HFBoards

Go Back   HFBoards > NHL Western Conference > Pacific Division > Edmonton Oilers
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
Notices

EV+/- and the Powerplay

Closed Thread
 
Thread Tools
Old
07-19-2004, 06:25 PM
  #1
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
EV+/- and the Powerplay

I've been thinking about EV+/- all day here now and after reading something in the Nedved post, something kinda struck me. I posted it there, but I wonder if it doesn't deserve its own thread.

How many points must a player create on the PP to overcome a poor EV+/-????

If you make the arguement that a player with a poor EV+/- actually hurts the team 5on5, how many points does he have to score on the PP to make up for his 5on5 shortcomings?


T

This seems to me to be a lot like the question, 'how many doubles does poorly fielding baseball player have to put up in order to make up for his fielding deficiency?' (For you Moneyball fans, the question in the book was about Albert Belle)

Thoughts??

Big T is offline  
Old
07-19-2004, 06:35 PM
  #2
Lowetide
Registered User
 
Lowetide's Avatar
 
Join Date: Feb 2002
Country: Canada
Posts: 13,281
vCash: 500
Quote:
Originally Posted by Big T
I've been thinking about EV+/- all day here now and after reading something in the Nedved post, something kinda struck me. I posted it there, but I wonder if it doesn't deserve its own thread.

How many points must a player create on the PP to overcome a poor EV+/-????

If you make the arguement that a player with a poor EV+/- actually hurts the team 5on5, how many points does he have to score on the PP to make up for his 5on5 shortcomings?


T

This seems to me to be a lot like the question, 'how many doubles does poorly fielding baseball player have to put up in order to make up for his fielding deficiency?' (For you Moneyball fans, the question in the book was about Albert Belle)

Thoughts??

Great question, and I'm certainly not the guy who can supply the answer. I do think that this question holds the key to the Oilers anemic PP though, because MacT loves to roll 4 lines that can take care of the puck. I think that's why Adam Oates poor season was such a disappointment and why Ales Hemsky's future development is so vital.

Lowetide is offline  
Old
07-19-2004, 07:12 PM
  #3
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by Big T
I've been thinking about EV+/- all day here now and after reading something in the Nedved post, something kinda struck me. I posted it there, but I wonder if it doesn't deserve its own thread.

How many points must a player create on the PP to overcome a poor EV+/-????

If you make the arguement that a player with a poor EV+/- actually hurts the team 5on5, how many points does he have to score on the PP to make up for his 5on5 shortcomings?
...
It is a very good question. And it depends on how you use the player. In todays game, where set lines have virtually gone the way of the dinosaur ... you can't really shelter more than one line effectively.

So if you are going to have a 'liability line' ... (and MacTavish always has, out of necessity or philosophy is a question only he can answer ... I'd guess the former) They better darn well contibute on the powerplay ... because they are eating up the soft minutes where better all-around players could outscore. Be it the Comrie line or the Oates line or any line that Hemsky is on or the 'Carter at centre' debacle ... there always seems to be one.

I really have to wonder about the Oilers PP anyways. The personnel decisions seem Burnsesque. I'd rather they didn't aim for the "PP guy" addition, personally.

igor* is offline  
Old
07-19-2004, 09:13 PM
  #4
mackdogs*
 
Join Date: Oct 2003
Location: Van, left coast
Country: Canada
Posts: 907
vCash: 500
I think the best example of this as a player right now is Sergei Gonchar. He was a -14 last year but scored 58 points in 71 games. The worst player to fit this mold has to be a defenseman but I would love to have Gonchar here.

It's a very hard question to answer but I'd say 35 or so pp points for a forward of this type (say -15ish) in an 82 game sched and 30 ppp or so for a defenseman.

mackdogs* is offline  
Old
07-20-2004, 10:42 AM
  #5
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by mackdogs
I think the best example of this as a player right now is Sergei Gonchar. He was a -14 last year but scored 58 points in 71 games. The worst player to fit this mold has to be a defenseman but I would love to have Gonchar here.

It's a very hard question to answer but I'd say 35 or so pp points for a forward of this type (say -15ish) in an 82 game sched and 30 ppp or so for a defenseman.
Usually Gonchar is a stellar 5on5 guy too. What happened to him this year, anyways? I only saw one WSH game all year.

