That would be awesome
One of these is not like the others.
Do you honestly believe that if every player in the league played those pre determined minutes that their p60 numbers would stay exactly how they are?
Better pair to build around: Matthews/Marner or Petterson/Hughes?
- Thread starter Northern Avs Fan
- Start date
that's not their raw production, which is what players get signed based off of
Your posts are built upon the false assumption that production rates are always completely irrelevant for the purposes of determining contract value in the league, even in circumstances in which the obvious and significant issues with raw production are amplified to ridiculous degrees, but you have literally never provided evidence which shows that production rates are irrelevant for the purposes of determining contract value.
TomasHertlsRooster
Don’t say eye test when you mean points
I have provided evidence on many occasions that raw production is the primary factor for driving a forward’s contract value, as well as supporting evidence that the ice time which a forward plays may actually reduce contract value if all else is equal. Here are a few examples:
5V5 P/60 and Cap Hit:
Points and Cap Hit:
(The charts use data from all forwards all forwards with a cap hit above $1M who signed their contract between the end of the 2017-2018 season and the start of the 2018-2019 season.)
Again, a citation is necessary for this. You’ve made the claim and been provided evidence to suggest the contrary, so it’s really on you to prove this.
Tell you what: I will concede this argument and delete my account if you can definitely prove that P/60 is a bigger driver of contract value in the NHL than raw points. That should be plenty enough motivation for you to quickly prove something that you claim to know with absolute certainty.
That is not evidence. 5v5 P/60 is not what I have put forth, much less exclusively. You are ignoring more than half of production. That is you making a graph with pretty horrible methodology over an irrelevant sample for a completely different thing that introduces a whole host of other issues, and then applying it to my analysis to try and discredit it, because you have no actual answer for it. The tweet you provided is actually further evidence that production rates are a factor in negotiations.
Last edited:
TomasHertlsRooster
Don’t say eye test when you mean points
The post with the chart literally says that it is 5V5 P/60; try to read. Further, just because you do not understand regression analysis does not mean that testing the R^2 between two metrics and cap hit is a horrible methodology at all.
The tweet that I posted does not prove that production rates are a factor; the Tweet suggests that if all else is equal, including production, a player will be paid less if they played fewer minutes.
I realized that right after my post. Which actually makes this horribly worse. You are ignoring more than half of production. You are taking my analysis, throwing away all PP production, throwing away all non-5v5 ES production, taking a 1-year sample of production and a few months of contracts from a few wildly different players, comparing it to cap hit with no consideration for term, and then declaring an entire set of stats wrong and irrelevant in all instances, when it correlates worse than something which includes way more information. That's some of the worst and most misleading methodology I've ever seen.
It literally talks about production rates being a factor in the negotiation; actually a big sticking point between them. Thank you for providing evidence for my position.
Last edited:
TomasHertlsRooster
Don’t say eye test when you mean points
I realized that right after my post. Which actually makes this horribly worse. You are taking my analysis, throwing away all PP production, throwing away all non-5v5 ES production, taking a 1-year sample of production and contracts from wildly different players, and then declaring an entire set of stats wrong, when it correlates worse than something which includes way more information. That's some of the worst and most misleading methodology I've ever seen.
It literally talks about production rates being a factor in the negotiation; actually a big sticking point between them. Thank you for providing evidence for my position.
The 5V5 P/60 and raw points in the one year sample was a quick way to look at things, and the fact that the raw points had an overwhelmingly higher R^2 value was rather telling.
McKenzie does not specify in the Tweet that Konecny’s camp was using rates to define “he’s amongst the better even-strength goal-scorers for his age/experience over last two seasons”, and based on the way that league executives discuss things, it’s very likely that he was referring to raw even strength goal totals, rather than rates, where Konecny was still among the better scorers in the league. The only thing that McKenzie clearly states in that Tweet is that Philadelphia was offering less based on ice-time, which wouldn’t really make any sense if teams paid players based on what you claim they do.
