Saturday, May 06, 2006

PP tidbit #2

The way the NHL calculates PP and PK efficiency is a pet peeve of mine. The goals/opportunity method of determining the efficacy of a PP or PK unit is extremely misleading. Furthermore, it just so happens that the NHL doesn't filter out the EN goals out of these numbers either further compounding the issue in my opinion.

Recently, mudcrutch has posted his situational team TOI numbers for the 2005-2006 regular season here. Since he's already done this, it's really easy for me to just steal the data and do something with it.

The first thing is to just take out the EN goals. It doesn't really change things all that much but it does tweak it a little bit for a few teams including the Oilers, who scored a couple of EN PP goals this year. Here are the top 10 in NHL and adjusted NHL PP efficiency:
The next thing to do is recalculate PP efficiency as PPGF/hr of PP icetime or PPGD/hr of PP icetime using mudcrutch's team TOI numbers. This approach to PP efficiency is superior because it effectively weights the PP opportunities by the number of PP minutes. Essentially this gets rid of the situation where a team gets stuck with a 10 sec man advantage and an 0/1 on the PP that it probably couldn't avoid no matter how dangerous their PP unit happens to be.

Lastly, I favour the PPGD/hr number. Why? The other numbers junkies out in cyberspace have a perfectly good reason for trying to ignore them most of the time. They pollute the +/- stat and they happen fairly rarely compared to the other situations in which goals are scored. The SH goal is also frequently touted as a "momentum changer" by hockey traditionalists, so I think they like to ignore it particularly because the mainstream thinks they are such critical events. ;) However, I do think they matter to evaluating special teams - some teams score them in piles, and therefore, some teams give up a significant number of them as well. The top 10 lists for both measures (plus the Oilers and Sharks) are here:

This changes the lists somewhat, especially once the SHGA get factored in. It's interesting to me that both Sutter and MacT like to use 2 defensemen on the PP a significant amount of time, and yet they aren't really able to avoid SH goals against much more than teams like the Rangers or Thrashers that routinely employ 4 or 5 forwards.

The next step for evaluating the PP is to find out the 5-on-4, 5-on-3 and 4-on-3 icetimes for each team. This would allow a better idea of how teams are getting their PPGD/hr. As of right now, all I can do is show the percentage of goals a given PP unit has scored against 3-man PK units (ie. both 5-on-3 and 4-on-3 advantages). I'm not sure if it really does bear out but based on the number of 4-on-3 goals scored, the likelihood that there's very little 4-on-3 icetime and by my eye, it seems to me that the 4-on-3 advantage is better than the 5-on-4. Admittedly, this might be a erroneous assumption.


The league average for %PPGF vs. 3-man PK units was 17.8%. I think what you'll find (maybe mudcrutch's script can lift these numbers form NHL.com?) is that the teams with the higher percentages probably had significantly more icetime against 3-man PK units and therefore had an easier time of it with respect to scoring PP goals.

Basically, if you're comparing the Shark and Oiler power plays, I think the edge should go to the Oilers. They had fewer SH GA and less than league average %PPGF against 3-man PK units.

The last thing left for adjusting PP efficiency would be to handicap the numbers for the PK units they had to face. I won't put up anymore tables here, but the Oilers' NW division opponents had a SH GD/hr of -5.26 while the Sharks' Pacific opponents put up an SH GD/hr of -6.26. Granted, the Sharks had something to do with that number, but I think the 1 GD/hr difference probably means that the Sharks had an easier time in their own division anyway in terms of facing solid PK units.

2 Comments:

Anonymous The Rage said...

It seems like Edmonton has the definate edge in terms of Special teams (especially when you consider penalty killing). They also have the edge in goaltending. As long as they can keep things relatively even five on five, then I think the Oilers have the slight edge in this series.

5/06/2006 11:09 pm  
Blogger sacamano said...

Man the intarweb is cool. What a terrific post.

5/07/2006 8:46 am  

Post a Comment

<< Home