HFBoards

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

Some of our team stats look really bad...

Reply
 
Thread Tools
Old
02-16-2013, 04:31 PM
  #126
Fourier
Registered User
 
Fourier's Avatar
 
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 8,655
vCash: 500
Quote:
Originally Posted by Stoneman89 View Post
Look, no one on here saying the #1 line doesn't get fed more Schultz than does the #2 line. The point was, and was settled by stats, is that it is not enough of a margin to make a huge difference. What's next, first in line to get skates sharpened?
I can add to what Bob said in the other thread:

Here is the % time on ice numbers for each of top 6 with the 6 main dmen.

Code:
	J. Schultz	N. Schultz	Smid	Petry	Whitney	Potter
Hall	46.9	                37.9	             34.3	  33.5	   20	16.1
Eberle	48.7	                39	             33.7	  32.8	   19	14.5
RNH	49.1	                37.9	             33.3	  33.2	   17.7	13.7
Gagner	34.6	                31.5	             34	  37.9	   24.3	20.6
Hemsky	33.6	                31.5	             33.4	  36.7	   26.9	22.3
Yakupov	34	                31.5	             35.1	  37.5	   24.2	20.1
These distributions are actually quite similar with the only real exception being that the top line has a fair number of additional minutes with J. Schultz.

Here are the raw minutes:

Code:
	J. Schultz	N. Schultz	Smid	Petry	Whitney	Potter
Hall	90.4	                 73.0	             66.1	64.5	 38.5	 31.0
Eberle	92.3	                 74.0	             63.9	62.2	 36.0	 27.5 
RNH	87.6	                 67.6	             59.4	59.3	 31.6	 24.5
Gagner	57.1	                 51.9	             56.1	62.5	 40.1	 34.0
Hemsky	51.6	                 48.4	             51.3	56.4	 41.3	 34.3
Yakupov	48.5	                 44.9	             50.1	53.5	 34.5	 28.7
Here are the raw GA 5 vs 5 and the rates normalized to 60 minutes.

Code:
	      GAOn	GAOn/60
Hall	         5	              1.58
Eberle	         5	              1.56
RNH	         6	              2.02
Gagner	         9	              3.27
Hemsky	         9	              3.51
Yakupov	         8	              3.37
J.Schultz	         6	              1.72
N. Schultz        7	              2.26
Petry	         8	              2.4
Smid	       10	              3.06
Whitney	         6	              2.72
Potter	         5	              2.62
These stats are pretty clear. The second line is giving up goals at nearly twice the rate of teh first line when you factor in minutes played and abiut 80% more goals in the raw. Schultz has an effect on this but only slightly since even if if you were to replace the extra minutes played with Schultz with minutes spent with Whitney you would expect actual GA for the top line to rise by less than 1/2 a goal.

As far as the extra time the 2nd line has played with Whitney and Potter about 3 minutes more this year than the top line. This would be expected to add about 2.75*3/60= .1375 actual goals against to the 2nd lines GA total.

So what these stats show is that right now there is no evidence to hang the GA differential on differences in pairings. The second line has simply given up more goals in fewer minutes no matter who they were playing with.

[All stats are from behind the net.]

Fourier is offline   Reply With Quote
Old
02-16-2013, 07:48 PM
  #127
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by Fourier View Post
I can add to what Bob said in the other thread:


These stats are pretty clear. The second line is giving up goals at nearly twice the rate of teh first line when you factor in minutes played and abiut 80% more goals in the raw. Schultz has an effect on this but only slightly since even if if you were to replace the extra minutes played with Schultz with minutes spent with Whitney you would expect actual GA for the top line to rise by less than 1/2 a goal.

As far as the extra time the 2nd line has played with Whitney and Potter about 3 minutes more this year than the top line. This would be expected to add about 2.75*3/60= .1375 actual goals against to the 2nd lines GA total.

So what these stats show is that right now there is no evidence to hang the GA differential on differences in pairings. The second line has simply given up more goals in fewer minutes no matter who they were playing with.

[All stats are from behind the net.]
I see it a little differently. This is how I calculated it.

Gagner currently has 164mins played. 34.6% with J Schultz. Convert into minutes=57mins played with J Schultz.

If Gagner instead had RNH allotment of 49.1% with J Schultz hes getting 80.5Mins of Justin.

So an additional 24mins(rounded) of J Shultz.

Now lets look at GA for J Schultz. Each 60mins his GAON =1.72 GAOFF=2.75. So theres more than 1GA allowed for each 60mins J Shultz isn't on the ice.

Multiple 24mins/60mins above by 1.03 =.412 less GA/60mins

Now one would do the same calculation for RNH

178minsplayed X49.1% with J Schultz = 87.3mins

Use Gagners coefficient and its 61.5mins. So RNH would be getting 26.2mins less J Schultz with Gagners allotment.

