As anybody will tell you, the Red Wings' penalty kill has not gotten the job done this season. Tied for 25th in the league with a 77.4% effectiveness rating, Detroit has allowed 19 goals to the opposition on power plays through 21 games; three other teams share this dubious distinction with the Wings as the worst in the league at preventing goals while shorthanded.
As part of a look at the advanced stats for the Wings, Chris over at Nightmare on Helm Street went over the Wings' PKers who average more than a minute of short-handed ice time per game in an article called 'Who's Killing the Penalty Kill? - Part 2 of a Statistical Look at the Wings.'
In his post, Chris uses Behindthenet.ca's rate stats for 4-on-5 goals against per 60 minutes of play. The whole chart can be viewed there, but the conclusions boil down to this:
So Filppula, Cleary, and Quincey are doing alright, and everyone else is pretty crap. Including Datsyuk, which does surprise me a bit.
So the TL:DR of part two? Zetterberg for MVP. Huskins is good when he doesn’t have to skate.
Huskins hasn't been on the ice for a PP goal against yet this season and Zetterberg's pace is approaching legendary at 1.8. The other three mentioned are at least close enough to his comparison to average (sub-7) to get an "alright" rating and then everybody else is "pretty crap."
This conclusion caught me by surprise. Partly because basing it solely on a stat that is mostly out of a player's control can get a bit tricky (essentially, just like plus/minus, rate stats which rely on goals happening are also fairly heavily influenced by uncontrollable noise), and partly because the Red Wings' penalty kill got miles better after Mike Babcock pulled Justin Abdelkader and Patrick Eaves out of that role in favor of more PK time for Datsyuk and Zetterberg.
In looking for context, I focused on Drew Miller, who leads the team in PKTOI/G among forwards. Miller's 2:28 per game on the PK is comparable to a few of the Wings' defensemen and is higher than the next-closest forward's time by more than 30 seconds.
As far as PKing goes, we're essentially looking at two different epochs for the Wings: the first goes from the beginning of the season to February 5th against Calgary. Through that game, Detroit gave up 13 power play goals on 43 opportunities for a 70% efficiency rating. Epoch 2 saw the Wings make adjustments to their PK strategy and personnel. Since that time, Detroit has killed 35 of 41 penalties, good enough for an 85% efficiency rating.
Do you want to guess the last time Drew Miller was on the ice for a power play goal against? If you guessed February 5th against Calgary, then congrats. If you further guessed that Abdelkader was on the ice for that goal, then you just read my mind. If you THEN guessed that the goal was essentially Abdelkader's fault then... well you're wrong. Drew Miller was given a minus in that game's CSSI for being out of position.
But astute readers will point out that February 5th was nearly a month ago. In that time, Miller has been on the ice for 25:59 of short-handed hockey without seeing a power play goal go into the net.
For comparison's sake, that time-frame is longer than the total amount of PK time for Mr. I-haven't-been-on-for-a-single-PP-goal-against Kent Huskins over the course of the entire season. It's also longer than Valtteri Filppula, Cory Emmerton and Justin Abdelkader have spent killing penalties all season.
Of course, that means that during the dark time prior to the end of the Calgary game, Miller was on the ice for 7 PPGA in about 26 minutes of ice time, meaning that on Feb. 5th, his per-60 rate was just below Abdelkader's at about 16.
However, if you dig farther, you see one big ugly outlier staring at you. Detroit allowed four power play goals to St. Louis in the season opener. Miller was on the ice for all four of them. If you remember back to that game, two of those power play goals were breakaways for the Blues. Bury your head in the sand to eliminate the outlier and Miller's completely unacceptable season-long 8.13 rate of PPGA/60PKTOI becomes a beautiful 3.69 rate.
So is Miller an awesome PKer or a terrible one? We've got conflicting reports on this. The truth seems closer to the former than the latter. Last year's numbers for Miller were the best on the team among forwards. He's not an offensive threat shorthanded, but since Babcock has spent more time putting him with bigger guns than Abdelkader, his quiet and effective shot-blocking and positioning have bolstered the PK.
One of the biggest problems with rate stats for things like special teams on individual players is that it takes a lot of games to get enough time to avoid the sample size problem. Only one Wings player has actually played more than 60 minutes shorthanded so far this season and 60 minutes is hardly enough short-handed ice time to make conclusive judgments. Looking at last year's stats, Chris is just about right on to call the 6.0 range a good average. Looking at that season, the three middle-of-the-pack teams for PK efficiency allowed 6.1 power play goals against per 60-minutes of 4-on-5 ice time. While I'd like to do better than that, it's at least the standard we can set to know whether or not they're being suckier than is acceptable.
However, with sample-size issues for everybody, the number can skew wildly. Pavel Datsyuk's rate of 10.57 is staggering considering the eyeball test suggests he's doing a good job short-handed. Then you look at the fact that Datsyuk is getting a .750 save percentage behind him. While shot quality is a fun discussion point, this year's CSSI doesn't indicate that Datsyuk is driving the opposition to be shooting that well against him. Sometimes you just get shitty luck.
On average, save percentages go down while shorthanded. If we consider .880 to be "average" and recalculate the 24 shots on goal Datsyuk has faced while shorthanded, that cuts his goals against in half (from 6 to 3). This also cuts his GA/60 pace in half from 10.57 to 3.8, which turns his performance from "pretty crap" to "pretty good". Hilariously, the save percentage difference between Datsyuk and Zetterberg is essentially what's driving the fact that there's a 9-goal difference per 60 minutes of play between them, considering Zetterberg has seen more shots against while shorthanded than Datsyuk in only 13 seconds more PK time.
At 4-on-5, Datsyuk is tops among forwards at the shot-attempts-against rate per 60.
If we treat shots on goal as the focus instead of the goals themselves (as neither really accurately account for the pesky and nebulous "shot quality" argument), we can take a look at the expected scores for everybody relative to how well they...or at least how well their entire units do at preventing shots on goal.
You'll notice that Emmerton and Miller switch places for this chart compared to the one above; same for Zetterberg and Cleary. Just remember that the chart above is shot attempts (shots on goal, blocked shots, and missed shots), while the one below only factors shots on goal.
|Player||Expected Goals Against/60|
Now just factoring shots that make it through guys trying to block them and ones which don't ring the iron or miss just-wide doesn't make it any more accurate than just basing on results, but factor in everything and we've at least got a better idea of why Datsyuk's PK rate looks terrible when the eyeball test shows that he's doing a good job. If I had to take only one stat to judge PK effectiveness for individuals, I'd lean more towards shot attempts against per 60, but I don't want to completey downgrade exactly how good Zetterberg has been on the penalty kill this season.
Hopefully the Wings' PK can continue to improve as the season goes along. I like the 85% effectiveness lately, but you're always just one game away from allowing 3 PPGA and ending up right back where you started.