HFBoards

Go Back   HFBoards > NHL Western Conference > Pacific Division > Edmonton Oilers
Notices

PP Ice time Stuff

Closed Thread
 
Thread Tools
Old
09-27-2003, 01:46 PM
  #1
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
PP Ice time Stuff

Ok, so I'm busy procrastinating, and I thought I'd kick around some of the PP ice time stats, to see if I could shake anything interesting loose.

I've been kind of thinking about this for a while, and I have a couple of areas that I'd like to pursue here. The first relates to what percentage of a forwards ice time is PP time. I think that this can tell us a few things. First off, what forwards are either so good on the PP that the team reserves them for that situation, or so bad at even-strength play that the team reserves them for that situation. Just as a general bit of info, I've confined myself to forwards who played more than 120 minutes on the PP over the course of last season. There were 195 forwards who fit that description last year, which works out to roughly six per team. The average player in this group spent 74.33% of his time on the ice at even strength, 19.68% of his time on the PP, and 5.89% of his time on the ice shorthanded. The breakdown of players by team is as follows:

ANA-8 ATL-7 BOS-5 BUF-6 CAL-5 CAR-7 CHI-6 CLB-6 COL-6 DAL-8 DET-8 EDM-8 FLO-7 LA-4 MIN-8 MON-4 NAS-8 NJ-6 NYI-5 NYR-7
OTT-7 PHI-8 PHO-6 PIT-4 SAN-5 STL-8 TB-7 TOR-9 VAN-6 WAS-6

Some notes on this. Players are shown as being with the team that they ended the situation, so teams that engaged in fire sales may have lower numbers. Hellooooo Pittsburgh. I think you can also look to injuries as being factors that drive the number higher as well-Edmonton is an example of that. I found Boston to be the most interesting-they have a couple key guys that they rely on very heavily.

Anway, moving on we can look at how players vary from the "average" player in terms of ice time. Lets look at the top 20 for % of ice time coming on PP.

Code:
PLAYER	             PP/TOI
RONNING	             29.57%
WHITNEY	             29.13%
MOROZOV	28.28%
WEIGHT	             27.66%
FRANCIS	             27.26%
NASLUND	             27.22%
BERTUZZI	26.66%
RICHARDS	26.17%
SANDERSON	26.09%
JOHANSSON	25.97%
O'NEILL	             25.94%
PARRISH	             25.85%
YASHIN	             25.81%
LEMIEUX	             25.43%
BRUNETTE	25.35%
KOVALCHUK	25.30%
HAVLAT	             25.25%
HEATLEY	             25.21%
ANDREYCHUK	25.05%
BONDRA	             24.80%
Alright, so I can hear many of you saying, well, so what? Why is this relevant? it is incredibly relevant when evaluating a player as a potential acquisition for your team.

Let's take Whitney as an example (full name: Ray Whitney whosedadfloydpracticesasanextragoaliefortheOilers) . As you can see from the chart, he played nearly 10% more than the league average in terms of PP time in relation to ice time. Detroit signed him in the offseason, drastically changing the circumstances under which he will operate. Lets take a look at the variety of factors that would have altered the season he put together last year. First off, ice time.

Whitney played exactly 21 minutes per game last year, numbers he is not likely to repeat in Detroit. Looking at it from a macro level, what could we expect as his ice time fell? Assuming his mix stays constant, here is what his ice time would look like at various percentages.

Code:
	TOI	ES	PP	SH
100%	21.00	12.82	6.12	2.05
90%	18.90	11.54	5.51	1.85
80%	16.80	10.26	4.90	1.64
70%	14.70	8.97	4.28	1.44
60%	12.60	7.69	3.67	1.23
50%	10.50	6.41	3.06	1.03
In order to get something out of this, we're going to need to bring in his levels of production-specifically, at what rate does he produce, given various situations? In 1701.12 minutes of total ice time last year, he scored 24 goals and 52 assists, giving him a GPM of 0.0141 and an APM of 0.0305. He had 14 even strength goals for an ESGPM of 0.0135 and 17 ESA for an ESAPM of 0.0164. He had 8 power play goals for a PPGPM of 0.0161 and 34 PPA for PPAPM of 0.0686. He had 2 SHG for a SHGPM of .012 and 1 SHA for a SHAPM of 0.0060. I'm just going to put these in a table so that its a bit easier to see. This is key to where I'm going next.

