HFBoards

Go Back   HFBoards > General Hockey Discussion > By The Numbers
By The Numbers Hockey Analytics... the Final Frontier. Explore strange new worlds, to seek out new algorithms, to boldly go where no one has gone before.

Adjusted stats - how valuable?

Reply
 
Thread Tools
Old
11-15-2012, 12:25 PM
  #401
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Iain Fyffe View Post
That assumption is not made. Adjusted stats does not make any assumption about who would score what if they actually played in another scoring environment. It's just a scaling of the actual results in the actual games they played, a transformation of the curve to make it comparable to other seasons in a systematic way. It does not say "this is how someone would have scored has they played at another time" it says "this is how impressive this guy's totals are compared to other players in other times."
So you're saying it's not a replacement for Raw stats, it's just a simple measure of the value of those Raw stats based solely on league GpG and no other factors what so ever?
You and your friends better get your vocabulary straight then because words like "replacement" and "alternative" sure seem to imply a different purpose and use.


Quote:
You misunderstand again. I didn't say it needs to be verified in the sense that it hasn't been yet. I'm saying it is something that needs to be verified before being used. Until something like this is verified it's a conjecture.
You'll have to explain this better because right now it looks like you're using a bunch of double talk to avoid agreeing with what I said.


Quote:
This tends to get lost in all the inaccurate statements made about what adjusted scoring does in this thread, including some made by you.
So you would agree then that anyone suing AS's at face value is just as wrong as the person using Raw stats at face value, correct?
That there are many more factors besides just AS's and Raw stats to be considered before actually arriving at a reasonable conclusion to what a player would have scored in another season?

Quote:
Originally Posted by eva unit zero View Post
You misunderstand the purpose of AS. While it does have its flaws, it is intended to show how the player's performance would have compared to the league in another year. The idea behind removing the player's own numbers is "His stats compare to this year as X; when adjusted to a common average they equal Y. Thus he can be compared against another season's performance, as the "everyone else" group is simply replaced with a different group of player/performances.

As for the Jagr/Gretzky comparison, the use of AS is a valid way of looking at things because the GPG is not the same in any given year and there was in fact wide variation across the 90s.
Ok listen, I was told that the criteria for AS's being more valuable in the Jagr/Gretzky case was because they were at different points in their career and it was unfair to Jagr because the tail end of Gretzky's prime fell in the higher scoring years and that it "pads" his stats too much. Somehow, this is more fair even though it's been well established that there is inflated value attached to DPE stats to which Jagr, in his prime, is getting this benefit over an about to retire Gretzky and I'm supposed to continue to view this as "fair"?

All that aside for a second, I also asked earlier if you would do the same for OV vs Crosby and I was answered yes.
So tell me, what's the criteria for that? Both players at all but identical times of their careers, playing through exactly the same time?

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 02:36 PM
  #402
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
One advantage of adjusted stats over raw stats is that they convert data into a format that allows for comparison based on value.

Let's say there are two supermarkets, one with all prices in Korean Won and the other with all prices in Chilean Pesos. Adjusted stats converts all prices into dollars, which allows for easy comparison. It doesn't change the proportionate prices, as steak will still be the same % higher than potatoes whether or not the prices are converted into dollars. It's possible that the prices (in dollars) for all products are the same at each supermarket, but it's not likely. Maybe one of the markets has a "buy $100, get $5 free" special, which helps wealthier shoppers (better players), while the other doesn't. Maybe the other market has great deals on cheaper goods like top ramen and potatoes, which are more of an advantage to less wealthy shoppers (like cheap/garbage goals for lesser players) for which they represent a larger % of likely purchases. In either case, these distortions would not be the fault of the exchange rate, they would simply be qualities unique to each market (each season) which would have to be studied and their combined effects quantified to determine which market offered better to various types of shoppers (players). It may be that the same market offers "buy $100, get $5 free" and also some sales that would be especially advantageous to poorer shoppers, and that the actual % savings would be about equal. To point to one or the other as "helping" a certain type of shopper would be misleading in the context of the whole range of prices at the market. Each season tends to have a number of these "specials" or "sales", which each need to be examined and need to be combined into the whole picture. Still, there's no denying that using a common currency saves a lot of time and really no information is lost in the process, so it's a superior system.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 02:55 PM
  #403
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
One advantage of adjusted stats over raw stats is that they convert data into a format that allows for comparison based on value.
But they're not a replacement right(even though you constantly call them such) and one has to realise that the value being presented from AS's is soley and completely based only on league GpG.
That a whole host of other factors have yet to be factored in.

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 03:00 PM
  #404
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,340
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
So you're saying it's not a replacement for Raw stats, it's just a simple measure of the value of those Raw stats based solely on league GpG and no other factors what so ever?
Good to have you back!

It is a replacement for raw stats in the sense that it takes the information contained in the raw stats, and adds more information. You really don't lose anything intra-season, and it allows better inter-season comparisons.