As expected, you’ve provided absolutely no evidence to support your position, just looked at somebody else’s evidence and said “actually, this fits my position!” when it clearly doesn’t. You can’t provide anything to show that P/60 is what drives pay in this league.
I’m not sure whether you know it deep down and you’re just burying your head in a wrong argument, or you’ve just convinced yourself of it with no supporting evidence but either way it’s hilarious. Get a grip.
It's not telling of anything. You specifically set up one graph to correlate less, by excluding a whole ton of information with horrible methodology, and then when it does correlate less, you use it to try to discredit other people's work, even when it has nothing to do with it. What's telling, is that over all this time, you have specifically avoided making a graph that is an accurate representation of my analysis methodology.
This is entirely your personal assumption, based on nothing.
If Konecny was talking about raw totals, and the team was coming back with a counter of low ice time, that would make no sense. That would be an argument in favour of Konecny. What this tweet seems to suggest, is that Konecny was discussing his great ES goal scoring rates, and the team attempted to counter that by claiming that those rates took place in a smaller sample. Each side fighting to skew the contract in the direction that they want; in other words, a negotiation.
This says absolutely nothing about how production rates impact contracts; it just shows that TOI is considered along with production.
As expected, you've provided absolutely no evidence to support your position. You have provided useless graphs that have no relation to anything I have shown, and a tweet that you are making a whole bunch of assumptions about; that is actually more likely to support my position.
TomasHertlsRooster
Don’t say eye test when you mean points
What work of yours, which shows that P/60 is a bigger driver of contract value than points, have I tried to discredit? Let’s start there?
(Hint: This work doesn’t actually exist.)
False and unsubstantiated. You have never shown any evidence that this is what is used as a basis for contracts
Pretty much everything I have ever posted.
Literally nobody except the people who have been directly involved in NHL contract negotiations know exactly what is used in contract determination, and how things are weighted. It's likely not exactly the same in every negotiation anyway; there are different impacting factors. None of you have shown evidence of raw points being the sole or best basis for contracts, yet you go around acting like that is all that needs to be considered, and go around attacking and trying to discredit the in-depth analysis of others, with literally no evidence, while demanding evidence of everything under the sun from everyone else.
It was asked why the contracts were what they were. I gave an accurate answer. When evaluating comparable contracts around the league over the past decade and a half, my method has provided overwhelmingly more accurate results than raw points.
I have consistently provided the methodology behind my analysis, why I am doing it in that why, and have answered questions people have related to my work. I've broken it down into point, primary point, and goal production. I'm always open to looking at additional context or factors, like linemate quality, age, etc. I am always looking for ways to improve my model, and have done so countless times. I have shown evidence of TOI differences within the ranges we are discussing having no significant impact on production rates. I have provided abundant evidence that this is an effective tool for player evaluation and contract determination. I have even done the work to display it in an easy to understand form.
You seemingly have no issue with players that are overpaid because they are gifted with inflated opportunity and PP TOI. You seemingly have no issue with players who are overpaid for no good reason at all. Yet for some reason, when a Leaf player and his contract isn't judged with the assumption that they would completely forget how to play and be completely useless on the PP if given the same number as everybody else, suddenly it's a travesty.
TomasHertlsRooster
Don’t say eye test when you mean points
So let me get this straight: you’ve been using P/60 over raw points to evaluate contract value for the past decade and a half?
If that’s the case, why were you on this forum less than two years ago telling another poster that the use of per-minute rates was “ridiculous” and “makes everything messy”?
Now, let me get another thing straight: your “model”. That is pro-rating every player’s point, primary point, and goal totals if they had played exactly 16:30 at even strength and 3:30 on the PP and maintained their exact scoring rate over the past 3 seasons, correct? You believe that your model is a more accurate method of what drives forward contract value than raw points or points per game, correct? And theoretically, if somebody were to test the correlation between cap hit and points, and compare that to cap hit and the results of your model, they’d find that your model is more closely correlated with what players are actually paid, correct?