Code:
ESGPM	ESAPM	PPGPM	PPAPM	SHGPM	SHAPM
0.0135	0.0164	0.0161	0.0686	0.012	0.006
Ok so now we know how he produces in certain situations, in addition to knowing how often he experiences certain situations. Now, what would happen if he we changed the variables? I'm going to suggest that a player's situation is much easier/likely to change than is his ability to produce. First, let's look at what he would do if he had the same mix of ice time, in various percentages of last years ice time.

Code:
%	ESG	ESA	PPG	PPA	SHG	SHA	PTS
100%	14	17	8	34	2	1	77
90%	13	16	7	31	2	1	69
80%	11	14	6	28	2	1	62
70%	10	12	6	24	1	1	54
60%	9	10	5	21	1	1	46
50%	7	9	4	17	1	1	38
I'm assuming that he plays in 82 games, by the way. Nevertheless, this seems fairly interesting. Detroit is going to be a much more competitive environment in terms of earning minutes on the ice.

Code:
	TOI	ES	PP	SH	%ES	%PP	%SH
HOLMS	12:27	9:39	2:47	0:00	77.51%	22.36%	0.00%
LARION	13:54	10:40	3:05	0:08	76.74%	22.18%	0.96%
FEDOROV	21:10	14:28	4:27	2:14	68.35%	21.02%  10.55
ROBITA	12:48	10:20	2:26	0:01	80.73%	19.01%	0.13%
HULL	18:07	13:18	3:22	1:26	73.41%	18.58%	7.91%
SHANAH	18:38	13:35	3:20	1:42	72.90%	17.89%	9.12%
DATSYUK	15:27	12:57	2:20	0:09	83.82%	15.10%	0.97%
ZETTER	16:18	12:48	2:02	1:27	78.53%	12.47%	8.90%
Some of those players are gone, but Yzerman didn't make that list, and Detroit has an interest in seeing guys llike Zetterberg and Datsyuk get the ice. Also note specifically the time on ice. The only guys who managed to get significantly above 18 minutes were the guys who were getting time on the PK. Based on what I've seen of Ray Whitney, he won't be seeing any time on the PK in Detroit. I'll be generous, and say that he sees maybe 20 seconds per game of PK time. As for PP time, this is going to be interesting too. He won't get the 6 minutes a game he was accustomed to in Columbus, although with Fedorov gone, there are significant PP minutes at centre to be had. Lets say he gets most of those minutes, and is looking at 3:00 a game on the PP. As for even strength, this is just gut feel, but I'm going to guess about 13 minutes a game. So he's looking at 18 minutes a night. Right off the bat, this drops him down to a 62 point season. Now lets consider the new breakdown.

Code:
       ESG	     ESA	 PPG	PPA	SHG	SHA	PTS
       14	     17	  4	17	0	0	53
Now for those of you who point out that Detroit has a much better PP, I would argue that this will be counteracted by the fact that Detroit had 90 fewer opportunities than did Columbus last year. They were 5% better on the PP for what its worth. As to the differences in the team, I guess time will tell, but I'd be surprised if Whitney scored much above 50 points this year, assuming an ice time mix similar to this one. Anything above, I'd attribute to the players that he plays with. Its going to be interesting. The other non-numerical factor would be the style that the teams play. I have the feeling Whitney could pretty much play however he wanted in Columbus-things won't be so loose in Detroit. Would you want to commit three years and 9.75 million to this?

Obviously, this sort of thing can be done with any player, but I just picked Whitney because he was a fairly high profile offseason signing. Hey GuyF, if you happen across this, you're tight with the Oilers brass, do teams crunch numbers like this on guys, or is this just all numerical mumbo jumbo? Is it more just a gut feel and what your scouts think when you make a move?

mudcrutch79 is offline  
Old
09-27-2003, 05:15 PM
  #2
PecaFan
Registered User
 
PecaFan's Avatar
 
Join Date: Nov 2002
Location: Ottawa (Go 'Nucks)
Posts: 8,904
vCash: 500
Quote:
Originally Posted by mudcrutch79
Now for those of you who point out that Detroit has a much better PP, I would argue that this will be counteracted by the fact that Detroit had 90 fewer opportunities than did Columbus last year. They were 5% better on the PP for what its worth. As to the differences in the team, I guess time will tell, but I'd be surprised if Whitney scored much above 50 points this year, assuming an ice time mix similar to this one. Anything above, I'd attribute to the players that he plays with. Its going to be interesting. The other non-numerical factor would be the style that the teams play. I have the feeling Whitney could pretty much play however he wanted in Columbus-things won't be so loose in Detroit. Would you want to commit three years and 9.75 million to this?