26.2/60mins X 1.03(J Schultz ON/off differential as cited earlier)= .449 more GA/60mins.


Now take these numbers and apply them to the present GA/60 mins.

So, If RNH, and Gagners allotment of J Schultz was REVERSED what we get is:

Gagner: 3.27 - .412 = 2.85GA/60mins

RNH 2.02 = ..449 - 2.47GA/60mins


So with JUST the J Schultz factor applied we see that theres not much difference above in the figures. Now if one cares to factor in the Consequence of playing with Hemsky and Yakupov vs playing with Eberle and Hall then Gagner would come out ahead. That calculation would take a while...

Just to note as well that Gagner is -1 this year. Hemsky is -4 and Yak -6. Even though they play the vast majority of time together its clear thus far that GAgner has been more successful then his linemates when they are not together.

Replacement is offline   Reply With Quote
Old
02-17-2013, 02:22 PM
  #128
bellagiobob
Registered User
 
bellagiobob's Avatar
 
Join Date: Jul 2006
Location: Sin City
Country: Barbados
Posts: 758
vCash: 500
Quote:
Originally Posted by Replacement View Post
I see it a little differently. This is how I calculated it.

Gagner currently has 164mins played. 34.6% with J Schultz. Convert into minutes=57mins played with J Schultz.

If Gagner instead had RNH allotment of 49.1% with J Schultz hes getting 80.5Mins of Justin.

So an additional 24mins(rounded) of J Shultz.

Now lets look at GA for J Schultz. Each 60mins his GAON =1.72 GAOFF=2.75. So theres more than 1GA allowed for each 60mins J Shultz isn't on the ice.

Multiple 24mins/60mins above by 1.03 =.412 less GA/60mins

Now one would do the same calculation for RNH

178minsplayed X49.1% with J Schultz = 87.3mins

Use Gagners coefficient and its 61.5mins. So RNH would be getting 26.2mins less J Schultz with Gagners allotment.

26.2/60mins X 1.03(J Schultz ON/off differential as cited earlier)= .449 more GA/60mins.


Now take these numbers and apply them to the present GA/60 mins.

So, If RNH, and Gagners allotment of J Schultz was REVERSED what we get is:

Gagner: 3.27 - .412 = 2.85GA/60mins

RNH 2.02 = ..449 - 2.47GA/60mins


So with JUST the J Schultz factor applied we see that theres not much difference above in the figures. Now if one cares to factor in the Consequence of playing with Hemsky and Yakupov vs playing with Eberle and Hall then Gagner would come out ahead. That calculation would take a while...

Just to note as well that Gagner is -1 this year. Hemsky is -4 and Yak -6. Even though they play the vast majority of time together its clear thus far that GAgner has been more successful then his linemates when they are not together.
Manipulating averages like this is not valid mathematically. Here is why I say this:

Lets say that for the rest of the year Gagner only plays with J. Schultz. Then he will play roughly 450 minutes more minutes with Schultz than he has now. Using your method above and your 2.85GA/60 number here is what we would get in this scenario:

2.85- 450/60 x 1.03 = -4.78

So if your method above was actually correct then exactly the same reasoning would show that if Gagner played the whole season with only Schultz going forward then we could expect Gagner to have a -4.78 GA/60 at the end of the year. This would mean that not only would Gagner not be on for any goals against the whole year but in fact eventually every time he stepped onto the ice goals that had already been scored against the Oilers would be removed from record.


Here is what the calculation you are trying to do really should look like:

Playing an extra 24 minutes with Schultz would add on average

24/60 x 1.72 =.688 goal

assuming that you maintained Schultz numbers.

This would increase his GA from 9 to 9.688.

However he would now have 188 min played so his new GAON/60 would be

9.688/188 x 60 =3.09

For RNH losing 26.6 minutes with Schultz would reduce his GA by the same .763 but now he would have
151.8 min played. His GAON/60 would be:

5.237/158.8 x 60 = 2.07

So in reality even if you gave Gagner the "Schultz advantage" that RNH currently enjoys the stats would look pretty much the same.


Last edited by bellagiobob: 02-17-2013 at 02:50 PM.
bellagiobob is offline   Reply With Quote
Old
02-17-2013, 02:56 PM
  #129
The Bored Man
#94
 
The Bored Man's Avatar
 
Join Date: Jul 2009
Location: Edmonton
Country: Canada
Posts: 5,379
vCash: 537
Quote:
Originally Posted by Replacement View Post
I see it a little differently. This is how I calculated it.

Gagner currently has 164mins played. 34.6% with J Schultz. Convert into minutes=57mins played with J Schultz.

If Gagner instead had RNH allotment of 49.1% with J Schultz hes getting 80.5Mins of Justin.

So an additional 24mins(rounded) of J Shultz.

Now lets look at GA for J Schultz. Each 60mins his GAON =1.72 GAOFF=2.75. So theres more than 1GA allowed for each 60mins J Shultz isn't on the ice.

