Monday, July 26, 2010

Penalties: Part 2 - Hooking and Intereference

So last time, the big news was that hooking and interference were down by 27% from the second period to the third period in the 2009-10 NHL regular season.

In this post, we'll take a look at some of the refs who appear to be responsible for this sharp inconsistency.

First off, a few general numbers:
  • Hooking and Interference were #1 and #4 on the overall list of penalties called (excluding Fighting, which isn't called at the discretion of the referee).
  • H&I (Hooking and Interference) dropped by 27% from the second to the third period.
  • All other calls (excluding Fighting) dropped by just 8% combined.

Why would this be the case? Is it consistent across all of the refs?

I can answer the second question pretty clearly.

First, a quick note about these numbers - there are two referees at each game and the NHL does not identify which referee made the call (or, at least, they don't publish it). That means the best we can do is "associate" each ref with the penalties that were called while they were on the ice. Despite the lack of individual accountability, I think the results are still interesting.

Given that the average drop is 27.49% (which is way too high in my mind), we'd expect to see some guys on each end of the spectrum.

In the charts below, the numbers for each period are the number of Hooking and Interference calls made in games worked by the ref. The % Change shows the percentage difference between the second and third period.

Here are the more consistent refs:
Referee First Second Third OT % Change
#18 Greg Kimmerly 48 37 43 2 16.22%
#41 Chris Ciamaga 19 19 20
5.26%
#5 Chris Rooney 59 54 52 1 -3.70%
#9 Dan O'Rourke 52 64 59 1 -7.81%
#43 Brian Pochmara 54 72 65 2 -9.72%
#13 Dan O'Halloran 51 57 51
-10.53%
#11 Kelly Sutherland 56 53 47 2 -11.32%
#20 Tim Peel 57 57 49 3 -14.04%
#28 Chris Lee 58 60 51
-15.00%
#32 Tom Kowal 52 68 57 2 -16.18%
#3 Mike Leggo 57 51 42 2 -17.65%
#27 Eric Furlatt 82 72 58 1 -19.44%

The actual number of calls isn't very useful here since I haven't shown the number of games worked by each ref. The thing you can count on, however, is that if one of these guys is wearing the stripes, you're more likely to see H&I called in the third.

Here is the other end of the ref pool. These fellows have what I consider to be truly frightening numbers because they are wildly inconsistent between the second and third period for H&I:

Referee First Second Third OT % Change
#45 Justin StPierre 81 89 40 2 -55.06%
#44 David Banfield 17 27 14 1 -48.15%
#25 Marc Joannette 42 66 35 1 -46.97%
#15 Stephane Auger 90 85 47 2 -44.71%
#48 Frederick L'Ecuyer 36 34 19 1 -44.12%
#7 Bill McCreary 63 60 34 2 -43.33%
#6 Dan Marouelli 37 44 26 1 -40.91%
#38 Francois StLaurent 58 80 48
-40.00%
#26 Rob Martell 63 58 35 2 -39.66%
#23 Brad Watson 68 73 45 1 -38.36%
#14 Dennis LaRue 57 60 38
-36.67%
#10 Paul Devorski 44 71 47 1 -33.80%

You have to wonder what some of these guys are thinking sometimes. StPierre, Auger, McCreary, Martell, Watson and LaRue are pretty consistent through the first two periods and then largely put the whistles away for the third.

You have to think the NHL has these numbers. You also have to think they're aware of the tendencies of these guys. Heck, they probably know (unlike us) which ref actually made the call, which would make this case even more strongly.

The problem with these numbers is that the players don't know what to expect from one night to the next. They get one night with a Kimmerly or a Rooney and the calls are pretty consistent across the entire game, and then you get an Auger or StPierre and the third period deteriorates into clutch and grab hockey.

Pretty interesting, I think. Thoughts on this?

This will be cross-posted to Hockey Independent (albeit in a summarized fashion with Part 1) later in the week (maybe later tonight if I can get my act together).

Monday, July 19, 2010

Penalties: Part 1

This is the first part in a series on penalties in the NHL. The data comes from various sources on the web, much of it from NHL.com in their play-by-play reports on each game.