Obviously, this sort of thing can be done with any player, but I just picked Whitney because he was a fairly high profile offseason signing. Hey GuyF, if you happen across this, you're tight with the Oilers brass, do teams crunch numbers like this on guys, or is this just all numerical mumbo jumbo? Is it more just a gut feel and what your scouts think when you make a move?
Interesting analysis. I always appreciate a well thought out article.

I think 18 minutes a game will be right, but I think your point prediction is way too low. He's put up 60 and 70 point seasons before on 18 minutes a game, on teams that were no where close to as good as this Wings team is.

18 minutes a game on the highest scoring team in the league is a much better situation than 20 minutes on one of the lower scoring teams. For example, Detroit may have had 90 less power plays, but they still scored more PP goals.

He'll be 60 points for sure, as he usually is. 70+ if he finds chemistry with a Hull or Datsyuk.

PecaFan is offline  
Old
09-29-2003, 07:23 PM
  #3
oilswell
Registered User
 
Join Date: Mar 2002
Location: Wandering the globe
Posts: 1,760
vCash: 500
Quote:
Originally Posted by mudcrutch79
Ok, so I'm busy procrastinating, and I thought I'd kick around some of the PP ice time stats, to see if I could shake anything interesting loose.
Glad you frittered your time in this manner!

Quote:
The first relates to what percentage of a forwards ice time is PP time. I think that this can tell us a few things. First off, what forwards are either so good on the PP that the team reserves them for that situation,
Also, I suppose it says a lot about the team.

Quote:
Lets look at the top 20 for % of ice time coming on PP....
No real surprises there (except maybe Parrish for me...bit of an eye opener), which, I suppose, is good. Interesting thing about Havlat. I understand why he needs the PP time and that with him being on the 3rd line he would have a greater proportion on PP, but...could part of his shine be related to PP opportunities he's been given?

Quote:
Now for those of you who point out that Detroit has a much better PP, I would argue that this will be counteracted by the fact that Detroit had 90 fewer opportunities than did Columbus last year. They were 5% better on the PP for what its worth. As to the differences in the team, I guess time will tell, but I'd be surprised if Whitney scored much above 50 points this year, assuming an ice time mix similar to this one. Anything above, I'd attribute to the players that he plays with.
IMO you're missing the more important issue with Whitney. In Colombus, I daresay the opposition could more-or-less key in on him. In Detroit, he may be a little more anonymous and therefore find himself defended less heavily. His rates may well go up for this reason too.

oilswell is offline  
Old
09-30-2003, 06:04 AM
  #4
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by mudcrutch79
... do teams crunch numbers like this on guys, or is this just all numerical mumbo jumbo?
Well, mudcrutch, I had written a thorough response to this thread ... then HF crashed on me, and I'm too lazy to retype everything.

Anyhow, NHL teams seem to be ahead of the other major sports in terms of using data to evaluate players. Largely due to the successful work of Roger Nielsen in the 80s. Though obviously this forms just part of the overall picture. Here's a thread from OilFans that might interest you. It references the database that the NYR put together about 12 years ago.
http://oilfans.com/forum/index.php?t...rt=15&count=15

On the Whitney thing, and IMO ... in terms of outscoring the opposition 5 on 5, not so good. But a really effective powerplay guy, he'll do even better in this regard in Detroit. Especially with the departure of Larionov (one of the most productive powerplay players ever) to New Jersey (who desparately needed a player like Igor IMHO). And the loss of Federov as well, who was certainly an effective poewrplay guy too, and could be trusted to man the point.

I think Whitney will get a tonne of powerplay time in Detroit ... and with the talent around him there ... he'll do really well I think.

BTW:

Another way of looking at powerplay effectiveness. PP points per hour of PP ice-time (though NHL teams seem to prefer your 'per minute' unit ... I like bigger numbers, easier on the eyes
PLAYER _____ PP minutes PP points per hour
RONNING ____ 411 3.1
WHITNEY ____ 493 5.1
MOROZOV ____ 143 7.1
WEIGHT _____ 390 5.5
FRANCIS ____ 441 3.1
NASLUND ____ 437 7.4
BERTUZZI ___ 443 5.7
RICHARDS ___ 418 5.5
SANDERSON __ 399 3.8
JOHANSSON __ 244 4.2
O'NEILL ____ 403 3.9
PARRISH ____ 336 3.8
YASHIN _____ 379 4.4
LEMIEUX ____ 394 6.9
BRUNETTE ___ 300 4.8
KOVALCHUK __ 389 3.2
HAVLAT _____ 278 3.9
HEATLEY ____ 415 5.3
ANDREYCHUK _ 297 5.6
BONDRA _____ 353 3.2


igor* is offline  
Old
09-30-2003, 06:27 AM
  #5
oilswell
Registered User
 
Join Date: Mar 2002
Location: Wandering the globe
Posts: 1,760
vCash: 500
Quote:
Originally Posted by igor
Another way of looking at powerplay effectiveness.
Yes, I was thinking about this myself. One thought I had is that really really really good PP players never get much PP time.