Multiple 24mins/60mins above by 1.03 =.412 less GA/60mins

Now one would do the same calculation for RNH

178minsplayed X49.1% with J Schultz = 87.3mins

Use Gagners coefficient and its 61.5mins. So RNH would be getting 26.2mins less J Schultz with Gagners allotment.

26.2/60mins X 1.03(J Schultz ON/off differential as cited earlier)= .449 more GA/60mins.


Now take these numbers and apply them to the present GA/60 mins.

So, If RNH, and Gagners allotment of J Schultz was REVERSED what we get is:

Gagner: 3.27 - .412 = 2.85GA/60mins

RNH 2.02 = ..449 - 2.47GA/60mins


So with JUST the J Schultz factor applied we see that theres not much difference above in the figures. Now if one cares to factor in the Consequence of playing with Hemsky and Yakupov vs playing with Eberle and Hall then Gagner would come out ahead. That calculation would take a while...

Just to note as well that Gagner is -1 this year. Hemsky is -4 and Yak -6. Even though they play the vast majority of time together its clear thus far that GAgner has been more successful then his linemates when they are not together.
What...? That's not how those stats work at all, as the post above me demonstrates. Yikes. This is shades of the "Horcoff > Crosby" madness from a few years ago.

The Bored Man is offline   Reply With Quote
Old
02-17-2013, 07:44 PM
  #130
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by bellagiobob View Post
Manipulating averages like this is not valid mathematically. Here is why I say this:

Lets say that for the rest of the year Gagner only plays with J. Schultz. Then he will play roughly 450 minutes more minutes with Schultz than he has now. Using your method above and your 2.85GA/60 number here is what we would get in this scenario:

2.85- 450/60 x 1.03 = -4.78

So if your method above was actually correct then exactly the same reasoning would show that if Gagner played the whole season with only Schultz going forward then we could expect Gagner to have a -4.78 GA/60 at the end of the year. This would mean that not only would Gagner not be on for any goals against the whole year but in fact eventually every time he stepped onto the ice goals that had already been scored against the Oilers would be removed from record.


Here is what the calculation you are trying to do really should look like:

Playing an extra 24 minutes with Schultz would add on average

24/60 x 1.72 =.688 goal

assuming that you maintained Schultz numbers.

This would increase his GA from 9 to 9.688.

However he would now have 188 min played so his new GAON/60 would be

9.688/188 x 60 =3.09

For RNH losing 26.6 minutes with Schultz would reduce his GA by the same .763 but now he would have
151.8 min played. His GAON/60 would be:

5.237/158.8 x 60 = 2.07

So in reality even if you gave Gagner the "Schultz advantage" that RNH currently enjoys the stats would look pretty much the same.
All I was going for with the calulations above is if the J Schultz allotment was REVERSED and both RNH and Gagner still played the minutes they currently do only with the percentage the other center gets of J Schultz. i.e. Gagner getting 49% and RNH 34% paired with J Schultz. Basically an inverse of what it is now. The numbers show there is an effect. Theres also an effect playing with Yak who has the worst +/- and GA/60mins on the club as a forward, and Hemsky who can be occasionally daft.

Replacement is offline   Reply With Quote
Old
02-17-2013, 07:52 PM
  #131
McClelland
Registered User
 
McClelland's Avatar
 
Join Date: Aug 2011
Location: Bergen
Country: Norway
Posts: 1,063
vCash: 50
Quote:
Originally Posted by Replacement View Post
All I was going for with the calulations above is if the J Schultz allotment was REVERSED and both RNH and Gagner still played the minutes they currently do only with the percentage the other center gets of J Schultz. i.e. Gagner getting 49% and RNH 34% paired with J Schultz. Basically an inverse of what it is now. The numbers show there is an effect. Theres also an effect playing with Yak who has the worst +/- and GA/60mins on the club as a forward, and Hemsky who can be occasionally daft.
Nice to have one on the board who can give us correct stats

Who the hell is Gagner going to play with in this team? Seems everyone destroys his good numbers

McClelland is offline   Reply With Quote
Old
02-17-2013, 07:58 PM
  #132
SeriousBusiness
T.Hall da man
 
SeriousBusiness's Avatar
 
Join Date: Oct 2003
Country: Canada
Posts: 3,582
vCash: 500
Quote:
Originally Posted by Beerfish View Post
Tambo will be in no danger at all no matter where the oilers finish.
Nah, he'll be gone.

SeriousBusiness is offline   Reply With Quote
Old
02-18-2013, 06:07 AM
  #133
Fourier
Registered User
 
Fourier's Avatar
 
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 8,655
vCash: 500
Quote:
Originally Posted by McClelland View Post
Nice to have one on the board who can give us correct stats

Who the hell is Gagner going to play with in this team? Seems everyone destroys his good numbers
As bob's numbers show there is no evidence that the distribution of the defensive pairings has more than a negligible impact on the statistical evidence supporting the assertion that the first line has been far better defensively than the second.