If you’re going to run around on a forum telling people that you’ve developed a model for contract evaluation that is superior to the methodology which they’ve previously used, and trash the works of other published models such as Evolving Hockey’s and Matt Cane’s, you need to actually show us what that model is and ideally test it yourself, although I’m happy to run some tests for you if you don’t know how.
No, I've been evaluating contracts signed over the cap era; a time period of a decade and a half, with way more accurate results than is gotten with points. You sure do love posting random cherry picked parts of other people's posts from years ago.
I don't think there is one magic way to determine all contracts ever; that is your assumption, despite all graphs you have ever done showing you that is not the case. As I said, there are different impacting factors to varying degrees in different negotiations. You seem to think that if some nothing depth player correlates better with points, then that automatically means that every single player has to follow the exact same formula, even when considering an extremely unique situation; one of the best ES and PP producers in the cap era, for which there is overwhelming evidence, who has temporarily had abnormally restricted opportunity for production relative to his peers and comparables, for reasons entirely outside their control.
Though yes, at least for the purposes of the types of players and contracts being discussed, production rates considered properly provide more accurate results for the purposes of player evaluation and contract determination than raw points.
I have no idea what you would personally find, as you've proven to manipulate your graphs in ways to get your desired results, and have given no care to the massive issues you introduce into them, or how you misrepresent other people's work.
You didn't care that your were representing 5v5 P/60, when it excludes all PP and non-5v5 ES production. Even putting it up against a stat that does include that information, creating misleading and incorrect conclusions.
You didn't care that you were taking a 1 year sample of production, despite me consistently using and putting emphasis on the importance of significant samples.
You didn't care that you were comparing against cap hit, with no consideration for term.
You didn't care that you were looking at contracts solely over the period of a few months.
You didn't care that you were comparing 1m replacement depth to 11m superstars.
You didn't care that you were comparing contracts of varying situations (post-ELC, RFA, UFA) or age.
Etc, etc.
There is no methodology for most people. They look at one raw point number, which has never been proven to be the sole factor at all, and claim that that is all that matters, disregarding all context.
TomasHertlsRooster
Don’t say eye test when you mean points
You say you’ve got a “model” which is far more representative of contract value, yet not only do you refuse to show us proof that your “model” is actually more representative of contract value, you refuse to even tell us what that “model” is so that somebody else can actually test whether or not it is accurate.
It is obvious what you are doing here. You are being intentionally nebulous and extremely careful not to say anything concrete in order not to get caught in a lie by somebody who actually does the work to prove you wrong. It’s even more pathetic than just making a claim without doing your own work.
Last edited:
When asking Dekes for Days to prove a claim, why does he constantly not?
I mean if you say you have a model, then show it. It can’t be that hard.
I mean if you say you have a model, then show it. It can’t be that hard.
I have shown consistent proof over the last year. I have also repeatedly explained my model, which is actually pretty simple, and shown why it is important. You have actively avoided accurate and representative testing for over a year; instead posting wildly misleading graphs and incorrect claims in attempts to discredit my work. You have absolutely zero evidence of anything you are suggesting; you just consistently complain about the work and analysis that me and others do, making demands of them despite no substantiation of your own.
TomasHertlsRooster
Don’t say eye test when you mean points
Especially when it’s a “model” that you have made, that you have so much confidence in, and that you have made thousands of posts on an internet forum defending.
TomasHertlsRooster
Don’t say eye test when you mean points
If you’ve already shown the consistent proof that your “model” is more accurate, and repeatedly explained your “model”, then where can I find this proof and these explanations?
In one of the hundreds of identical discussions we have had where you have needlessly attacked me and my work with no evidence or substantiation.
TomasHertlsRooster
Don’t say eye test when you mean points
If this is true, then it shouldn’t be hard to find one example of you explaining your “model” and showing proof that your “model” is more accurate.
Ad
Upcoming events
-
-
-
BET ON ONLY ONE HORSE WATCH LIVE - FINGER LAKES RACE 1 - Farmington, NYWagers: 2Staked: $540.00Event closes- Updated:
-
-