Wednesday, January 30, 2008

Power Play Opportunity Analysis

As I said yesterday, because I've been thinking so much about power play opportunities lately, I decided to take a look at the data over the last 5 years (well, four previous years plus this season) to see if there are any trends based on official or opponent.

One important thing to mention is that I do not believe that you can look at a box score and say "That was a poorly officiated hockey game." It's not fair to the official, who does have one of the hardest jobs in sports. I'm not trying to make a judgment upon the competency of an official solely based on this data. I've got opinions on most of these officials already, however.

I do believe, however, that as a general rule of thumb, the higher-skilled team should generally end up on the power play more often than its opponent. It's not always the case--I'm on record as really liking the games that Sergott has called this year and he didn't exactly favor us in that regard--but I've said for years that I don't understand how we always seem to end up short-handed so much when we play LSSU.

Basically what I wanted to find out is if there are any trends as to which referees tend to give Michigan more power plays, which refs have left us short-handed more often, which refs tend to let us play, what opponents do we take more penalties against, what teams should we expect to be playing a man up against, do we have a home ice advantage with the refs, etc. In the long run, I would fully expect a team as talented as Michigan to end up with significant edges in power plays over most of the teams in the CCHA, with the exception of Michigan State. Michigan and MSU are the only CCHA teams that have proven to be able to sustain success over the long haul, though Miami and Notre Dame are up at the moment. Time will tell if they're here to stay.

I considered including the score of the game as well as the shot totals to see if the team that "carried the game" was getting the benefit of more power plays as well, but I decided that's still not necessarily fair--and it would've been a huge undertaking. There are examples, however, that blew my mind. One that jumps to mind, which wasn't included since it was an exhibition game, came in 2004 when we played Windsor. We won the game 12-1, outshot them 59-14 (including 20-1 in the second period), and yet we had less power plays (9-8). Justify that one to me, Aaron. They had almost as many PPs as shots.

Onto the data. I hope you find it as interesting as I did.

The first thing to look at is the number of power play opportunities we've gotten versus our opponents over the years. Following the totals, I've expressed the number of games that we've received more PPOs, the # of times we had less PPOs, and the number of times they were even (I'm calling this the "record" for the duration of the article).

2007-08: Michigan: 132, Opponents: 141, 8-14-4
2006-07: Michigan: 241, Opponents: 241, 18-19-4
2005-06: Michigan: 254, Opponents: 233, 22-14-5
2004-05: Michigan: 273, Opponents: 244, 21-12-9
2003-04: Michigan: 241, Opponents: 175, 33-3-7

The first thing that jumps out is the 2003-04 season. Holy crap. We were only short-handed more than our opponent on three occasions: The tournament game against BC, a game at MSU, and a game at Alaska. It was the 28th game of the season before we faced more power plays than we were given.

But as much as that jumps out, look at this season. Michigan is (was?) off to the best start in the history of the program, has been #1 for a fair portion of the season, is one of the fastest and most-skilled teams in the country, and, I would point out, they don't have a player that takes an absurd number of penalties. Yet, in over half their games, they've had a deficit in PPOs. This includes games against NMU, UNO, LSSU (twice, including an absurd -7, which I'll get to later), BGSU (twice), and WMU, which is a rarity for that team.

What does it mean? I'm really not sure. Boston College, North Dakota, and New Hampshire also are in the negative, and DU and CC are close. But in the league that seems to call more penalties than any other, the CCHA, the elite teams have benefited. Miami is +20, Notre Dame is +26, and MSU is +24. And yet we're -9. I find it very strange, especially with the "Officials love Michigan" notion that is bandied about by fans throughout the conference.