For that matter, what happened to that whole Caps team? Yeesh, just a nightmare season. They got outscored at 5on5 at a Columbus-type rate ... and they were a stifling defensive (albeit boring) team very recently. Having said that ... the Oilers always seemed to play great against them in Northlands, usually really good games.

igor* is offline  
Old
07-20-2004, 10:58 AM
  #6
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by Big T
...
How many points must a player create on the PP to overcome a poor EV+/-????...
I'll take another run at it, by way of example:

If Nedved plays the season with very good 5on5 players for the most part ... say guys like Dvorak. And if he doesn't have to go toe to toe with the better players on the other team too much either ... then he'll probably end up a bit better than evens ... say EV+/- of +5.

Put a healthy Reasoner in the exact same role, similar quality line-mates and opposition ... and he'll probably be between EV+/- +15 and +20.

So if they are being paid the same ... then Nedved would have to contribute to the powerplay to the tune of 10 to 15 EXTRA goals, over and above what Reasoner would have done in the same role.

Makes sense, no?

.
.
.

As an aside ... how the oddsmakers see it:

The betting line moved a staggering 0.2 goals, on the individual game line, when Reasoner was injured. The oddsmakers love the guy ... that equals about a +16 goal differential over the course of the season.

I missed most of the betting line changes, and it is impossible to automate the tracking, but: Down .15 goals when Dvorak was a late scratch before the VAN game in March or April. Up nearly half a goal when Forsberg was a late scratch vs the Oilers in COL in March.

The season line (92ish points from most sources for the Oil last year) didn't move a smidge when Comrie held out. But when Heatley's injury became public the ATL line dropped like a rock and was pulled completely within a couple of hours.

igor* is offline  
Old
07-20-2004, 12:27 PM
  #7
Hairball
Registered User
 
Hairball's Avatar
 
Join Date: Oct 2002
Location: Oiler Nation
Country: Canada
Posts: 2,533
vCash: 500
The way I see it is like this. The Oilers were a very good team 5 on 5 last year. The PP was very bad. We know that adding someone who is great on the PP will improve our PP. Adding someone who is not great defensively will not necessarily hurt our 5 on 5 play though because that is a team responsibility. So in the Oilers case, adding someone like this should have more benefit than detriment. So to answer the question, I don't think there has to be that much more contribution on the PP to make up for the even strength play. Only in our situation though.

Hairball is offline  
Old
07-20-2004, 12:30 PM
  #8
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
Quote:
Originally Posted by igor
I'll take another run at it, by way of example:

If Nedved plays the season with very good 5on5 players for the most part ... say guys like Dvorak. And if he doesn't have to go toe to toe with the better players on the other team too much either ... then he'll probably end up a bit better than evens ... say EV+/- of +5.

Put a healthy Reasoner in the exact same role, similar quality line-mates and opposition ... and he'll probably be between EV+/- +15 and +20.

So if they are being paid the same ... then Nedved would have to contribute to the powerplay to the tune of 10 to 15 EXTRA goals, over and above what Reasoner would have done in the same role.

Makes sense, no?


I wonder if this question can be solved independent of the possible replacement player (in this case, independant of Reasoner)????

Or is at simple as every PP+/- a player has above even, contributes equally to offset his poor EV+/-???? (and Vice versa ie: if Reasoner was great at even strength but for some reason had a negative PP+/-, this would offset point for point from his E/V+/- against his TOTAL+/- (new term????)

I'm not sure if that equal relationship between the two is correct. I'll have to think on that for a bit. Does it make a difference how many PP opportunities a player gets or does it only matter his total number of goals he contributes to???? (by simply being on the ice???)



Quote:
Originally Posted by igor
As an aside ... how the oddsmakers see it:

The betting line moved a staggering 0.2 goals, on the individual game line, when Reasoner was injured. The oddsmakers love the guy ... that equals about a +16 goal differential over the course of the season.

I missed most of the betting line changes, and it is impossible to automate the tracking, but: Down .15 goals when Dvorak was a late scratch before the VAN game in March or April. Up nearly half a goal when Forsberg was a late scratch vs the Oilers in COL in March.

The season line (92ish points from most sources for the Oil last year) didn't move a smidge when Comrie held out. But when Heatley's injury became public the ATL line dropped like a rock and was pulled completely within a couple of hours.
If the odds moving half a goal in the Oilers favour simply from the removal of one player (Forsberg) doesn't speak to how dominant that player is, I don't know what does.

Where do you guys get these odds from? Soounds interesting.


T

Big T is offline  
Old
07-20-2004, 12:36 PM
  #9
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
A few more questions

Is TOTAL+/- really a better (more accurate) picture of how a player contributes on the ice???