In the 2009-2010 season, there were 12328 penalties called, which works out to about 10 per game (as there are 1230 games in each regular season).

The breakdown by period is:
  • First period: 4062
  • Second period: 4420
  • Third period: 3745
  • Overtime: 101
So, right off the bat, it looks as though the "they put away the whistles in the third" idea has some merit. We'll take a closer look at that later in the this post, and later in the series when we look at individual referee data.

The top 10 penalties called in 2009-2010 were:
Penalty Calls
Hooking 1750
Roughing 1502
Fighting 1423
Tripping 1412
Interference 1065
Holding 953
Slashing 783
Hi-sticking 747
Cross checking 483
Boarding 310

Some notables that are missing from the top 10 include: Delay of game (puck over glass), Too many men, Goaltender interference, Holding the stick and the bloody version of Hi-sticking.

All told, there were 50 distinct types of penalties called in 2009-2010 (I was surprised how many different ones there were). Here are a few that were almost never called:
  • Spearing was called 3 times, and one additional time as a double minor.
  • Abusive Language - Bench was called twice
  • Just one Illegal Stick penalty, on Brett Carson of the Hurricanes at 18:41 of the third period in a 5-1 blowout over the Rangers. Not sure the point of that one was...
Now let's take a look at the most commonly called penalties, broken out by period (I combined both the regular and double minor versions of Hi-sticking):

Period
Penalty 1 2 3 OT Total
Boarding 94 109 104 3 310
Cross checking 147 172 160 4 483
Delaying Game-Puck over glass 91 94 106 1 292
Fighting 674 454 295
1423
Hi-sticking 257 277 298 12 747
Holding 293 353 300 7 953
Holding the stick 54 67 42
163
Hooking 594 661 480 15 1750
Interference 376 394 287 8 1065
Interference on goalkeeper 67 85 79 2 233
Roughing 440 541 512 9 1502
Slashing 210 303 261 9 783
Too many men/ice - bench 74 104 60 3 241
Tripping 433 538 421 20 1412

We already saw previously that calls from the second period to the third period dropped from 4420 to 3745, a 15% drop. Most of these are judgement calls on the part of the refs. You can see they are pretty consistent on many of them - Roughing, Too many men, Boarding, Cross checking, Hi-sticking, Slashing, etc.

Take a second look at three calls that have a definitive impact on the speed and flow of the game - here they are with the drop from the second to the third period:
  • Hooking - dropped by 27%
  • Interference - dropped by 27%
  • Tripping - dropped by 22%
I'm willing to let Tripping drop - it's essentially flat with the first period, so I could argue that the long change has something to do with that.

Hooking and Interference calls, however are way down from the first period numbers too. You can't tell me that it's not happening as much in the third period - I don't buy it. I think the refs are exercising their "judgement" (such as it is) instead of calling what they see on these penalties.

Any thoughts on the analysis?

Friday, July 16, 2010

Sidney Crosby and Faceoffs: Part 3

So for the last part, I thought I'd take a look at which players Sid had the most success against (or lack thereof).

With the linesman, we couldn't be certain who had dropped the puck, so we couldn't say for sure what the actual success rate was. With the players, it's obviously very specific and certain.

These numbers speak for themselves. I'm not showing anyone he faced less than 10 times, because I think the sample size is too small. That leaves us with 58 players Sid faces 10 times or more in 2009-2010.

Here are the top 10 guys that Sid dominated in the faceoff circle:
Player Wins Losses Total Win %
23 DRURY 29 7 36 80.56%
17 CARTER 16 4 20 80.00%
19 CONNOLLY 12 4 16 75.00%
26 CHRISTENSEN 8 3 11 72.73%
19 THORNTON 8 3 11 72.73%
91 STAMKOS 13 5 18 72.22%
27 REINPRECHT 14 6 20 70.00%
47 PEVERLEY 14 6 20 70.00%
9 WEISS 23 10 33 69.70%
11 BETTS 18 8 26 69.23%

Poor Chris Drury - he really had a hard time against Sid this season. Jeff Carter isn't far behind (just for reference, Sid was 33-21 against Mike Richards too).