The analysis of games at home vs. games on the road is probably what you'd expect. Over the five years studied, Michigan is +8 in PPOs on the road with a "record" of 28-24-12. At neutral sites, we're +10 at 17-12-5. But at home, the Yost advantage comes into play. We're +89 in PPOs at home and are 57-26-12 in that span. We've been negative once under each of Aaron (5 games), Hall (9), Hill (2), Hoberg (2), Kotyra (4), Lisko (4), and Murphy (1). Piotrowski has done 22 of our games, we're +52 in that span, and we've been negative just three times--never more than -1. Disclaimer: The road/neutral numbers may be slightly off. I'm fairly confident I counted some of the neutral site Joe Louis games as road games. It likely wouldn't change things though.

Strangely enough, even though Hall has had us in the negative just one time in nine game at Yost, at the six neutral site games and two road games he's done, we've been positive just twice, and even once.

Breaking it down by referee and by opponent is where it gets really fun, however.

I'm sure it will shock no one to hear that Hoberg was the most whistle-happy official that we've had more than once or twice. He's called an average of 16 power plays in the five games he's done, and never less than 11. The CCHA ref that lets teams play the most? Matt Shegos actually, at 9.6 power plays per game. For the record, this "Wolverine lover" has done 26 of our games in the last 5 years, and we're +13 in power plays (about one additional PP per weekend). Amazingly, since it's the Sparties that love to accuse him of being Wolverine fan #1, we're actually -4 in the nine Michigan/MSU games that he's done.

Here's the breakdown by referee (PPO/game numbers are for the teams combined):
Aaron: 12.2 PPO/game, +11, 10-3-4 over 17 games
Brown: 11.2 PPO/game, +4, 3-0-2 over 5 games
Hall: 10.8 PPO/game, +4, 7-6-4 over 17 games
Hill: 11.3 PPO/game, -7, 1-3-0 over 4 games
Hoberg: 16 PPO/game, +2, 2-2-1 over 5 games
Klosowski: 8.7 PPO/game, +4, 1-0-2 over 3 games
Kotyra: 10.8 PPO/game, +3, 2-2-1 over 5 games
Lisko: 10.6 PPO/game, -8, 3-4-0 over 7 games
McInchak: 13.4 PPO/game, +9, 15-9-3 over 27 games (that's too much McInchak FTR)
Murphy: 18 PPO/game, -4, 0-1-0 over 1 game
Piotrowski: 10.4 PPO/game, +66, 27-7-4 over 38 games
Pochmara: 14.5 PPO/game, +7, 3-3-0 over 6 games
Sergott: 13 PPO/game, -1, 1-2-0 over 3 games
Shegos: 9.6 PPO/game, +13, 16-8-2 over 26 games
Wilkens: 10.8 PPO/game, +6, 6-7-5 over 18 games

Since, at least to my knowledge, everyone around the conference has viewed Piotrowski as the standard of excellence in officiating over the years, I think it's pretty telling when you look at the numbers by his name. None of the other refs that we frequently have (double digit times over 5 years) is higher than +.65/game, whereas Pio is at +1.74/game for Michigan. Dare I say that's probably closer to how it should be, given that Michigan is always in the top 3 in the conference?

And before he's accused of being nothing but a Michigan-lover, he did 4 Michigan/MSU games (that's all?) and we're -3 in those games.

Hill and Lisko (at least on paper) deserve the biggest punches in the face. Brian Aaron would be up there but I can't track waived-off goals. We're 4-7-0 in PPOs when we've had Hill or Lisko and are -15 in those games. So you figure they've done some games where we've played against top-notch competition and we took some penalties due to the opposing speed right? Wrong. Here are our opponents in those games: BGSU four times, WMU three times, LSSU three times, and Quinnipiac. We were negatives in all five road games that those two did.

As for the breakdown by opponent, this stems from my anger over the fact that it seems like we always seem to be short-handed when we play LSSU. It's completely inexplicable. My theory has been that officials can't possibly call a team for everything when they clutch and grab all the time, so they wait for a flagrant violation to whistle them. Meanwhile, when they've got a team that doesn't play that way, the little things are more obvious, so they get called for those. Personally I think there's some merit to it.