Moreover, if you add to this that the first line has been consistently seeing the best players from the other team it is even more noteworthy.

Prior to the last game against Colorado here were the first lines most common opponents:

From behind the net:

MARC-EDOUARDVLASIC 10.9%, BRAD STUART 10.4%, Joe Thorton 9.4%, Patrick Marleau 9.2%, P.J. Parenteau 8.8%

Matt Duchene 8.8%, Alex Edler 8.2%, Henrik Sedin 7.6%, Daniel Sedin 7.3%

(The %ages are for RNH but they are very similar for all three).


For the second line right now the most common opponents are:

PAULSTASTNY 12.5%, DAVID JONES 8.2%, Dan Boyle 7.9%, Greg Zanon 7.8%,

Matt Irwin 7.7%, Kevin Beiksa 7.6%, Ryan O'Byrne 7.5%, Matt Huniwick 7.5%,

John Mitchell 7.1% , JAMIE MCGINN 7.0%

Fourier is offline   Reply With Quote
Old
02-18-2013, 08:43 AM
  #134
I am the Liquor
Registered User
 
I am the Liquor's Avatar
 
Join Date: Oct 2008
Location: Sunnyvale
Country: Canada
Posts: 34,178
vCash: 5496
Its all Hemsky's, Yakupov's, Omark's, Penner's, Paajarvi's, Belanger's, Cogliano's, Nilsson's fault.

I am the Liquor is offline   Reply With Quote
Old
02-18-2013, 10:12 AM
  #135
McClelland
Registered User
 
McClelland's Avatar
 
Join Date: Aug 2011
Location: Bergen
Country: Norway
Posts: 1,063
vCash: 50
Quote:
Originally Posted by Fourier View Post
As bob's numbers show there is no evidence that the distribution of the defensive pairings has more than a negligible impact on the statistical evidence supporting the assertion that the first line has been far better defensively than the second.

Moreover, if you add to this that the first line has been consistently seeing the best players from the other team it is even more noteworthy.

Prior to the last game against Colorado here were the first lines most common opponents:

From behind the net:

MARC-EDOUARDVLASIC 10.9%, BRAD STUART 10.4%, Joe Thorton 9.4%, Patrick Marleau 9.2%, P.J. Parenteau 8.8%

Matt Duchene 8.8%, Alex Edler 8.2%, Henrik Sedin 7.6%, Daniel Sedin 7.3%

(The %ages are for RNH but they are very similar for all three).


For the second line right now the most common opponents are:

PAULSTASTNY 12.5%, DAVID JONES 8.2%, Dan Boyle 7.9%, Greg Zanon 7.8%,

Matt Irwin 7.7%, Kevin Beiksa 7.6%, Ryan O'Byrne 7.5%, Matt Huniwick 7.5%,

John Mitchell 7.1% , JAMIE MCGINN 7.0%
I should marked that post with a I meant that Bob is the real deal here and his calculating hasnt any bias.

I mean Replace has throwed out miles of home made stats true out the years to support his "players" and next to no one has that time over to get into that debate with him, even if you have a feeling that something is wrong.

McClelland is offline   Reply With Quote
Old
02-18-2013, 10:42 AM
  #136
Red Deer Rebel
Registered User
 
Red Deer Rebel's Avatar
 
Join Date: Apr 2008
Location: Red Deer
Posts: 2,315
vCash: 500
Maybe it's Gagner who's destroying Whitney's stats?

Red Deer Rebel is offline   Reply With Quote
Old
02-18-2013, 11:11 AM
  #137
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by McClelland View Post
I should marked that post with a I meant that Bob is the real deal here and his calculating hasnt any bias.

I mean Replace has throwed out miles of home made stats true out the years to support his "players" and next to no one has that time over to get into that debate with him, even if you have a feeling that something is wrong.
The fact of the matter is the Oilers are sheltering the young stars with the best possible D pairing available. Theres a big drop off after that in the D pairings. This isn't an area of conjecture. Nobody dissagrees that its the Schultz's, then Smid-Petry, then Whitney-Potter with the last pair being a severe dropoff.

Next, Nuge gets to play with two heart and soul guys in Hall and Eberle. These are already bonafide experienced NHL Players that have excelled at the game.

Gagner's rarely if ever recieved this type of sheltering and has instead been assigned years of less than stellar players like Cogs, Nillson, Omark, Paajarvi who were either not ready, too young, or weren't going to be consistent NHL players. Nuge has worked with star players from the word go.

Soon as GAgner gets his allotment of star players to work with its predictable how far he can take it. The most productive line on the club and against the odds.