And on the other end of the spectrum:
Player Wins Losses Total Win %
9 MORRISON 3 9 12 25.00%
42 ANISIMOV 5 12 17 29.41%
28 GAUSTAD 8 18 26 30.77%
91 GOMEZ 5 10 15 33.33%
37 BERGERON 17 28 45 37.78%
50 VERMETTE 4 6 10 40.00%
46 KREJCI 7 10 17 41.18%
14 PLEKANEC 12 17 29 41.38%
23 SLATER 5 7 12 41.67%
9 ROY 5 7 12 41.67%

Gaustad, Bergeron and Plekanec were easily his toughest challenges all year, because he faced them with regularity and wasn't terribly good against them. I don't have time to do it now, but it would be interesting to know if he struggles against right-handed players more than left...

Sidney Crosby and Faceoffs: Part 2

This part will take a look at how Sid fared with games worked by specific linesman.

I always wondered if the way the linesman drops the puck has anything to do with who wins the faceoff. Maybe there was a pattern that some guys could pick up on and use to their advantage. Maybe some linesman allow some guys to cheat more and gain an advantage.

Despite the fact that we're talking about a very small percentage difference between average (50%) and very good (Sid won about 56%), the sheer volume of draws makes any small advantage significant. Sid took almost 1800 draws. The difference of a percentage point would be 18 more wins, so for Sid, he won 108 draws more than average. Since the Pens play a puck possession style, that number is significant.

So, in looking at the linesman, there is no way to identify who is dropping the puck. There are two linesman working each game and the data doesn't specific who drops each puck.

What we can do, however, is to "associate" the linesman with the faceoffs, so that in games worked by Marc Pare, we know that Sid has a specific success rate (53.7%).

Here are the top 10 linesmen Sid was most successful against:
Linesman Total # Win %
#97 Jean Morin 107 66.36%
#70 Derek Nansen 70 62.86%
#47 Dan Schachte 53 62.26%
#68 Scott Driscoll 93 61.29%
#59 Steve Barton 85 61.18%
#75 Derek Amell 231 58.87%
#50 Scott Cherrey 89 58.43%
#57 Jay Sharrers 86 58.14%
#66 Darren Gibbs 43 58.14%
#96 David Brisebois 156 57.69%

If any of these guys are announced as the linesman for a game, Sid is likely to have a good night.

On the flip side, here are the guys he has the least success against:
Linesman Total # Win %
#73 Vaughan Rody 62 46.77%
#88 Mike Cvik 49 48.98%
#55 Shane Heyer 42 50.00%
#54 Greg Devorski 159 50.31%
#90 Andy McElman 129 50.39%
#77 Tim Nowak 68 51.47%
#65 Pierre Racicot 33 51.52%
#92 Mark Shewchyk 64 51.56%
#61 Lyle Seitz 44 52.27%
#86 Brad Lazarowich 17 52.94%

You only really need to worry about linesman he faced more than 100 times on this list. I'd say the other numbers are insignificant due to small data size.

The last piece of information I tried to get is how many times Sid was kicked out of the faceoff. This data is not recorded anywhere, but I looked at any faceoff where Sid was on the ice but did not take the draw. I think in those situations, it is normally safe to assume that he had been kicked out of the draw.

According to that methodology, Sid was kicked out 63 times during the 2009-2010 season. Here are the top 5 guys that kicked him out of the draw:
Linesman Faceoffs Waved Out% Win %
#66 Darren Gibbs 43 8.51% 58.14%
#57 Jay Sharrers 86 7.53% 58.14%
#79 Mark Pare 149 7.45% 53.69%
#71 Brad Kovachik 148 6.92% 56.76%
#78 Brian Mach 179 6.77% 57.54%

You can see it didn't appear to have any impact on the win % as Sid was above his season average on all but one of those guys.

On the other end of things, Sid was waved out a grand total of once all year against these five guys (though the volume is low)
Linesman Faceoffs Waved Out% Win %
#47 Dan Schachte 53 0.00% 62.26%
#68 Scott Driscoll 93 0.00% 61.29%
#61 Lyle Seitz 44 0.00% 52.27%
#65 Pierre Racicot 33 0.00% 51.52%
#96 David Brisebois 156 0.64% 57.69%

I think the Driscoll number is significant. Sid is well above average against him with a decent number of faceoffs.