They score in the first few seconds. (!)

Thus, the cynic (hi!) might say the only way to get a lot of PP time is to be pretty crummy at it yet be on a team that can't ice a better option. Maybe this explains a few players' numbers (Ronning?).

oilswell is offline  
Old
05-15-2004, 04:24 PM
  #6
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
This is going to come off as tooting my own horn a bit, but I thought I'd update this thread, as fruitlessly search for something that Igor said on some thread that I can't quite remember.

Anyway, based on the numbers above, if you multiply them out by Whitney's actual ice time in Detroit, you'd expect him to have 11.5 ESG (actual number was 10), 2.79 PPG (actual number was 3) and .80 SHG (actual number was 1). Assistwise, you'd expect 14.07 ESA (actual number was 19), 12 PPA (actual number was 9) and .4 SHA (actual number was 1). Overall, you'd expect 41.57 pts, and he actually had 43. I recognize that this is a rather small sample size (actual number of this sample size is 1), but it seems interesting nonetheless.

I just bought my copy of Baseball Prospectus, which I recommend to anyone with an interest in how performance analysis gets done in sports. Anyway, they discuss things that are skill as opposed to luck. They differentiate based on whether an event is dependent on the player in question or on what other players do. I'd speculate, and this is something that I'm looking into, that things like assists are more likely to be "luck", given the presence of the second assist, and that you don't have much control over the skill set of the guys playing with you.

In any event, interesting stuff, and something that I might try and push further. I think it provides a further suggestion that Slats could have pretty easily figured out that Anson Carter would go in the toilet upon becoming a NYR, as well as suggests that maybe statistical analysis still has a ways to go in hockey.

mudcrutch79 is offline  
Old
05-15-2004, 05:51 PM
  #7
Lowetide
Registered User
 
Lowetide's Avatar
 
Join Date: Feb 2002
Country: Canada
Posts: 13,281
vCash: 500
Mudcrutch:
They differentiate based on whether an event is dependent on the player in question or on what other players do. I'd speculate, and this is something that I'm looking into, that things like assists are more likely to be "luck", given the presence of the second assist, and that you don't have much control over the skill set of the guys playing with you.



One of the things I've always wondered about is why they must give a second A on a goal. If it hit a player's skate, then the play continued and it's clearly a fluke, don't award it. However, on a goal like the TBay goal the other night, that was poetry and everyone deserved a point on the play.


Mudcrutch:
In any event, interesting stuff, and something that I might try and push further. I think it provides a further suggestion that Slats could have pretty easily figured out that Anson Carter would go in the toilet upon becoming a NYR, as well as suggests that maybe statistical analysis still has a ways to go in hockey.


There are always going to be players in every era whose numbers don't really reflect their true skill. The question becomes what to measure and who is going to do the work. Categories like "quality chances" are already measured, and "giveaway/takeaway" was an actual stat not long ago.

What categories would you like to see measured?

Lowetide is offline  
Old
05-15-2004, 07:11 PM
  #8
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
Two stats that I'd love to see, although one of them is so subjective that I don't know that it would be of much worth. The first is a breakdown of icetime by who players play with and against. This would make it much easier to compare players properly. The second, more subjective one, would be some sort of measure on breakout passes made by defencemen, like a percentage, on how many they manage to put on the tape or something. In many ways, this is the most important move a dman makes offensively, sending the team down the ice, and guys who can't do it are a complete liability. It's an important skill, and one that can have big benefits without generating the assist necessarily. One more, I'd like to see penalties broken out into those that actually put a team down a man, and those that are coincidental-I've looked at this before, and in my opinion, PIM is an utterly useless stat. Oh, and one more subjective one, drawn penalties.

Basically, I'm looking for stats that can help us value and evaluate players more fairly compared to one another. God forbid that the NHL ever help us out with this though.

mudcrutch79 is offline  
Closed Thread

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 12:46 AM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2014 All Rights Reserved.