Quote:
Originally Posted by Rhiessan71 View Post
You and your friends better get your vocabulary straight then because words like "replacement" and "alternative" sure seem to imply a different purpose and use.
Me and my friends? What with all the lolcats and great big bold fonts, you seem to be taking this quite personally. Chill.

Quote:
Originally Posted by Rhiessan71 View Post
Ok listen, I was told that the criteria for AS's being more valuable in the Jagr/Gretzky case was because they were at different points in their career and it was unfair to Jagr because the tail end of Gretzky's prime fell in the higher scoring years and that it "pads" his stats too much. Somehow, this is more fair even though it's been well established that there is inflated value attached to DPE stats to which Jagr, in his prime, is getting this benefit over an about to retire Gretzky and I'm supposed to continue to view this as "fair"?
How much is the "inaccuracy" inherent to the DPE figures? How does this "inaccuracy" compare to that from the basic change in scoring level? If the latter is greater, then yes, AS are more fair in this case. You're making assumptions about the degree of these inaccuracies, so to speak. If you have something solid to show that the former is significantly greater than the latter, please share.

Based on your comments it would seem that AS should have something of a tier adjustment added to it. If that were done, this objection of yours would presumably disappear, and raw stats would be rendered even more obsolete for this type of analysis.

Iain Fyffe is offline   Reply With Quote
Old
11-15-2012, 03:05 PM
  #405
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,340
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
But they're not a replacement right(even though you constantly call them such) and one has to realise that the value being presented from AS's is soley and completely based only on league GpG.
That a whole host of other factors have yet to be factored in.
The same can be said about raw stats, of course. The caveats that apply to AS also apply to raw stats, and AS adds a bit of additional information, so it's not unreasonable to call AS a replacement.

Where I agree with your argument that they should not be considered the end-all is that AS do lack certain contexts (which raw stats also lack). So you can't say one player was necessarily more impressive than another because his adjusted points were higher. You need to consider his linemates, PP opportunities, etc., which you would also have to do if using raw stats. One thing you don't have to do it consider the league-wide scoring environment, because that's already been considered.

Adjusted stats over raw stats? Sure. Adjusted stats as a final answer to a particular question? Not likely.

Iain Fyffe is offline   Reply With Quote
Old
11-15-2012, 03:06 PM
  #406
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Rhiessan71 View Post
But they're not a replacement right(even though you constantly call them such) and one has to realise that the value being presented from AS's is soley and completely based only on league GpG.
That a whole host of other factors have yet to be factored in.
To me, retaining all the previous info (rankings, proportions) and gaining new info (value in context of league scoring) essentially means replacement, but I'm not preventing anyone from using raw data.

I agree that there's a whole host of other factors yet to be factored in. How important those factors are and how they can be analyzed and quantified properly are what we should be discussing. Some of those will either be quite difficult to quantify (but may be captured in the "error" portion of certain models and/or by other variables) or will basically be inconsequential (most ignore aerodynamics when describing a large boulder's path down a hill). I think most of the important variables can be defined and quantified to a large degree, and hopefully that will improve the accuracy of adjusted stats for the "predictive" purposes to which you've alluded.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 04:15 PM
  #407
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Iain Fyffe View Post
Good to have you back!

It is a replacement for raw stats in the sense that it takes the information contained in the raw stats, and adds more information. You really don't lose anything intra-season, and it allows better inter-season comparisons.
No, it's not a replacement.
First off, you and countless others have repeatedly tried to say that AS's are not predictions, they are just an estimate of value.
In what formula would you replace Raw data with estimations when the formula still has equations to be done that need to run off the Raw data?
I have already shown that AS's have inaccuracies in them compared to the raw stats. You call them rounding errors, whatever, it doesn't matter, they're there.
If you replace the Raw stats with Adjusted stats before all other factors are determined, you are working from a modified and inaccurate base for those other factors to be used on.


Quote:
Me and my friends? What with all the lolcats and great big bold fonts, you seem to be taking this quite personally. Chill.
Because it's god damned frustrating when I say that all info should be used and weighed, that nothing should be thrown out. That Adjusted Stats should never be used as a final answer, that they are only part of the equation, along with everything else.
That when things that are proven to exist like Tier scoring changes, special teams and inflated value of DPE points are deemed as "conjecture" not because they don't exist but only because no one has figured out how to incorporate them globally yet.
That in one sentence I'm being told that Adjusted stats are only a measure of value and not a prediction, only to have them being used as exactly that in the next sentence telling me that this player actually did outscore that player over this time frame.
That after saying and then hearing all this, I am then told my views are all but "barbaric" or that I have some kind of bias against AS's.
No sir, I think I have a pretty balanced view of AS's and what their true weight should be. A weight that will change per comparison because they are part of the equation and not the equation itself.
That it's not I with a bias against AS's, it's that I have a bias against people who are biased for AS's.

I have never said to throw out AS's. I have always said to use both.
Yet throughout this thread, there are numerous examples of people just throwing Raw stats out or the avocation of that behavior and I'm not the impartial one eh...riiight.

Quote:
How much is the "inaccuracy" inherent to the DPE figures? How does this "inaccuracy" compare to that from the basic change in scoring level? If the latter is greater, then yes, AS are more fair in this case. You're making assumptions about the degree of these inaccuracies, so to speak. If you have something solid to show that the former is significantly greater than the latter, please share.
See...this leads to even more frustration. Common sense.
Delving into DPE values, it doesn't take long to see that there's some funky **** and, at times, some large anomalies going on.
Everyone knows there's some issues there, we just can't figure out exactly where or how yet and therefore can't address it with an exact offsetting value yet.

If your car is acting up and you haven't been to the mechanic to figure out exactly what's wrong with it yet. You might continue to drive it around town to run errands and go to work (using DPE values against other DPE values) but you're not going to drive out of town or go on a long trip with it(take DPE values out of the DPE) because you simply can't trust it.


Quote:
Based on your comments it would seem that AS should have something of a tier adjustment added to it. If that were done, this objection of yours would presumably disappear, and raw stats would be rendered even more obsolete for this type of analysis.
It would help to give the values AS's relay more weight for the equation but I am NEVER going to replace Raw stats completely.
I think I have made that pretty clear by now and why.


Last edited by Rhiessan71: 11-15-2012 at 04:28 PM.
Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 04:20 PM
  #408
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 22,912
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
Because it's god damned frustrating when I say that all info should be used and weighed, that nothing should be thrown out. That Adjusted Stats should never be used as a final answer, that they are only part of the equation, along with everything else.
That when things that are proven to exist like Tier scoring changes, special teams and inflated value of DPE points are deemed as "conjecture" not because they don't exist but only because no one has figured out how to incorporate them globally yet.
That in one sentence I'm being told that Adjusted stats are only a measure of value and not a prediction, only to have them being used as exactly that in the next sentence telling me that this player actually did outscore that player over this time frame.
Who is telling you, in this thread or otherwise, that Adjusted Stats are the be-all and end-all final answer?

Anyone who has told you that Adjusted Stats are perfect is lying to you. However, I haven't seen that claim offered up in this thread.

Chalupa Batman is offline   Reply With Quote
Old
11-15-2012, 04:33 PM
  #409
habsfanatics
Registered User
 
Join Date: May 2012
Posts: 3,384
vCash: 500
Quote:
Originally Posted by Taco MacArthur View Post
Who is telling you, in this thread or otherwise, that Adjusted Stats are the be-all and end-all final answer?

Anyone who has told you that Adjusted Stats are perfect is lying to you. However, I haven't seen that claim offered up in this thread.
I tend to side with the guys who know the numbers better than I do, like yourself and Iainn and czech your math, but I see some valid arguments from Rhiessan that seemed dismissed out of hand.

I have seen throughout this thread that even guys that know the numbers better than I do and insist themselves that there are some flaws and then turn around and immediately post about so and so doing this and using AS as their only source of proof. Czech has done it several times now, even admitting to not seeing many of the seasons or games discussed prior to the lockout. This isn't meant to bash, but I see them being used as the be-all end-all quite often.

The only point I have and have no idea if it really has any merit is that it would seem likely that the distribution of points wouldn't remain static from top tier to bottom tier players. If the league gets tougher, it may not impact the skilled guys as much as the tweeners. Seems pretty logical, no?

habsfanatics is online now   Reply With Quote
Old
11-15-2012, 04:40 PM
  #410
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 22,912
vCash: 500
Quote:
Originally Posted by habsfanatics View Post
I tend to side with the guys who know the numbers better than I do, like yourself and Iainn and czech your math, but I see some valid arguments from Rhiessan that seemed dismissed out of hand.

I have seen throughout this thread that even guys that know the numbers better than I do and insist themselves that there are some flaws and then turn around and immediately post about so and so doing this and using AS as their only source of proof. Czech has done it several times now, even admitting to not seeing many of the seasons or games discussed prior to the lockout. This isn't meant to bash, but I see them being used as the be-all end-all quite often.

The only point I have and have no idea if it really has any merit is that it would seem likely that the distribution of points wouldn't remain static from top tier to bottom tier players. If the league gets tougher, it may not impact the skilled guys as much as the tweeners. Seems pretty logical, no?
If anyone - Czech or otherwise - is using adjusted stats as the alpha/omega of an argument, that's a problem.

Regarding your last paragraph, I agree completely - although if we're attempting to measure "value", then it raises different questions entirely.