As an aside for Fourier to be quoting Qualcomp specific player matchings right now this early in the season as if it means much of anything is spurious logic. We play 14 different teams. We've only played 14 games. Mentioning opponent names at this point as if its indicative of how all things go in all games is silly. Both the topsix lines have had matching. Any sentient team would do that, its where all our goals are.

Replacement is offline   Reply With Quote
Old
02-18-2013, 11:15 AM
  #138
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by Fourier View Post
As bob's numbers show there is no evidence that the distribution of the defensive pairings has more than a negligible impact on the statistical evidence supporting the assertion that the first line has been far better defensively than the second.

Moreover, if you add to this that the first line has been consistently seeing the best players from the other team it is even more noteworthy.

Prior to the last game against Colorado here were the first lines most common opponents:

From behind the net:

MARC-EDOUARDVLASIC 10.9%, BRAD STUART 10.4%, Joe Thorton 9.4%, Patrick Marleau 9.2%, P.J. Parenteau 8.8%

Matt Duchene 8.8%, Alex Edler 8.2%, Henrik Sedin 7.6%, Daniel Sedin 7.3%

(The %ages are for RNH but they are very similar for all three).


For the second line right now the most common opponents are:

PAULSTASTNY 12.5%, DAVID JONES 8.2%, Dan Boyle 7.9%, Greg Zanon 7.8%,

Matt Irwin 7.7%, Kevin Beiksa 7.6%, Ryan O'Byrne 7.5%, Matt Huniwick 7.5%,

John Mitchell 7.1% , JAMIE MCGINN 7.0%
If I can establish the framework of your logic here J Schultz, with 49% of his toi with Nuge and the topline has a "negligible effect" while you're quoting off names like Joe Thornton at 9.4% on ice opponent as having a salient effect..

I don't even know what to say here.

Its possible to make much better arguments than the above.


Early in the season especially Qualteam is a more statistically significant, salient variable than Qualcomp due to sample size. To even be suggesting what Qualcomp really looks like this early is misleading. The sample is too limited to make much out of any very limited differentiation. In terms of % we're talking minute differences in toi, maybe secs, making up a % difference.

Replacement is offline   Reply With Quote
Old
02-18-2013, 12:15 PM
  #139
Beerfish
Registered User
 
Beerfish's Avatar
 
Join Date: Apr 2007
Posts: 10,655
vCash: 500
Quote:
Originally Posted by SeriousBusiness View Post
Nah, he'll be gone.
If the Oilers finish dead last he will still be here. No question in my mind. There is always a huge barrell of excuse in oiler land.

Beerfish is offline   Reply With Quote
Old
02-18-2013, 05:25 PM
  #140
Fourier
Registered User
 
Fourier's Avatar
 
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 8,655
vCash: 500
Quote:
Originally Posted by Replacement View Post
The fact of the matter is the Oilers are sheltering the young stars with the best possible D pairing available. Theres a big drop off after that in the D pairings. This isn't an area of conjecture. Nobody dissagrees that its the Schultz's, then Smid-Petry, then Whitney-Potter with the last pair being a severe dropoff.

Next, Nuge gets to play with two heart and soul guys in Hall and Eberle. These are already bonafide experienced NHL Players that have excelled at the game.

Gagner's rarely if ever recieved this type of sheltering and has instead been assigned years of less than stellar players like Cogs, Nillson, Omark, Paajarvi who were either not ready, too young, or weren't going to be consistent NHL players. Nuge has worked with star players from the word go.

Soon as GAgner gets his allotment of star players to work with its predictable how far he can take it. The most productive line on the club and against the odds.

As an aside for Fourier to be quoting Qualcomp specific player matchings right now this early in the season as if it means much of anything is spurious logic. We play 14 different teams. We've only played 14 games. Mentioning opponent names at this point as if its indicative of how all things go in all games is silly. Both the topsix lines have had matching. Any sentient team would do that, its where all our goals are.



Quote:
Originally Posted by Replacement View Post
If I can establish the framework of your logic here J Schultz, with 49% of his toi with Nuge and the topline has a "negligible effect" while you're quoting off names like Joe Thornton at 9.4% on ice opponent as having a salient effect..

I don't even know what to say here.

Its possible to make much better arguments than the above.


Early in the season especially Qualteam is a more statistically significant, salient variable than Qualcomp due to sample size. To even be suggesting what Qualcomp really looks like this early is misleading. The sample is too limited to make much out of any very limited differentiation. In terms of % we're talking minute differences in toi, maybe secs, making up a % difference.
Not surprisingly, your posts show that you really don't understand what these statistics mean. They have nothing to do with Qualcomp or any +/- related statistic. They are a direct measure of how much time each player has played against various opposition players. For example:

RNH has played about 34:05 ES minutes in the two games against SJ. The stats show that his line played roughly 19:25 minutes against the Sharks top pairing of Vlasic and Stuart.

In contrast Gagner played 25:09 mins ES vs SJ. His line faced Matt Hunwick, SJ's #6 defenceman, a little over 14 minutes at ES.