But yeah, so I was mainly curious to see what the breakdown of our games against LSSU has been. And I was right. In the last 10 games against LSSU, we're incredibly -7, and are 4-4-2 with regard to PPOs. Less surprisingly, we were +4 and 2-0-0 when Piotrowski called our games with them.

The breakdown by opponent:
Alabama-Huntsville: 17 PPO/game, -5, 0-1-0 in 1 game (thanks, Hoberg)
Alaska: 13.4 PPO/game, +14, 7-3-3 in 13 games
BC: 9 PPO/game, +4, 2-1-1 in 4 games
BGSU: 13.1 PPO/game, +5, 6-5-1 in 12 games
BU: 10.7 PPO/game, +2, 2-1-0 in 3 games
CC: 12.5 PPO/game, -5, 0-2-0 in 2 games
Connecticut: 20 PPO/game, -2, 0-1-0 in 1 game
Ferris: 11.3 PPO/game, +12, 6-3-3 in 12 games
LSSU: 10.1 PPO/game, -7, 4-4-2 in 10 games
Mercyhurst: 11 PPO/game, +1, 1-0-0 in 1 game
Merrimack: 18 PPO/game, +6, 1-0-0 in 1 game
Miami: 12.4 PPO/game, +10, 7-3-0 in 10 games
Michigan Tech: 12.6 PPO/game, +7, 4-1-0 in 5 games
Minnesota: 10.8 PPO/game, -1, 3-2-1 in 6 games
MSU: 9.0 PPO/game, -9, 9-10-3 in 22 games
Nebraska-Omaha: 9.6 PPO/game, +6, 6-5-4 in 15 games
New Hampshire: 10.5 PPO/game, +1, 1-0-1 in 2 games
Notre Dame: 11.6 PPO/game, even, 6-6-3 in 15 games
Northeastern: 16.3 PPO/game, +3, 1-2-0 in 3 games
Niagara: 11 PPO/game, +3, 1-0-0 in 1 game
NMU: 10.8 PPO/game, +23, 10-2-2 in 14 games
NoDak: 11.5 PPO/game, -1, 1-1-0 in 2 games
OSU: 10.8 PPO/game, +19, 10-2-2 in 14 games
Providence: 9 PPO/game, -1, 0-1-0 in 1 game
Quinnipiac: 13.5 PPO/game, +8, 3-0-1 in 4 games
St. Lawrence: 15 PPO/game, -1, 0-1-0 in 1 game
Western Michigan: 10.8 PPO/game, +8, 7-3-2 in 12 games
Wisconsin: 10.8 PPO/game, +7, 4-2-0 in 6 games

Now that I think about it, you can't really use PPOs to determine which games are the most chippy, as I had stated yesterday. Those type of games are going to have a lot of scuffles after the whistle which, unless Shegos is reffing, don't usually lead to a team being short-handed. Proof: The games against MSU actually have the least amount of power plays.

The teams which we have a significant edge against: NMU, OSU, and Alaska. We've got decent edges over Miami, Ferris, and Western. Seems appropriate for the most part (especially when you consider that the Notre Dame and Miami series have swung back in their favor since they got good recently).

It's data like the above which makes that LSSU number so odd to me. I can see being in a deficit against MSU (though we've never been above +2 in those games, and that's only happened three times, whereas we've been -3 or worse on four occasions). We've been short-handed lately against Miami and Notre Dame. But Lake State? I don't understand that one. The degree of those negatives is strange as well: -4, -6, -7, -2. How exactly is this happening? How does that team get power plays? Let alone 7 more than a team like Michigan. Weird...

The Bowling Green one seems a bit low too. Then again, we've been -3, -1, -3, and -3 the last four times we've played them.

I think most everyone agrees that we've got a problem in the CCHA: The only good ref (Piotrowski) has retired. And the data, at least in my opinion, seems to back that up. I'm not asking for Michigan to always have the upper hand, but it seems that the other officials around the league try and even out the penalties more often than not, regardless of the skill level of the teams or how the game is going. I also believe they have the tendency to "ref the score" so to speak, which is equally annoying.

