campbellandwales

A Final Look at Special Teams in 2009-10

In Hockey on September 5, 2010 at 4:29 pm

By Ryan Wagman

As I’m sure many of you have waited with bated breath for my final special teams rankings for the 2009-10 season, I would like to start off with an apology. With the post-season, comes a certain malaise, born of the knowledge that no matter how much we can write about hockey, think about hockey or argue/fight about it, there is no hockey. Just backroom drama.
So, too, is there backroom drama within this writer’s life. Since my last entry, my job description has changed drastically, I took on other hockey writing projects, first with draftamerica.com and now with premiumscouting.com, and I managed to squeeze in a short vacation in San Francisco with the Mrs.
Now tomorrow is Labour Day and many pre-season hockey rags are already out. My former colleagues at Hockey Prospectus (soon-to-be-formerly Puck Prospectus) are about to release their first ever annual.So without any further ado, (and no real commentary) I give to you last season’s final rankings.
Power Play Efficiency (the average time between goals when up by a man. Two man advantages are double-counted in time)
1) Was 382.266
2) SJ 449.877
3) Mon 453.719
4) Van 463.319
5) Phi 476.621
6) LA 476.797
7) Ana 477.190
8) TB 513.841
9) Det 522.593
10) NYR 534.200
11) Dal 535.525
12) Min 536.328
13) NJ 536.765
14) Clm 549.800
15) Col 551.429
16) Chi 560.442
17) Pit 568.643
18) Buf 571.945
19) Edm 583.115
20) StL 584.226
21) Car 594.589
22) Ott 595.449
23) Bos 612.818
24) Nas 620.213
25) Cal 653.023
26) Atl 653.520
27) NYI 658.061
28) Pho 686.000
29) Fla 700.467
30) Tor 758.523
Penalty Kill Efficiency (counted as with the Power Play, but in reverse)
1) StL 784.044
2) Buf 772.816
3) Bos 760.838
4) Chi 715.737
5) SJ 657.420
6) Ott 648.980
7) Pit 646.404
8) Pho 641.816
9) NYR 640.300
10) Det 627.302
11) Mon 594.642
12) NJ 592.000
13) Cal 585.778
14) Phi 585.649
15) Atl 582.561
16) Min 567.113
17) Van 543.220
18) Clm 531.738
19) LA 517.305
20) TB 511.831
21) Car 511.532
22) Col 503.083
23) Fla 485.793
24) Ana 473.821
25) Was 472.851
26) Edm 451.701
27) Dal 431.338
28) Nas 428.103
29) NYI 407.239
30) Tor 384.123
And the combined ranking, being the power play efficiency number, minus the penalty kill efficiency number. The lower the number, the better the organizations’ special teams were last season. This is as it is desirable to go longer between power play goals allowed by your team’s penalty killers, while you hope your team can score power play goals as often as possible
1) SJ -207.543
2) Buf -200.871
3) StL -199.818
4) Chi -155.295
5) Bos -148.020
6) Mon -140.923
7) Phi -109.028
8) NYR -106.100
9) Det -104.709
10) Was -90.585
11) Van -79.901
12) Pit -77.761
13) NJ -55.235
14) Ott -53.531
15) LA -40.508
16) Min -30.785
17) TB 2.010
18) Ana 3.369
19) Clm 18.062
20) Pho 44.184
21) Col 48.346
22) Cal 67.245
23) Atl 70.959
24) Car 83.057
25) Dal 104.187
26) Edm 131.414
27) Nas 192.110
28) Fla 214.674
29) NYI 250.822
30) Tor 374.400
OK, so I lied about the commentary. Now would be a good time to look at how my special team efficiency socres differ from the common version’s results.
Let’s start with the power play numbers. The traditional measures also had the Capitals as sporting the game’s best power play, clicking 25.2% of the time. That worked out to be over 15% better than the cluster of teams between 20.9-21.8%.
In that case, we agree again, as Washington’s power play score was also just over 15% better than the 2nd-ranked Sharks’ unit. At the other end of the spectrum, the traditional system does not quite appreciate how bad the lowly Leafs’ power play was last year. Scoring 14% of the time, they seemingly finished just below Florida, a difference of less than 1.5%. Looking at the game on a mor granular level, as I have attempted to do, shows the Buds to have fallen behind the Panthers by a much wider margin, being 7.65% less effective than Florida. The actual rankings don’t vary too much between the traditional system and mine, unless you’re a Sharks fan (move from 4th-2nd) or support the Rangers (13th-10th), but the granularity is interesting.
On the penalty kill, the changes in raw ranking are minimal, generally being the difference between placing in tight clusters, such as the Coyotes dropping from 6th in the traditional method to 8th here. They were in a cluster with the Rangers, Senators and Penguins that was separated by 0.4% in the traditional method and 8.5 seconds of efficiency here. Unlike the power play, there was not a single team that breezed past its peers like the Capitals. The Blues, leaders on both forms of measurements, were 1% more efficient than the 2nd-ranked Sabres in the traditional method and the same here. On the bottom, the Leafs (again – that must have been historically bad among special teams), were around 2% less likely to kill a penalty than the 29th ranked Islanders in the traditional method, while the granular data showed that they were, in fact, nearly 6% less efficient at killing penalties than the Isles, or any other team.
Looking at the universal special teams’ rankings, I never could have expected such a spread between best and worst of 581.943. Even if we remove the Leafs (I wish I could forget), we still end up with a number of 458.365. In seconds, that’s over 7.5 minutes of efficiency difference between the great San Jose and the poor Islanders. Nearly ten minutes if we include the Maple Leafs.
Before the 2010-11 season gets underway, let’s ponder the numbers and compare them to this summer’s transactions – did your team adjust based on their weaknesses in special teams play? How responsible was Chris Mason for the Blues’ ability to kill penalties? Evgeny Nabokov for the Sharks? Will a full season of Ilya Kovalchuk raise the Devils’ power play? Will a full season of Dion Phaneuf and Phil Kessel and a healthy Mike Komisarek and the absence of Vesa Toskala improve the fortunes of the Maple Leafs? I could go on, but you get the picture.
I hope to continue to track special teams efficiencies during 2010-11, to see if we can learn more, and frankly, because no one else is doing it.
Happy hockey everyone.
Advertisements
  1. Hi Ryan,

    I don’t know if I’ve commented on this before, but I think
    you should invert the numerator and denominator of your
    efficiency equation, and express it as goals / 60 min,
    or something like that. The reason is that you cannot
    sum “efficiencies”, as you are doing, while you can
    sum goal differential (because hockey games are scored
    by goal differential, not by efficiencies). Also, the
    units are not very intuitive: I doubt anybody but you
    understands what it means that there is an efficiency
    difference of 7.5 minutes between San Jose and the Islanders.

    Also, if you are going to quote the traditional percentages,
    you should do it consistently with your stats: for example,
    when comparing the penalty killing of Toronto and Long
    Island, you’d be better off saying that Toronto was scored
    against at a 25.3% rate and Long Island at a 23.7% rate.
    That means that Long Island was about 7% better, the exact
    same conclusion you come to.

    All in all, measuring special teams by time rather than
    opportunities is a strict improvement, so I welcome
    whatever work you do in this.

    • Tom,
      Thanks for the feedback. I can’t remember hearing your comments before, but I think your suggestion to invert numerator and denominator is good – much more intuitive and easier to measure.
      I’ll add that to my work sheets for this season. I want a few seasons of data before I begin to draw further conclusions.
      Cheers,
      Ryan

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

%d bloggers like this: