Who is the NHL's best penalty killer?
Picking up on something Matt Fenwick was talking about over at The Battle of Alberta, I've been working on an analysis of the league's best penalty killers.
The results will probably surprise you.
Using data from Behind the Net and taking only players who have played at least 120 minutes shorthanded as of Feb. 18 (approximately two minutes per game), the player with the lowest number of power-play goals scored against per two minutes short-handed ice time is Vancouver Canucks forward Josh Green.
In fact, the Canucks — owners of the league's top-ranked penalty kill — have the four highest-ranked players in the category, and five of the top seven players. This isn't a measurement that takes into account the strength of a player's teammates or the goaltender (hello Roberto Luongo!), so take it for what it's worth.
A look through the chart toppers reveals a list of players you likely wouldn't name off the top of your head if pressed for defensive dynamos: Green, Fernando Pisani, Jeff Friesen, Alex Burrows, David Legwand, Jozef Stumpel and Todd White lead the way for forwards, while Lukas Krajicek, Sami Salo, Mattias Ohlund, Daniel Tjarnqvist, Kim Johnsson and Branislav Mezei top the blueliners.
Here's a look at the top 30 overall. I'll do my best to get a full spreadsheet up later on if there's enough interest:
| Player | Team | Pos | PPGA | MINS | PPGA per two mins SH | |
| 1 | GREEN | VAN | F | 3 | 142 | 0.042 |
| 2 | KRAJICEK | VAN | D | 5 | 141 | 0.071 |
| 3 | SALO | VAN | D | 8 | 187 | 0.086 |
| 4 | OHLUND | VAN | D | 10 | 231 | 0.086 |
| 5 | PISANI | EDM | F | 7 | 159 | 0.088 |
| 6 | FRIESEN | CGY | F | 7 | 149 | 0.094 |
| 7 | BURROWS | VAN | F | 10 | 209 | 0.096 |
| 8 | LEGWAND | NSH | F | 8 | 160 | 0.100 |
| 9 | TJARNQVIST | EDM | D | 7 | 134 | 0.104 |
| 10 | STUMPEL | FLA | F | 10 | 187 | 0.107 |
| 11 | WHITE | MIN | F | 9 | 167 | 0.108 |
| 12 | JOHNSSON | MIN | D | 13 | 234 | 0.111 |
| 13 | MEZEI | FLA | D | 7 | 124 | 0.113 |
| 14 | STURM | BOS | F | 8 | 139 | 0.115 |
| 15 | JOVANOVSKI | PHX | D | 7 | 122 | 0.115 |
| 16 | MARCHANT | ANA | F | 10 | 168 | 0.119 |
| 17 | KESLER | VAN | F | 10 | 162 | 0.123 |
| 18 | REASONER | EDM | F | 11 | 178 | 0.124 |
| 19 | WALLIN | CAR | D | 10 | 162 | 0.124 |
| 20 | ROZSIVAL | NYR | D | 11 | 176 | 0.125 |
| 21 | SMOLINSKI | CHI | F | 10 | 155 | 0.129 |
| 22 | ROLSTON | MIN | F | 12 | 186 | 0.129 |
| 23 | MORRISON | VAN | F | 9 | 139 | 0.129 |
| 24 | LAROSE | CAR | F | 10 | 154 | 0.130 |
| 25 | MOEN | ANA | F | 12 | 185 | 0.130 |
| 26 | MCCLEMENT | STL | F | 11 | 169 | 0.130 |
| 27 | STAJAN | TOR | F | 9 | 137 | 0.131 |
| 28 | DALEY | DAL | D | 11 | 165 | 0.133 |
| 29 | ZANON | NSH | D | 14 | 209 | 0.134 |
| 30 | JONES | PHI | D | 11 | 163 | 0.135 |
The list looks pretty solid at first glance, albeit there are a lot of surprising names on there. What we're missing are some of the biggest minute penalty killers, who are probably seeing some of the toughest PK minutes.