So while the top line was playing well over half the game against the Sharks top pairing, Gagner and the second line were playing a significant majority of the time against the Sharks weakest defenseman.

Similarly, RHN's line faced the Sharks top forward line about 17 minutes of the 34 minutes they played at ES. Similarly, the top line faced the Sedin line about 1/2 the time against Vancouver. Against Colorado they played the majority of the game against Duchene's line. Against Phoenix the top line saw OEL for almost 80% of the time they were on the ice. IF you want to look at the TOI stats of every gae the Oilers have played the pattern will be the same.

No one is using Qualcomp here. It is undeniable that these players are the best the opposition has to offer.

The top line spends the majority of time playing against the opposition’s top forwards and best defensemen. Gagner's line gets a mixture of whoever else was left over.


As far as sheltering the top line with better defense, I thought bob had already shown this "fact" to be wrong. They have played more minutes with Justin Schultz but if you look at absolute time on the ice with all of the Oilers other defensemen there is very little difference between the two lines. Furthermore, it has already been shown that this extra time with Schultz has almost no real impact on their stats as afar as thing like GAON/60 are concerned.

In contrast it seems clear that if the first line is playing significantly more minutes against the opposition best offensive players. This you dismiss as meaningless. If there is any spuriuos logic here it would seem to be this.

If you want to convince yourself that you have made your case go ahead. It would seem that no one else is buying it.

Fourier is offline   Reply With Quote
Old
02-18-2013, 05:30 PM
  #141
nexttothemoon
Eight Straight
 
nexttothemoon's Avatar
 
Join Date: Jan 2010
Location: Northern Alberta
Country: Canada
Posts: 8,726
vCash: 50
Quote:
Originally Posted by Beerfish View Post
If the Oilers finish dead last he will still be here. No question in my mind. There is always a huge barrell of excuse in oiler land.
If they finish dead last they need Tambo at least until the lottery draw. Need his ridiculous troll luck to deliver the #1 pick again because the odds are worse for last place teams now.

If I were Tambo I'd wear one of those troll hairdo's to the lottery show.






Definitely a family resemblance there.

nexttothemoon is offline   Reply With Quote
Old
02-18-2013, 05:53 PM
  #142
Red Deer Rebel
Registered User
 
Red Deer Rebel's Avatar
 
Join Date: Apr 2008
Location: Red Deer
Posts: 2,315
vCash: 500
Fourier, is there a statistics website that shows who was on the ice for all Oiler GF and GA, or do we have to sift through the game reports on NHL.com?

The reason I ask is that I would like to look at the video of all GA where Whitney and Gagner are on the ice together, to see how it is that Whitney is supposedly killing Gagner's stats. Being on ice at the same time does not imply causation, but none of these websites seem to take that into account.

Red Deer Rebel is offline   Reply With Quote
Old
02-18-2013, 05:56 PM
  #143
Fourier
Registered User
 
Fourier's Avatar
 
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 8,655
vCash: 500
Quote:
Originally Posted by Red Deer Rebel View Post
Fourier, is there a statistics website that shows who was on the ice for all Oiler GF and GA, or do we have to sift through the game reports on NHL.com?
I am looking for this as well but I think you need to sift through the game reports. I might do this if I get some time. It does not take that long.

Fourier is offline   Reply With Quote
Old
02-18-2013, 06:22 PM
  #144
nexttothemoon
Eight Straight
 
nexttothemoon's Avatar
 
Join Date: Jan 2010
Location: Northern Alberta
Country: Canada
Posts: 8,726
vCash: 50
Yes there's an excellent site with a lot of that info for every player across the NHL.

Here's the page for Gagner for example:

http://stats.hockeyanalysis.com/show...012-13&sit=5v5

Tht site is chock full of data.. you can click all day and it never ends.

Goals
Shots
Corsi
Fenwick


Not much it doesn't have. I'm surprised some team hasn't paid the guy to shut it down and come work for them because there is a crapload of useful stats there.


One thing to be aware of though, I think the stats are always a game or two behind as it takes them a while to update. Still very useful stats database though.

nexttothemoon is offline   Reply With Quote
Old
02-18-2013, 06:33 PM
  #145
Red Deer Rebel
Registered User
 
Red Deer Rebel's Avatar
 
Join Date: Apr 2008
Location: Red Deer
Posts: 2,315
vCash: 500
Response to Fourier

Well, it would be interesting, because according to the stats that you have compiled above, it would seem that Gagner has played a total of 40.1 minutes with Whitney, while RNH has played a total of 31.6 minutes with Whitney. Whitney has only been on for 7 even-strength goals against all year, Gagner has been on for 10, and Nugent-Hopkins has been on for 7.