If we were even to compute such a stat, what would it include (EV+/-, PP+/-, PK+/-???) and what kind of relationship would the parts vary by?? ( (EV+/-) + 0.80(PP+/-) + 0.70(PK+/-) = TOTAL+/- ??? )



T

Big T is offline  
Old
07-20-2004, 04:42 PM
  #10
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by Big T
Is TOTAL+/- really a better (more accurate) picture of how a player contributes on the ice???

If we were even to compute such a stat, what would it include (EV+/-, PP+/-, PK+/-???) and what kind of relationship would the parts vary by?? ( (EV+/-) + 0.80(PP+/-) + 0.70(PK+/-) = TOTAL+/- ??? )
Personally, I really think that's heading down the wrong path.

Hockey is a simple game, its all about creating and preventing scoring chances, or in the big picture creating and preventing goals. And if you look at things simply and clearly ... the impact of a player in helping his team accomplish this is a large part of his true value. And, barring injury, it is remarkably consistent. Its the fans and analysts that are erratic IMHO.

But the PP, SH, 5on5, 4on4 and EN situations are completely different. And they have to be treated that way IMO.

In fact EV+/- is too fabricated. EV+ and EV- (or PP+ and SH-, whatever) is the cleanest and clearest.

The amount of ice-time ... and what another player (maybe the median player, whatever) would have accomplished with the same opportunity is the measuring stick.

igor* is offline  
Old
07-20-2004, 06:03 PM
  #11
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
Quote:
Originally Posted by igor
Personally, I really think that's heading down the wrong path.

Hockey is a simple game, its all about creating and preventing scoring chances, or in the big picture creating and preventing goals. And if you look at things simply and clearly ... the impact of a player in helping his team accomplish this is a large part of his true value. And, barring injury, it is remarkably consistent. Its the fans and analysts that are erratic IMHO.

But the PP, SH, 5on5, 4on4 and EN situations are completely different. And they have to be treated that way IMO.

In fact EV+/- is too fabricated. EV+ and EV- (or PP+ and SH-, whatever) is the cleanest and clearest.

The amount of ice-time ... and what another player (maybe the median player, whatever) would have accomplished with the same opportunity is the measuring stick.
1) OK igor, I'll take your word for it that EV+/- is too fabricated. Although, EV+/- does seem awfully simple. Could you explain why you feel this way????

2) I'm a dumba$$. What's the difference between EV+ / EV- and EV+/-????? How are they related???

3) I guess my point is if EV+/- is the best indicator we currently have of an individuals contribution to a team winning, can it be improved? Or put another way, can we really have an accurate picture of each player's contributions if we ignore parts of the game? (PowerPlay and PenaltyKill)

4) I get that the quality of opposition and the quality of the players you are with contribute to each individuals EV+/-, and that the amount of ice time would affect the outcome in different ways. Did you not post something some time ago comparing a given player including something to factor in the difficulty of his opponents and the quality of his on ice teamates??? I'm fuzzy on that.

I sure appreciate the help in explaining this stuff though. I'm just hoping to provide a different viewpoint/perspective on how to evaluate a player. Hope I'm not going over something already discussed.


T

Big T is offline  
Old
07-20-2004, 08:08 PM
  #12
Allan
Registered User
 
Join Date: Aug 2003
Location: Edmonton
Country: United Nations
Posts: 1,405
vCash: 500
Quote:
Originally Posted by Big T
1) OK igor, I'll take your word for it that EV+/- is too fabricated. Although, EV+/- does seem awfully simple. Could you explain why you feel this way????
I think the problem with EV+/- is that it's a summary, as opposed to showing the distributions of the goals (when, with who, against who, etc.)

Quote:
Originally Posted by Big T
2) I'm a dumba$$. What's the difference between EV+ / EV- and EV+/-????? How are they related???
As an example, a player can be EV+12 and EV-10, but as EV+/-, is is just expressed as the total of +2.

Quote:
Originally Posted by Big T
3) I guess my point is if EV+/- is the best indicator we currently have of an individuals contribution to a team winning, can it be improved? Or put another way, can we really have an accurate picture of each player's contributions if we ignore parts of the game? (PowerPlay and PenaltyKill)
I think the power play and penalty kill have to be represented seperately from even strength play, since the expectations, contributions, and probability of a goal being scored are so different. I don't think +/- apply so much in those situations - a stat like time on ice per goal or point is more appropriate, IMO.