The chief adjustment that adjusted statistics make is to adjust for different goal environments. If a goal is scored in a league where the average score is 3-2, that's a lot more valuable than a goal scored in a league where the average score is 8-6. If you're going to do that, then you probably need to take it a step further, and instead of adjusting to league-wide scoring levels, you need to adjust to a schedule-weighted scoring level (since games in the 1980s Smythe Division had less-valuable goals than games in the 1980s Norris Division, for instance).

Then do you go further, and adjust for situations? (A goal scored in a 1-1 tie game is more valuable than a goal scored at the end of an 8-1 win). And at what point does the sample size wash out the value in the information?

Adjusted statistics are never the ultimate result - but they add information, and they're generally more informative than raw numbers.

Chalupa Batman is offline   Reply With Quote
Old
11-15-2012, 04:42 PM
  #411
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Taco MacArthur View Post
Who is telling you, in this thread or otherwise, that Adjusted Stats are the be-all and end-all final answer?
I guess I misunderstood all of you telling me straight up that Jagr had more value than Gretzky from 90/91-98/99. Was that not a final answer on your part? Did I miss a follow up conversation or explanation of other factors and how they were applied for you to make that determination that Jagr had more value?


Quote:
Anyone who has told you that Adjusted Stats are perfect is lying to you. However, I haven't seen that claim offered up in this thread.
I know, I'm confused too.
Why would anyone advocate the complete throwing away of Raw Stats if you didn't have something perfect to replace it with?
I mean, that's the only conclusion I can come up with when it seems much more rational and correct to me, to keep and use both at various weights.

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 04:44 PM
  #412
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 22,912
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
Why would anyone advocate the complete throwing away of Raw Stats if you didn't have something perfect to replace it with?
I mean, that's the only conclusion I can come up with when it seems much more rational and correct to me, to keep and use both at various weights.
Don't let "perfect" be the enemy of "good".

We're never going to get a perfect measure of a hockey player's contribution to games. It can't even be done in baseball, which is a game of discrete events. That doesn't mean that we shouldn't try to account for what we can account for.

Chalupa Batman is offline   Reply With Quote
Old
11-15-2012, 04:57 PM
  #413
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Taco MacArthur View Post
Don't let "perfect" be the enemy of "good".

We're never going to get a perfect measure of a hockey player's contribution to games.
Oh dude...I've been saying this since the beginning.
That making comparisons are about more than just the numbers.

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 05:01 PM
  #414
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by habsfanatics View Post
I have seen throughout this thread that even guys that know the numbers better than I do and insist themselves that there are some flaws and then turn around and immediately post about so and so doing this and using AS as their only source of proof. Czech has done it several times now, even admitting to not seeing many of the seasons or games discussed prior to the lockout. This isn't meant to bash, but I see them being used as the be-all end-all quite often.
Whos has using them as a be-all end-all? Adjusted stats, whether production or plus-minus, are not perfect, but I believe they are by far the most objective data available for evaluating forwards. I didn't say they are the only data that should be considered. However, people use rankings, awards, Cups, etc. with often little or no "adjustment" for how much easier or more difficult it may be to earn each in different eras and seasons. I understand there are potential misuses of adjusted stats, and I'm not saying I've never changed my interpretation of them as I had more information. Pretending that raw stats are generally as useful as adjusted stats is misleading as well IMO. Adjusted stats simply takes the first and by far the most important steps in the adjustment process. It corrects for the changing goals/win (league gpg), wins/season (length of schedule) and assists/goal (how often assists awarded on goals). That is what I call "simple adjustment". Anything past that, including HR.com's adjusting for player's own scoring and roster size, is entering the realm of adjusting further for difficulty/quality of the individual's season (e.g. a smaller roster doesn't change the value of the production... it changes the difficulty of attaining that production level... and of the minimum quality of player necessary to attain that level).

As to the bolded, if you mean before the '95 lockout, I generally watched less hockey then than in the several years following. That doesn't mean I wasn't watching the playoffs when I could, and following the seasons as they happened to some degree. The NHL was more widely available in the 90s on Fox, ESPN, etc. (those were some of the best days for free hockey) and I could/would go to bars/restaurants to watch playoffs on satellite as well. IIRC, when I was younger in the late 70s & much of the 80s, there just wasn't as much coverage of the NHL outside Canada. However, I remember watching Lafleur & Dryden on live TV, so it's not like I just showed up to watch Crosby & Ovechkin either.

Quote:
Originally Posted by habsfanatics View Post
The only point I have and have no idea if it really has any merit is that it would seem likely that the distribution of points wouldn't remain static from top tier to bottom tier players. If the league gets tougher, it may not impact the skilled guys as much as the tweeners. Seems pretty logical, no?
If everything stayed the same, then one would expect the distribution of points between tiers to remain equal. However, a number of things changed: avg. gpg decreased... new equipment... possibly new coaching/systems... more power plays... more Euro/US players, esp. at top tier forward... expansion. Without knowing which factors caused the observed effect, it's difficult to further adjust properly. I think "simple adjusted" is a good base from which to explore further. I've repeatedly suggested using regression analysis, because it measures multiple factors simultaneously, some of which may be correlated and/or contemporaneous changes.

It's unclear what you mean by the league getting "tougher" and how your conclusion follows from that. If you mean "lower scoring", then it depends on which factors are responsible, as to how (and how much) to adjust further. It also got tougher in terms of the total talent pool, and the average quality of top forwards. These factors could change the dynamic between tiers, since the lower tiers may not have improved as much as the top tiers during the Euro/Russian/US influx. A lot of things may seem logical, but that doesn't mean they're always true, esp. since there are multiple factors at work.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 05:08 PM
  #415
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Rhiessan71 View Post
I guess I misunderstood all of you telling me straight up that Jagr had more value than Gretzky from 90/91-98/99. Was that not a final answer on your part? Did I miss a follow up conversation or explanation of other factors and how they were applied for you to make that determination that Jagr had more value?
If you mean who was more valuable from '91-99, to me it's clearly Jagr. Adjusted plus-minus and even strength points are likely substantially more in Jagr's favor than adjusted points are. So it was an older Gretzky, who abandoned defense and didn't have the strength to play a possession game, versus a one-man wrecking crew who was skating through whole teams and was could only be stopped by 2-3 guys hooking and holding the whole shift. Is that enough follow up conversation for you?

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 05:29 PM
  #416
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
If you mean who was more valuable from '91-99, to me it's clearly Jagr. Adjusted plus-minus and even strength points are likely substantially more in Jagr's favor than adjusted points are. So it was an older Gretzky, who abandoned defense and didn't have the strength to play a possession game, versus a one-man wrecking crew who was skating through whole teams and was could only be stopped by 2-3 guys hooking and holding the whole shift. Is that enough follow up conversation for you?
No, actually it isn't and this should have been said at the time instead of simply presenting the AS's values as the answer which IS exactly what you and other did! TO WHICH MY POINT MOST DEFINITELY STILL STANDS!!!
Jagr was not a "one-man wrecking crew" for all 9 seasons in question, nor was Gretzky an older player who abandoned defense for all 9 either.
What you're describing is what happened in the last 4-5 years, not the first 4-5 years.

The way I see it is that Gretzky was a better and more valuable player in the last 3-4 years than Jagr was in the first 3-4 years.
They go back and forth pretty even in the middle leaving us 2-3 years of higher play from each to compare (Gretz in the early 90's vs Jagr in the late 90's) and I don't see the same gap there that you do apparently and I certainly do not see that gap, if one really exists, having enough value to overcome Gretzky's superiority in his late years vs Jagr's early years.


Last edited by Rhiessan71: 11-15-2012 at 05:35 PM.
Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 06:30 PM
  #417
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Rhiessan71 View Post
No, actually it isn't and this should have been said at the time instead of simply presenting the AS's values as the answer which IS exactly what you and other did! TO WHICH MY POINT MOST DEFINITELY STILL STANDS!!!
Jagr was not a "one-man wrecking crew" for all 9 seasons in question, nor was Gretzky an older player who abandoned defense for all 9 either.
What you're describing is what happened in the last 4-5 years, not the first 4-5 years.

The way I see it is that Gretzky was a better and more valuable player in the last 3-4 years than Jagr was in the first 3-4 years.
They go back and forth pretty even in the middle leaving us 2-3 years of higher play from each to compare (Gretz in the early 90's vs Jagr in the late 90's) and I don't see the same gap there that you do apparently and I certainly do not see that gap, if one really exists, having enough value to overcome Gretzky's superiority in his late years vs Jagr's early years.
Here's some more stats (some actual, some adjusted) for you ('91-99):

ES points (actual)
Gretzky 500
Jagr 598

ES points (adj.)
Gretzky 717
Jagr 892

ESGF/ESGA ratios (est. actual)
Gretzky: 1.53, 0.95, 1.24, 0.75, 0.68, 0.94, 1.19, 0.97, 0.69
Jagr: 0.96, 1.22, 1.74, 1.29, 1.71, 1.50, 1.47, 1.56, 1.37,

ESGF/ESGA (est. on/off)
Gretzky: 1.10, 1.00, 1.23, 0.88, 0.74, 1.25, 1.06, 1.34, 0.78
Jagr: 0.85, 1.20, 1.42, 1.35, 1.78, 1.44, 1.88, 1.63, 1.70

Avg. of ratios '91-99
Gretzky: 0.99 (est. actual), 1.04 (est. on/off)
Jagr: 1.42 (est. actual), 1.47 (est. on/off)

Gretzky's second best actual ES GF/GA ratio was less than Jagr's 7th best, and his best on/off ratio was less than Jagr's 7th best.

Gretzky's overall effectiveness at ES was nothing like it was in his peak years. He was racking up points on the PP in the early 90s, with a prime Robitaille, Sandstrom, and an older Kurri, with Blake and/or Zhitnik on the points. Meanwhile, Jagr was getting limtied or virtually no PP time in the early 90s, deferred to Lemieux running the PP when he has healthy, and then ran the PP with an older Francis (until he left) with Kevin Hatcher as the best PP d-man (he was not exactly Leetch). Gretzky had a substantial advantage due to skill, experience and opportunity in '91 & '92... they were close in points in '93 & '94, but Jagr was much more effective at ES... and then Jagr was clearly much better after and including '95.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 06:46 PM
  #418
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
Here's some more stats (some actual, some adjusted) for you ('91-99):

ES points (actual)
Gretzky 500
Jagr 598

ES points (adj.)
Gretzky 717
Jagr 892

ESGF/ESGA ratios (est. actual)
Gretzky: 1.53, 0.95, 1.24, 0.75, 0.68, 0.94, 1.19, 0.97, 0.69
Jagr: 0.96, 1.22, 1.74, 1.29, 1.71, 1.50, 1.47, 1.56, 1.37,

ESGF/ESGA (est. on/off)
Gretzky: 1.10, 1.00, 1.23, 0.88, 0.74, 1.25, 1.06, 1.34, 0.78
Jagr: 0.85, 1.20, 1.42, 1.35, 1.78, 1.44, 1.88, 1.63, 1.70

Avg. of ratios '91-99
Gretzky: 0.99 (est. actual), 1.04 (est. on/off)
Jagr: 1.42 (est. actual), 1.47 (est. on/off)

Gretzky's second best actual ES GF/GA ratio was less than Jagr's 7th best, and his best on/off ratio was less than Jagr's 7th best.

Gretzky's overall effectiveness at ES was nothing like it was in his peak years. He was racking up points on the PP in the early 90s, with a prime Robitaille, Sandstrom, and an older Kurri, with Blake and/or Zhitnik on the points. Meanwhile, Jagr was getting limtied or virtually no PP time in the early 90s, deferred to Lemieux running the PP when he has healthy, and ran the PP with an older Francis (until he left) with Kevin Hatcher as the best PP d-man (he was not exactly Leetch). Gretzky had a substantial advantage due to skill, experience and opportunity in '91 & '92... they were close in points in '93 & '94, but Jagr was much more effective at ES... and then Jagr was clearly much better after and including '95.
Sooo...you say that Jagr didn't get a lot of PP time for a while, therefore scored most of his points at ES.
Then you try and use that ES to PP ratio, that of course is going to show that Jagr scored a greater % of his points at ES, against Gretzky, who had a more even distribution of PP to ES because he played more on the PP in the beginning of the decade than Jagr did?
Just making sure I'm getting this right.
I mean really if anything it just goes to prove that Gretzky's value offensively was that much greater than Jagr's through those years.
And that those ratio's of ESGF/ESGA are all well and good but If Jagr is only playing 15 minutes a game while Gretzky plays 20-25.
And Gretzky is facing the other teams best checkers early in the decade while Jagr wasn't...how valuable are those first half decade numbers really?

Later on in the decade when Gretzky is a former shadow of himself and Jagr is in his prime. Jagr is blowing Gretz away at even strength.
Tell me though, how is counting these advantages as "equal value" any different than me trying to count Gretzky's earlier decade advantages as equal value over a rookie Jagr?

Maybe I'm wrong here but it sure looks like all you're trying to do is minimize Gretzky's advantage years while maximizing Jagr's advantage years.
Anyone else see the same?

Oh and rationalizing it like a madmen after, still doesn't change the FACT that you used Adjusted Stats as a final answer.


Last edited by Rhiessan71: 11-15-2012 at 07:00 PM.
Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 07:23 PM
  #419
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Rhiessan71 View Post
Sooo...you say that Jagr didn't get a lot of PP time for a while, therefore scored most of his points at ES.
Then you try and use that ES to PP ratio, that of course is going to show that Jagr scored a greater % of his points at ES, against Gretzky, who had a more even distribution of PP to ES because he played more on the PP in the beginning of the decade than Jagr did?
Just making sure I'm getting this right.
I didn't use the ES to PP ratio for any of those calculations. Adjusted ES points simply means actual ES points adjusted for the ES GPG in the league for each season.

Quote:
Originally Posted by Rhiessan71 View Post
I mean really if anything it just goes to prove that Gretzky's value offensively was that much greater than Jagr's through those years.
Say what? The player who relied more on a stacked PP for his points, but couldn't compare at ES was that much greater because of that? Interesting conclusion.

Quote:
Originally Posted by Rhiessan71 View Post
And that those ratio's of ESGF/ESGA are all well and good but If Jagr is only playing 15 minutes a game while Gretzky plays 20-25.
And Gretzky is facing the other teams best checkers early in the decade while Jagr wasn't...how valuable are those first half decade numbers really?
If you'd prefer total adjusted plus-minus for each season, which would generally help the player with more minutes:

Gretzky: 12, 0, 10, -10, -19, 13, 5, 18, -12 (+17 total)
Jagr: -10, 13, 25, 23, 44, 34, 40, 29, 42 (+142 total)

Quote:
Originally Posted by Rhiessan71 View Post
Oh and rationalizing it like a madmen after, still doesn't change the FACT that you used Adjusted Stats as a final answer.
I'm certainly not going to depend on the tally of a bunch of hungover North American sportswriters to form the main basis of opinion for more modern players. The adjusted numbers may not be perfect, but they show a substantial difference in value that matches what I saw. After '92, Gretzky's only real advantage was scoring more on the PP for a couple years, when Jagr wasn't getting nearly the opportunity on the PP, but was more effective in ES ice time (for which they had more similar opportunity). There's also a difference between being limited by opportunity, and being limited by age/injuries.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 07:37 PM
  #420
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Taco MacArthur View Post
If anyone - Czech or otherwise - is using adjusted stats as the alpha/omega of an argument, that's a problem.
That's where we disagree then. It depends on the amount and nature of data available, but it seems the other choices are:

The Eye Test- Arguing about what one thinks one remembers one's eyes seeing over (in most cases) a limited sample. Sure seems like a lot of room for bias, subjectivity, etc. Esp. difficult to reach consensus with others when there is initial disagreement.

Rankings amongst peers- From what I've seen, mostly or completely ignores the changing quality of peer talent, which has changed dramatically over time.

Awards- Same failure to recognize changing peer talent, but add in the bias and subjectivity of some oft-ignorant sportswriters.

Cups/Playoffs- Even those who use adjusted stats effectively often forget that playoff stats were affected in mostly very similar ways that effected regular season scoring. The role of team quality, % teams in playoffs, league size & parity, etc. are also often ignored by many or not fully appreciated.

They all can be used for evaluation. I don't tell anyone else what to use or that it's "wrong" when context is properly considered. However, there's a reverence, esp. for O6 players, that seems to prevent the realization that they were big fish in a much smaller pond... when it's a lot easier to be one of the big fish. Smaller population... no Euros/Russians to deal with... no huge leagues where lesser players can have fluke big seasons instead of being relegated to the third line on a powerhouse. I see respected posters claiming Howie Morenz should be highly ranked based on being a big fish in a much, much smaller pond... an ancient poll... and some yearly polls of "the usual suspected experts"... and I'm supposed to bow to that and agree that preferring objective evidence is "wrong." I can't see that happening.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 07:56 PM
  #421
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
I'm certainly not going to depend on the tally of a bunch of hungover North American sportswriters to form the main basis of opinion for more modern players. The adjusted numbers may not be perfect, but they show a substantial difference in value that matches what I saw. After '92, Gretzky's only real advantage was scoring more on the PP for a couple years, when Jagr wasn't getting nearly the opportunity on the PP, but was more effective in ES ice time (for which they had more similar opportunity). There's also a difference between being limited by opportunity, and being limited by age/injuries.
You have 3 choices here...

A) You can simply swallow your pride a bit and admit that you got carried away and used AS's as a final answer. Something which I cited previously as a big issue with them. You just add that you'll do your best not to do so again and we can both move on.

B) You can stubbornly keep on rationalizing it and I'll keep on telling you so until another mod comes in and tells us both to stop.

C) You can just not respond anymore, not swallow your pride a bit and just ignore that you did it and I won't have anything to respond to but myself and everyone else will know.

I'm cool with any of the options.

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 08:02 PM
  #422
SaintPatrick33
Conn Smythe Winner
 
Join Date: Aug 2012
Posts: 3,040
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
The Eye Test- Arguing about what one thinks one remembers one's eyes seeing over (in most cases) a limited sample. Sure seems like a lot of room for bias, subjectivity, etc. Esp. difficult to reach consensus with others when there is initial disagreement.
I think though, you aren't looking at a limited sample. It's more the case of people who have watched whatever being discussed for years.

More to the point, there's always going to be a measure of subjectivity involved since no human endeavor can ever be completely boiled down to a metric. Metrics are useful tools, but they are NEVER going to be definitive.


Quote:
Originally Posted by Czech Your Math View Post
Rankings amongst peers- From what I've seen, mostly or completely ignores the changing quality of peer talent, which has changed dramatically over time.
The problem here is that the changing quality of talent is balanced by the fact that since the Original Six the NHL has expanded as fast as the talent pool has grown.


Quote:
Originally Posted by Czech Your Math View Post
Awards- Same failure to recognize changing peer talent, but add in the bias and subjectivity of some oft-ignorant sportswriters.
Yet at the same time you end up disregarding the views of some very knowledgeable sports writers. It isn't necessary to throw out the baby with the bath water.


Quote:
Originally Posted by Czech Your Math View Post
Cups/Playoffs- Even those who use adjusted stats effectively often forget that playoff stats were affected in mostly very similar ways that effected regular season scoring. The role of team quality, % teams in playoffs, league size & parity, etc. are also often ignored by many or not fully appreciated.
Unless one is talking goaltenders I don't think the Cup argument is appropriate. And even then only in certain eras where goaltending dominates.

SaintPatrick33 is offline   Reply With Quote
Old
11-15-2012, 08:04 PM
  #423
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 3,373
vCash: 50
Quote:
Originally Posted by Rhiessan71 View Post
You have 3 choices here...

A) You can simply swallow your pride a bit and admit that you got carried away and used AS's as a final answer. Something which I cited previously as a big issue with them. You just add that you'll do your best not to do so again and we can both move on.
That's rather authoritarian of you. Get over yourself, I'm not 8 and no one's gonna tell anyone which data they can or can't use.

Quote:
Originally Posted by Rhiessan71 View Post
B) You can stubbornly keep on rationalizing it and I'll keep on telling you so until another mod comes in and tells us both to stop.

C) You can just not respond anymore, not swallow your pride a bit and just ignore that you did it.
Rationalize what? Using adjusted stats as the best estimate of production, and (in context) using them as a large portion of my basis of evaluation? There's no pride involved, it's what I find to be the fairest, most objective evidence available to establish what production for forwards. Considering that offensive production and overall ES effectiveness (both in context) seem the main priorities of most top forwards, then the best estimates of such seem to be worthy of strong consideration as very important factors in the process of evaluation. If any of you want to ignore all data completely and pick numbers out of a hat, I can't prevent you from doing so. That won't prevent me from stating my position and reasons for that position founded in fact (data) and logic, not blurry memories or someone else's opinion.

Czech Your Math is offline   Reply With Quote
Old
11-15-2012, 08:12 PM
  #424
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,881
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
That's where we disagree then. It depends on the amount and nature of data available, but it seems the other choices are:

The Eye Test- Arguing about what one thinks one remembers one's eyes seeing over (in most cases) a limited sample. Sure seems like a lot of room for bias, subjectivity, etc. Esp. difficult to reach consensus with others when there is initial disagreement.

Rankings amongst peers- From what I've seen, mostly or completely ignores the changing quality of peer talent, which has changed dramatically over time.

Awards- Same failure to recognize changing peer talent, but add in the bias and subjectivity of some oft-ignorant sportswriters.

Cups/Playoffs- Even those who use adjusted stats effectively often forget that playoff stats were affected in mostly very similar ways that effected regular season scoring. The role of team quality, % teams in playoffs, league size & parity, etc. are also often ignored by many or not fully appreciated.

They all can be used for evaluation. I don't tell anyone else what to use or that it's "wrong" when context is properly considered. However, there's a reverence, esp. for O6 players, that seems to prevent the realization that they were big fish in a much smaller pond... when it's a lot easier to be one of the big fish. Smaller population... no Euros/Russians to deal with... no huge leagues where lesser players can have fluke big seasons instead of being relegated to the third line on a powerhouse. I see respected posters claiming Howie Morenz should be highly ranked based on being a big fish in a much, much smaller pond... an ancient poll... and some yearly polls of "the usual suspected experts"... and I'm supposed to bow to that and agree that preferring objective evidence is "wrong." I can't see that happening.
Translation: You start with Adjusted Stats and discard anything that doesn't agree with them.
Exactly what I said you do, earlier in the thread.

And hey, that's fine. If that's the way you chose to evaluate players, that's your choice.
I'm just glad to finally get a straight answer about it.

I'm good now, got the confirmation of my suspicions and won't continue "muck up" the thread any more.


Cheers
R71

Rhiessan71 is offline   Reply With Quote
Old
11-15-2012, 08:21 PM
  #425
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,340
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
No, it's not a replacement.
First off, you and countless others have repeatedly tried to say that AS's are not predictions, they are just an estimate of value.
In what formula would you replace Raw data with estimations when the formula still has equations to be done that need to run off the Raw data?
Any of them, since converting between raw stats and AS is dirt simple. It's trivial to calibrate a raw-stats-based equation to use AS instead.

Quote:
Originally Posted by Rhiessan71 View Post
I have already shown that AS's have inaccuracies in them compared to the raw stats. You call them rounding errors, whatever, it doesn't matter, they're there.
So show your results with decimals, and don't remove the individual player stats when making the calculation. There, you're now using adjusted stats without these "inaccuracies". They're still adjusted stats, and have none the of (very small) inaccuracies you see.

Quote:
Originally Posted by Rhiessan71 View Post
Why would anyone advocate the complete throwing away of Raw Stats if you didn't have something perfect to replace it with?
Pardon? Shouldn't the criteria for replacement be better, not necessarily perfect?

Iain Fyffe 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 02:17 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.