If Whitney is the reason for Gagner's higher on-ice goals against (which is what Replacement is alleging) then presumably there would be a much higher number of goals against per minute when he is on the ice with not only Gagner, but RNH and the other centers as well. Since Whitney's only been on for a total of 7, I just don't think those numbers are going to work.

Whitney's GAON/60 for the entire year is about the same as Smid's and Nick Schultz's, but it would be interesting to see if his numbers are affected depending on who is playing forward.

My question for Replacement is this: If Whitney's mistakes and fumbles are causing the relatively high numbers of GA when Gagner is on the ice, why isn't the same thing happening when RNH is on the ice? Whitney is a constant, so why would his numbers be better when he's playing with RNH than Gagner, particularly when we know that RNH generally plays against better offensive players - which would presumably mean more scoring chances against when RNH is on the ice?

Red Deer Rebel is offline   Reply With Quote
Old
02-18-2013, 06:36 PM
  #146
Baby Nilsson
Registered User
 
Baby Nilsson's Avatar
 
Join Date: Apr 2006
Location: Victoria
Posts: 1,313
vCash: 500
Quote:
Originally Posted by Red Deer Rebel View Post
Well, it would be interesting, because according to the stats that you have compiled above, it would seem that Gagner has played a total of 40.1 minutes with Whitney, while RNH has played a total of 31.6 minutes with Whitney. Whitney has only been on for 7 even-strength goals against all year, Gagner has been on for 10, and Nugent-Hopkins has been on for 7.

If Whitney is the reason for Gagner's higher on-ice goals against (which is what Replacement is alleging) then presumably there would be a much higher number of goals against per minute when he is on the ice with not only Gagner, but RNH and the other centers as well. Since Whitney's only been on for a total of 7, I just don't think those numbers are going to work.

Whitney's GAON/60 for the entire year is about the same as Smid's and Nick Schultz's, but it would be interesting to see if his numbers are affected depending on who is playing forward.

My question for Replacement is this: If Whitney's mistakes and fumbles are causing the relatively high numbers of GA when Gagner is on the ice, why isn't the same thing happening when RNH is on the ice? Whitney is a constant, so why would his numbers be better when he's playing with RNH than Gagner, particularly when we know that RNH generally plays against better offensive players - which would presumably mean more scoring chances against when RNH is on the ice?
http://stats.hockeyanalysis.com/show...012-13&sit=5v5

Baby Nilsson is offline   Reply With Quote
Old
02-18-2013, 06:41 PM
  #147
Red Deer Rebel
Registered User
 
Red Deer Rebel's Avatar
 
Join Date: Apr 2008
Location: Red Deer
Posts: 2,315
vCash: 500
Quote:
Originally Posted by nexttothemoon View Post
Yes there's an excellent site with a lot of that info for every player across the NHL.

Here's the page for Gagner for example:

http://stats.hockeyanalysis.com/show...012-13&sit=5v5

Tht site is chock full of data.. you can click all day and it never ends.

Goals
Shots
Corsi
Fenwick


Not much it doesn't have. I'm surprised some team hasn't paid the guy to shut it down and come work for them because there is a crapload of useful stats there.


One thing to be aware of though, I think the stats are always a game or two behind as it takes them a while to update. Still very useful stats database though.
... and here is Whitney's page: http://stats.hockeyanalysis.com/show...012-13&sit=5v5.

According to this, when Gagner is on the ice with Whitney at even-strength, the other team scores at a rate of 1.302 per 20 minutes.

When RNH is on the ice with Whitney, the other team scores at a rate of 0.571 per 20 minutes.

How do you explain these numbers, Replacement?


Last edited by Red Deer Rebel: 02-19-2013 at 01:38 PM. Reason: spelling
Red Deer Rebel is offline   Reply With Quote
Old
02-18-2013, 08:01 PM
  #148
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by Fourier View Post
Not surprisingly, your posts show that you really don't understand what these statistics mean. They have nothing to do with Qualcomp or any +/- related statistic. They are a direct measure of how much time each player has played against various opposition players. For example:

RNH has played about 34:05 ES minutes in the two games against SJ. The stats show that his line played roughly 19:25 minutes against the Sharks top pairing of Vlasic and Stuart.

In contrast Gagner played 25:09 mins ES vs SJ. His line faced Matt Hunwick, SJ's #6 defenceman, a little over 14 minutes at ES.
I know exactly what the stats mean. Its not hard to discern. I was stating Qualcomp out of brevity being that I'm not writing an English essay.

My response is meant to denote that looking at who the opponents are early in a season, and ranking the top ordinal opponents, is less important factor in toi spent then looking at who the team mates are. I'm not sure how you would disagree. Unless one could see the top 50 or so opponents the list as is(top 8) is somewhat anecdotal and likely doesn't tell the whole story. What the opponent list does indicate that in the case of the SJ Sharks or AV's that theres some matching. See what I mean?

Quote:
Similarly, RHN's line faced the Sharks top forward line about 17 minutes of the 34 minutes they played at ES. Similarly, the top line faced the Sedin line about 1/2 the time against Vancouver. Against Colorado they played the majority of the game against Duchene's line. Against Phoenix the top line saw OEL for almost 80% of the time they were on the ice. IF you want to look at the TOI stats of every gae the Oilers have played the pattern will be the same.
The first game against the Nucks I could swear Sedins were matched against Gagner line and that they've had some first line matching in other games as well. I don't discount theres likely more games where topline is going against topline but I don't know its always the case.

Quote:
No one is using Qualcomp here. It is undeniable that these players are the best the opposition has to offer.
Yes, in the instances quoted, but again we're talking limited sample and 3, 2 games or 1 game against teams we've played.

The top line spends the majority of time playing against the opposition’s top forwards and best defensemen. Gagner's line gets a mixture of whoever else was left over.


Quote:
As far as sheltering the top line with better defense, I thought bob had already shown this "fact" to be wrong. They have played more minutes with Justin Schultz but if you look at absolute time on the ice with all of the Oilers other defensemen there is very little difference between the two lines. Furthermore, it has already been shown that this extra time with Schultz has almost no real impact on their stats as afar as thing like GAON/60 are concerned.
Justin Schultz alone thus far offers 1.03 less GA when he's on ice. It is a factor. Playing with the inexperienced Yak is also a factor for Gagner. Nor has Hemsky ever been confused with a complete zone player. I think who the respective centers are playing with is significant. You say its negligible, fine, we disagree.

Quote:
In contrast it seems clear that if the first line is playing significantly more minutes against the opposition best offensive players. This you dismiss as meaningless. If there is any spuriuos logic here it would seem to be this.
Pretty sure you're not trying to even consider my point. That the teammates known is a larger proportion of what we know right now then the incidental Thornton or Sedin opponents which may or may not be demonstrative of how all teams match us. I'm prepared to cede the point but I don't know its been definitively established yet.

Quote:
If you want to convince yourself that you have made your case go ahead. It would seem that no one else is buying it.
Nobody has to buy anything. Its simply discussion, a look at some of the stats, and trying to discern impacts.

Thanks as well for responding.


Last edited by Replacement: 02-18-2013 at 08:10 PM.
Replacement is offline   Reply With Quote
Old
02-18-2013, 08:09 PM
  #149
Replacement
Now with 9% more zen
 
Replacement's Avatar
 
Join Date: Apr 2005
Location: Hockey Hell
Country: Canada
Posts: 35,638
vCash: 500
Quote:
Originally Posted by Red Deer Rebel View Post
... and here is Whitney's page: http://stats.hockeyanalysis.com/show...012-13&sit=5v5.

According to this, when Gagner is on the ice with Whitney at even-strength, the other team scores at a rate of 1.302 per 20 minutes.

When RNH is on the ice with Whitney, the other teams scores at a rate of 0.571 per 20 minutes.

How do you explain these numbers, Replacement?
For the moment I would say that its early. So some of the numbers are statistically tenuous due to limited sample involved. Still, the 1.32/20mins is quite high. The 2nd line does indeed seem to struggle a lot when matched with Whitney. Yak is an enormous 2GA/20mins with Whitney suggesting disaster. How is this different then what I stated in terms of Whitneys impact on 2nd line?

Now should the 2nd line be able to withstand such a D, or D pair? Perhaps at this point, and thats an area of other debate. But we have Hemsky, Gagner, and a raw rookie being matched with a bad D. Its not been a helpful association. When Yak ceases being a raw rookie who is having an experience some games I'll reassess.

Nuge on the other hand is a number one pick so I expect more, and expect him to be better, and he's playing with another number one pick who I expect to be a pretty good player, and Eberle the top producer last year. On the first line there is no Yak. This is no knock on yak either, he's learning the NHL game and with some growing pains.

Replacement is offline   Reply With Quote
Old
02-19-2013, 04:51 AM
  #150
Fourier
Registered User
 
Fourier's Avatar
 
Join Date: Dec 2006
Location: Waterloo Ontario
Country: Canada
Posts: 8,655
vCash: 500
Quote:
Originally Posted by Baby Nilsson View Post
Thank you!

This is a great find!

Here are some other intersting observations...

Gagner's GA/20 in the 121 minutes he has played with with Yakupov ====== 0.991.

Gagner's GA/20 in the 54 minutes he has played without Yakuopv ======= 1.103


In fact it is interesting to note that two of the 5 ES goals that Hall and Eberle have been on the ice for have been scored with Gagner also on the ice.

None of this really tells you anything deep about Gagner's defensive play. But what it does do is to show that at least statistically he owns his own shortcomings. They are not someone else's fault.


Last edited by Fourier: 02-19-2013 at 05:32 AM.
Fourier is offline   Reply With Quote
Reply

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 04:55 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.