Allan is offline  
Old
07-21-2004, 12:26 PM
  #13
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
Quote:
Originally Posted by Allan
I think the power play and penalty kill have to be represented seperately from even strength play, since the expectations, contributions, and probability of a goal being scored are so different. I don't think +/- apply so much in those situations - a stat like time on ice per goal or point is more appropriate, IMO.

And that gets back to my question. I agree that even strength and the powerplay and the PK have to be taken separately. What I wonder is, if we were to try and combine these numbers, (or some other numbers like you suggest) to get a more clear picture of every player's on ice contributions for the entire game, not just even strength. What kind of relationship do they have with each other???

TOTAL+/- = ( EV+/- ) + ( Ice Time per point produced / 'x' ) + ( 'y' / Ice time between goals allowed )

In this example, what are the values you would use for 'x' and 'y'????


It's a very tough question I know, but one that would make personel (sp?) descisions that much more clear, no???

ie:1) Does Nedved make up enough in his ability to positively effect the powerplay to overcome his very average play while even strength???

ie:2) How many goals does a very defensive forward have to prevent on the penalty kill to overcome his lack of production at even strength???

etc....

What is the relationship between a given players even strength play and his PP and PK play???

T

Big T is offline  
Old
07-21-2004, 12:51 PM
  #14
oilswell
Registered User
 
Join Date: Mar 2002
Location: Wandering the globe
Posts: 1,760
vCash: 500
Quote:
Originally Posted by Big T
What kind of relationship do they have with each other?
I don't know, but perhaps one angle attack might be to scale in two dimensions: (1) scale special teams GF-GA based on the difference in penalties (or penalty minutes assigned) taken by the team, and (2) scale the importance of special teams by the goal event rate difference. For a highly penalized team, good PKers are more important (and so on). If PP scoring rates are double what ES scoring rates are, good/bad PP play should be scaled accordingly.

oilswell is offline  
Old
07-22-2004, 01:54 PM
  #15
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
Quote:
Originally Posted by oilswell
I don't know, but perhaps one angle attack might be to scale in two dimensions: (1) scale special teams GF-GA based on the difference in penalties (or penalty minutes assigned) taken by the team, and (2) scale the importance of special teams by the goal event rate difference. For a highly penalized team, good PKers are more important (and so on). If PP scoring rates are double what ES scoring rates are, good/bad PP play should be scaled accordingly.
It's a really difficult question isn't it?? I agree that the ES and PP/PK are different beasts and so must be treated differently.

First of all, when you look to a player's EV+/-, a positive value indicates he has a positive contribution to his team 5on5, correct? But if we look at the PP, a player is expected to score. So having a PP+/- of zero actually isn't a very good thing at all. To take oilswells point, the results have to be scaled. That is, we need to discover a 'new zero'.

My thinking is, if the mythical 'average player' is expected to have a EV+/- of zero, what is the 'average player's' expected PP+/-. Or, how many net points is the 'average player' expected to produce in the course of the season. And of course the inverse would be true for the PK. PK+/- NET ZERO would be below zero since the opposing team is expected to score.

I would also assume we would have to have a different NET ZERO for 5on4 and 5on3 situations. (If the NET ZERO PP+/- for 5on4 is +5, then the NET ZERO PP+/- for 5on3 might be +11, or something to this effect.)

One other consideration, which I believe brings up a point made earlier, is how do you scale for teams that have more than the average amount of PP opportunities. I'm just pulling numbers out of my head here, but if the 'average player' has say 250 minutes of PP time per season, how do you scale for players with more (or less) PP time? Is it as simple as just pro-rating a player's PP+/- as if he had exactly 250 minutes of PP time??? Possibly.

A tool like this would be very benificial to every GM in the league for a couple of reasons.

1) For a team that is typically quite heavily penalized (undisciplined), a player who has a high PK+/- would be more valuable than for a team that is less penalized.

2) For a team that draws more than the average number of penalties, a player with higher PP+/- is more valuable than for the average team.

GM's could decide more accurately how valuable PP and PK 'specialists' are to their team.

Big T is offline  
Old
07-22-2004, 04:20 PM
  #16
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by Big T
...First of all, when you look to a player's EV+/-, a positive value indicates he has a positive contribution to his team 5on5, correct?
Generally ya. And clearly a guy who is EV+50 and EV-40 has a better chance of repeating his success than a guy who played fewer minutes and is EV+20 and EV-10. Just because with the second guy ---> a few goalposts or bounces here and there and his numbers change a lot from year to year.

Quality of opposition that you faced makes a huge difference to a players value too. As does the quality of teammates and goaltending.

Quote:
But if we look at the PP, a player is expected to score. So having a PP+/- of zero actually isn't a very good thing at all. To take oilswells point, the results have to be scaled. That is, we need to discover a 'new zero'.
With powerplays is the PP+/- rate matters, i.e. the number of PP goals that your team scored when you were on the ice, less the SH markers against, all per hour of PP time. Andas you say, obviously you have to separate out the 5on4 and 5on3 powerplays too.

Points-per-hour on the PP is a pretty good indicator too. Its usually the same group of star players that lead in these two stats every year. The quality of linemates makes a huge difference too ... since puck movement (IMO) is the most important ingredient in PP success, if you have one or two guys on your PP unit that are crap ... they'll probably drag your numbers down with them.

Because such a small % of the game is played with the man advantage ... you inevitably get some pretty streaky numbers too.

Quote:
My thinking is, if the mythical 'average player' is expected to have a EV+/- of zero, what is the 'average player's' expected PP+/-.
The average player is about EV+/- -3. Because talent isn't evenly distributed ... the median NHLer is a lot closer in talent level to the 4th-liner-fringe-guy than he is to Forsberg.

Quote:
Or, how many net points is the 'average player' expected to produce in the course of the season.
I don't have an answer, but I think that's a really good way of looking at it. I think if you grabbed all the guys, from each team, that ranked between 7th and 12th in PP ice-time on their teams ... and looked at their PP-points-per-hour and PP+/- per hour ... they'd probably be a pretty good benchmark for NHL PP mediocrity

And probably the second PP units of vast majority of teams get pretty similar results ... equally bad.

.
.
.

I think that guys who are good on the powerplay, but less effective at 5on5 ... typically they cost a lot more, salary-wise, than a guy that helps you outscore the same amount just with good 5on5 play.

Guys who are good at both ... the Oilers simply can't afford that kind of player. But a guy with good PP skills and who ISN'T and 5on5 force ... if they are used right I think that they can help your PP without breaking the budget. Guys like Hemsky and Bergeron on this team ... if they are sent out with other good PP talent as a unit ... say with Smyth, York and ??? (the Oilers really are a bit thin on the ground in PP-type talent ) ... then they will probably have decent success.

igor* is offline  
Old
07-23-2004, 11:44 AM
  #17
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
What I'm really wondering here guys and girls is...

Is it possible to come up with one number or indicator that would give a good indication of how much an individual player contributes to his team?

I now see how it's important to break EV+/- into EV+ and EV-, so I'm sure any new indicator created would not be perfect. But is there some way to combine EV+/-, (or some other statistic that shows a player's effectiveness 5on5) with a PP effectivness indicator and a PK indicator??

Maybe these things will turn out to be better left on their own, but who knows??



T

Big T is offline  
Old
07-23-2004, 12:16 PM
  #18
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by Big T
What I'm really wondering here guys and girls is...

Is it possible to come up with one number or indicator that would give a good indication of how much an individual player contributes to his team?

I now see how it's important to break EV+/- into EV+ and EV-, so I'm sure any new indicator created would not be perfect. But is there some way to combine EV+/-, (or some other statistic that shows a player's effectiveness 5on5) with a PP effectivness indicator and a PK indicator??

Maybe these things will turn out to be better left on their own, but who knows??
I seem to be the only person responding to you here, and I don't think you agree with me, but WTH, one more post:

I think you're looking to combine good raw data into a universal 'magic number' for a players value, no?

And I don't think that makes sense.

The EV numbers are really solid and repeat well assuming good health for the player. Partly because almost all coaches shuffle their lines like crazy nowadays, and partly because most of the ice-time and goals occur at 5on5.

SH numbers have a lot more to do with goaltending and coaching than players. So while some guys are better PKers than others ... NONE are dominant PKers IMO (besides great goalies).

PP numbers are a bit tougher because of the limited ice-time and the really significant impact of team-mates. Still, you can make a good estimate of a players future impact on a PP by his past numbers ... previous linemates considered.

BUT ... it depends on the team. By way of example ... NSH have had terrible PPs ... so adding a guy like Zidlicky to the lineup made sense, because he has good PP skills, and their team really lacked those kind of guys. Zidlicky was/is a 5on5 misadventure ... but the good (PP) outweighs the bad (5on5) so they ran with him.

But on a tema like DET or OTT ... they don't need him on the PP, so he'd have been playing in the AHL.

A guy like Delmore or Tarnstrom might make sense to some teams because they can contribute to the PP. But most teams would prefer to use a forward on the PP point and have a more reliable d-man on the roster than go this route. All depends on your team, your coach and your bias I guess.