Here's a brief list of some of the other biggest minute PKers (250+ for defencemen, 180+ for forwards) and where they rank among the 250 players with a minimum of 120 shorthanded minutes:
Defence - Jay Bouwmeester, 303 minutes (53rd); Jason Smith, 251 minutes (61st); Chris Phillips, 309 minutes (70th); Bryan Allen, 260 minutes (75th); Dan Hamhuis, 275 minutes (104th); Kevin Bieksa, 242 minutes (105th); Derian Hatcher, 370 minutes (110th); Chris Chelios, 266 minutes (115th).
Forwards - Mike Johnson, 193 minutes (44th); Niklas Hagman, 191 minutes (46th); Jamal Mayers, 204 minutes (47th); Brian Sutherby, 199 minutes (52nd); Manny Malholtra, 195 minutes (56th); Anton Vermette, 189 minutes (72nd); Rod Brind'Amour, 222 minutes (78th); Stu Barnes, 210 minutes (79th).
Who are the worst penalty killers in terms of power-play goals allowed per two minutes?
Among those who have played 120 minutes shorthanded, Ryan Johnson from the St. Louis Blues ranks last among 242 players. The bottom 20, in order: Johnson, Dustin Brown, Paul Ranger, Rob Blake, Derek Morris, Niclas Havelid, Francois Bouillon, Bobby Holik, Shawn Bates, Aaron Miller, Bryan McCabe, Kyle McLaren, Brad Larsen, Keith Ballard, Brad Richards, Mattias Norstrom, Karlis Skrastins, Dallas Drake, Mike Peca and David Vyborny.
UPDATE I've put some of the raw data up on Google Spreadsheets.
Labels: Luongo
by James Mirtle @ 7:19 p.m. | Permalink |
18 comments
18 Comments:
What we're missing are some of the biggest minute penalty killers, who are probably seeing some of the toughest PK minutes.
I think that's somewhat of a missing key to this analysis, not that I'm blaming you to any degree, you're doing the best with what's out there.
But still, you gotta consider that most every team has two power play units, one better than the other. And often enough, there is a penalty-killer who has to face that top unit more than secondary units. I think a 'strength of opposition' metric (similar to BTN's 5-on-5 stuff) would help set the stage here.
Also, I think his numbers only look at 5-on-4 (top of my head), but 5-on-3 should probably be a part of this also. Sure, being the lone forward on a 3-on-5 means you're probably getting scored on more often, but it should be a consideration for 'best penalty killing forward' (if you're not the go-to-guy on your team, I doubt you should carry that title for the league).
Good start, though.
I'd also take a look at what type of save percentage these guys have behind them on the PK, but still interesting stuff.
Man, awesome work Mirtle.
I'm pretty sanguine about the "Quality of Opposition" thing on the PK, at least compared to ES. Two eyes tell me that there's probably 4 PK shifts and 2 PP shifts on the same 2-minute powerplay... half the time you're on the road, where you can't match up anyway... etc.
The goaltending thing is the biggest wildcard when comparing players on different teams (no one could seriously believe that one team has the 4 best PKers in the NHL), but comparing players on the same team, I suspect it's extremely "fair".
- Matt F.
Sleek brought up the point that I was thinking about...guys like Green and Krajicek are 2nd unit guys who play against 2nd unit guys. Green isn't out against Thornton as much as he's out against a guy like Pavelski.
Still, this is a good start and some food for thought. Analysis has to start somewhere, and now Josh Green has fuel for his contract negotiations ;)
Two eyes tell me that there's probably 4 PK shifts and 2 PP shifts on the same 2-minute powerplay.
Sure, the unsuccessful ones.
But I'm not so sure line-matching (H/R) matters too much. Most PPs start with the top option on either side--and that is kind of what's missing.
You can show me numbers that say that Joe Dipenta is a better PKer than Chris Pronger, but it doesn't quite fly. Looking at the list, there isn't a plethora of go-to PK guys (Pahlsson, Madden, Pandolfo, Lehtinen), and their absence is conspicuous.