It's pretty telling to me that in the 38 games that Pio--largely known as the best ref in the CCHA--did, we were given more PPOs 20 more times than our opponent was (27-7-4). In the 155 games that Pio didn't officiate, we were given more PPOs 20 more times than our opponent (75-55-25). It's kind of amazing.

I've uploaded the spreadsheet here, if anyone else would like to play around with it and see if anything jumps out that I've neglected to mention. I plan on maintaining my local copy so that I can use the data in weekend previews and whatnot.

8 comments:

streaker said...

Tim, that a mountain of work... do you have a life? ;)

Great analysis... alot of backup to battle the "Michigan gets all the calls" woofing that our (mostly) in-state rivals like to use.

Well, we'll see who we get this weekend against NMU.

Streak

Kurt said...

Tim,

Thanks for that. I can just see your annoyance with these results and (yes I am biased) you support your argument quite well. Normally, I would think that many of the results may be skewed with teams taking liberties late in the game and refs not calling it but getting us for retailiation (so many of the WMU, OSU games...) but the results seem to contradict that. Nice job with the number crunching.

THE UNH MEN'S HOCKEY BLOGGER GUY said...

Nice work!

Welcome to the 'penalize the top team' zone!

THE UNH MEN'S HOCKEY BLOGGER GUY said...

Make sure you get Michigan fans to vote for Porter, even on the Gerbe For Hobey page!

http://gerbe-for-hobey.blogspot.com/

Chris Willie Williams (your brother) said...

Hey Tim,
Thanks for telling me about this entry when we were talking yesterday. It's really impressive, and as always, you back up your arguments like nobody's business.

One thing that annoys me:

It's "fewer PPOs," not "less."

Sars from Tomato Nation explains:
Use "less" for quantity and "fewer" for number. Huh? Okay. You go shopping. You need apples. You don't see as many apples as you saw the last time you went to the store, so you see…what? Fewer apples. That's right. But maybe the store reduced its produce selection, so there's not as much fruit overall. So there's…what? Less fruit. Good. "Many/much" is analogous, and you can use it as a cheat to figure out which one is correct; if it's something you can enumerate, you'd use "many," and if you use "many," "fewer" is correct. If it's something you quantify, you'd use "much," and when you use "much," "less" is correct. Whew — that's complicated. Here's a simpler example: less noise, fewer decibels. Got it? En masse, it's "less." Individual units, "fewer."

Just thought I'd annoy you with a grammar lesson this morning.
-Yr Brother

Packer487 said...

Silly English majors :-P

And streaker, no, no I don't...

Yellow Blade said...

Chúng tôi cung cấp các dịch vụ vận chuyển gửi hàng từ nước ngoài về VN và ngược lại như Dịch vụ vận chuyển gửi hàng từ Việt Nam đi Pháp, ship hàng từ úc về việt nam, mua hàng Pháp, mua hàng Trung Quốc, Dịch vụ vận chuyển gửi hàng từ Việt Nam đi Úc... Hầu như ở mỗi nước chúng tôi đều cung cấp 2 dịch là chuyển về và chuyển đi. Có thể lấy ví dụ như ở Anh. Chúng tôi có dịch vụ gửi hàng đi Anh và mua hàng Anh. Ngoài chúng tôi còn phục vụ mua hang myvan chuyen hang di my. Để biết thêm chi tiết hãy liên hệ với chúng tôi nhé.

Quốc Lê Bảo said...

VietAir Cargo chúng tôi chuyên cung cấp ship hàng Mỹ và dịch vụ gửi hàng đi mỹ,gửi hàng đi mỹ giá rẻ, chuyển hàng đi mỹ, vận chuyển hàng đi mỹ, .... Chúng tôi còn là đơn vị hàng đầu Mua hàng trên amazon, Mua hàng amazon, Mua hàng trên ebay, Mua hàng ebay, mua hang my, mua hàng mỹ, mua hộ hàng mỹ. Hãy liên hệ ngay với VietAir Cargo chúng tôi nhé