistics into some kind of super-advanced statistic that relates points-per-minute-per-game versus penalty-minutes-per-game to try to get a better image of how
useless useful some players are.
I'd love to know what you guys think. I feel that the general consensus around here is that we don't need advanced statistics to tell us that Cam Janssen and Eric Boulton are useless players. Do you think there is a better way to compile this statistic? Should I use hits as well, even though that statistic is inconsistent from arena to arena? The better our ability to gauge the usefulness of physical players the better we can complain about Lou keeping around guys like Janssen, Boulton and Mills.