So ... if you consider a team of median players to be good for an EV goal differential of -12 or so and a special teams differential of about -10 to -15. Then that's a -25 goal differential and a $22M payroll. And you need to get to +40 goal differential to be a legitimate contender.

Thats about a 65 goal swing! A lot.

If you have a budget of $34M ... that extra $12M had BETTER get you good outscoring results, or your hooped. About a +1 in goal differential for every extra $200k in salary you are shelling out above the median salary ($1.1M roughly).

Generally speaking, 5on5 goal differential is cheaper than PP outscoring. Which is the reason that the Oilers were able to be such a good 5on5 outscoring team last year ... in spite of the relatively low budget, enormous centre issue and bouts of dodgy goaltending.

IMHO the cheapest way to build a successful team is pretty much the way that the Oilers are doing it. With a couple things I don't understand ...

1. They need less young D, more veteran D.
2. They need PP specialists ... and they need to get the 5 best PP guys out on the ice at the same time! Hemsky, Bergeron, York and Smyth ... these guys should ALL be out there at the same time on the first unit. And of the next guys in line ... at least the next four are forwards: (Dvorak, Reasoner, Torres, Isbister, Horcoff)

.
.

BTW: Sorry, I managed to turn this into an Oilers PP rant. I'm bad for that. :lol

BTW#2: The obvious simple way to build a cheap team that gets good results ... get a stellar goalie with a low salary. But since that's like planning make big money by buying a lottery ticket ... and its short-term ... its not realistic longterm.


Last edited by igor*: 07-23-2004 at 12:20 PM.
igor* is offline  
Old
07-23-2004, 01:36 PM
  #19
Big T
Registered User
 
Join Date: Sep 2003
Location: LLoydminster, AB
Posts: 338
vCash: 500
Quote:
Originally Posted by igor
So ... if you consider a team of median players to be good for an EV goal differential of -12 or so and a special teams differential of about -10 to -15. Then that's a -25 goal differential and a $22M payroll. And you need to get to +40 goal differential to be a legitimate contender.

Thats about a 65 goal swing! A lot.

If you have a budget of $34M ... that extra $12M had BETTER get you good outscoring results, or your hooped. About a +1 in goal differential for every extra $200k in salary you are shelling out above the median salary ($1.1M roughly).

Generally speaking, 5on5 goal differential is cheaper than PP outscoring. Which is the reason that the Oilers were able to be such a good 5on5 outscoring team last year ... in spite of the relatively low budget, enormous centre issue and bouts of dodgy goaltending.

IMHO the cheapest way to build a successful team is pretty much the way that the Oilers are doing it. With a couple things I don't understand ...
Now that's the kind of thinking I was really trying to get at. Obviously in a pretty round-about way though!!! I really like looking at things the way you have described igor. Awesome.

I still think it would be worthwhile to look in to finding a better indicator for PP and PK effectiveness though. But maybe not combining them.


T

Big T is offline  
Old
07-26-2004, 07:51 PM
  #20
mudcrutch79
Registered User
 
Join Date: Jul 2003
Location: The Big Smoke
Posts: 3,903
vCash: 500
Igor,

Why not adjust a player's EV- to reflect the goaltending he's playing in front of. You've said in the past that you don't think that the team has much of an impact on the goaltender. I'm not sure I necessarily agree, especially on the margins, but it seems to me that an EV-20 on the 1986 Oilers is different than an EV-20 on the early 90's Nordiques. One would think that by using save percentage of the various goalies, you could make teamwide adjustments...

It seems to me that this would then become more useful as an analytical tool..

mudcrutch79 is offline  
Old
07-26-2004, 08:32 PM
  #21
igor*
 
Join Date: Mar 2002
Posts: 3,276
vCash: 500
Quote:
Originally Posted by mudcrutch79
Igor,

Why not adjust a player's EV- to reflect the goaltending he's playing in front of. You've said in the past that you don't think that the team has much of an impact on the goaltender. I'm not sure I necessarily agree, especially on the margins, but it seems to me that an EV-20 on the 1986 Oilers is different than an EV-20 on the early 90's Nordiques. One would think that by using save percentage of the various goalies, you could make teamwide adjustments...

It seems to me that this would then become more useful as an analytical tool..
Absolutely. That's the logical first step IMO, and the reason that understanding EVsave% is important in the first place.

Than accounting for quality of opposition, then quality of teammates.

igor* 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 04:14 PM.

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.