As a real simple way of trying to factor out some of the inlfuence of goaltending you could take the adjust those values up or down based on the team goals against average vs the league goals against average.
For example, the league average goals against average is 2.89. Vancouver's goals against average is 2.43.
Josh Green's adjusted PPGA/2minSH would then be:
0.042 * 2.89 / 2.43 = 0.050.
It's pretty rudimentary but is fairly easy to calculate and would probably give you slightly better results.
That said, Josh Green would still rank really high, just as he does defensively in my player rankings.
Josh Green for the Selke!!!
I'm not sure why you're leaving out shorthanded goals for in this analysis, which is already included in Behind the Net's data. The fact that some guys like Martin St. Louis create lots of SH goals is quite significant.
You should also include a quality of ice factor as well, as PK units in places like Edmonton will struggle more as PP units have a better surface on which to move the puck and create scoring opportunities.
It's really a pretty pedestrian ranking without factoring that in. Shame on you.
We should not forget altitude as well. Killing penalties in high altitude, low oxygen levels Colorado is more difficult. That explains why Colorado has a better road PK% (80.4) than home PK% (79.9%).
See? Johnson gets what I mean. That's .5%, nothing to sneeze at!
I'm also working on a roomate factor. Seems obvious to me that if Madden and Pandolfo room together they have a distinct advantage over Lehtinen if he's forced to share space with Jon Klemm.
I'm not sure why you're leaving out shorthanded goals for in this analysis, which is already included in Behind the Net's data.
The whole point of playing the penalty kill is preventing goals; whether or not a Samuel Pahlsson scores while shorthanded or not I think it besides the point. Granted, it's a nice bonus, but we're trying to determine and isolate who the best defensive forwards are.
I think that should be Daniel Tjarnqvist named in your post, not Matthias Tjarnqvist.
whether or not a Samuel Pahlsson scores while shorthanded or not I think it besides the point.
Tough example. Sammy only scores shorthanded in my gameday predictions; he's neglected to score shorthanded in an actual game this year.
:)
Which is exactly why I used him — it doesn't particularly matter that he doesn't score on the PK in terms of how effectively he is defensively.
One season is a small sample since there are a lot of situational factors (goalie, teammates, opponents), but a few years of data can give a better picture. I looked at the PK rates (PP goals per 2 minutes on PK) I could find from the last 4 seasons for some high minute forwards. Here's how I'd classify them:
Elite (under .14): K. Maltby, T. White, J. Stumpel, T. Marchant
Good (.15 to .17): J. Lehtinen, K. Draper, S. Pahlsson, C. Drury, B. Rolston
Mediocre (over .18): R. Brind'Amour, M. Peca, J. Pandolfo, J. Madden, S. Horcoff, B. Holik
If you're just looking to isolate defensive performance on the PK, that sounds fine. But if I was a coach, I'd certainly feel comfortable putting St. Louis and/or Lecavalier out there, knowing their potential for scoring a SH goal.
Obviously there are late-game situations (protecting a lead vs. trying to tie) that might dictate leaning in one direction or another.
As a real simple way of trying to factor out some of the inlfuence of goaltending you could take the adjust those values up or down based on the team goals against average vs the league goals against average.
This assumes that a team's goals against average is entirely the result of the goaltender, and not the team in front of him. I've watched the Devils enough to know that this is a bad assumption. There are a lot of data interactions going on here.
The whole point of playing the penalty kill is preventing goals;
I disagree. The whole goal of the penalty kill is exactly the same as it is for the rest of the game: maximize the number of goals that you score, and minimize the number of goals the other team scores. This gets you to the objective: scoring more goals than the other guy does.
If your goal is simply to find out who are the best players at preventing goals, then you are taking the correct approach. If you are trying to determine who the best penalty killers are, then your definition is clearly flawed. Very obviously, one of the reasons that John Madden is considered a top penalty killer is that he's always a considered a threat to produce a short-handed goal; this is somewhat more reputation than reality at this point, but that reputation was very well earned.
Post a Comment
<< Home