Oh My God it's Finally Happening!

Non-league talk in here. Please make NSFW *links* and not pics.
User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19815
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1982 times
Been thanked: 2902 times

Re: Oh My God it's Finally Happening!

Post by RonCo » Sat Aug 11, 2018 3:32 pm

Ted wrote:
Sat Aug 11, 2018 3:10 pm

We're just assuming putouts are automatic. Have watch a stone fisted, lead footed clutz receive throws for most of season now, I can tell you they are not.
You're arguing for a modification of the adjustment, not removing it all together. Note that the adjustment of a 1B is different from a DH--which is needed to take into account the base fielding value of a 1B over a DH.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: Oh My God it's Finally Happening!

Post by Ted » Sat Aug 11, 2018 3:34 pm

Ron, I'm not arguing that their shouldn't be positional adjustments. I'm arguing that the method is BAD. Pujols does NOT add value by playing in the field. This is not true. He is WORSE than an average 1B. He can't freaking move anymore. Any metric that gives him more value simply for having a certain position penciled next to his name is needs improvement.

You can incorporate positional adjustments into WAR without a flat number being added or subtracted. I've read that fangraphs link you posted many many times are their are so many glaring issues with the explanation that it even ADMITS.

First, there's this

"In general, we want to add runs for players who play tough positions and subtract runs for players who play easier positions to account for the fact that average at one does not equal average at the other in terms of total run prevention. For example, for a shortstop, the adjustment is +7.5 runs per full season. For a left fielder, it’s -7.5 runs per full season."

WHAT????? Read that again. Average at one positions HAS to equal average at another position in terms of relative run prevention. It has to. It's mathematically impossible for it to not. In terms of ABSOLUTE run prevention, of course it doesn't, which is what that paragraph says of course. But why should I care around total or absolute run prevention when I'm trying to use a metric to compare a player to a theoretical replacement level in terms of RELATIVE value?

It goes on "That means the difference between an average left fielder and an average shortstop is about 15 runs per season. That’s a sizable gap."

Now we're just defining ourselves in terms of ourselves. That's just bad science.

Then there's this

"The actual numbers we use are based on some calculations that were done about a decade ago that used the performance of players who moved positions. It’s certainly reasonable to suggest that those numbers have changed as the game has changed, so use the adjustments as guides more than as firm rules. The DH adjustment might be too negative because it’s harder to hit when you’re not playing in the field, and the catcher adjustment might be a bit too large. There’s lots of room to disagree on the precise decimals and if you’re so inclined, I’d invite you to come up with a more accurate rendering of the numbers. Here is one recent example of such work."

The players used were primarily utility players who do NOT represent typical players. And here the author basically admits that the adjustments are wonky. And THIS is where my biggest problem with them lies. Baseball changes. We should not be using data from nearly two decades ago to be making a flat numerical adjustment to a value that is otherwise calculated relative to other current players. Period. Ever. This is bad. It should always be relative to current players.

You can literally just go paragraph by paragraph through this thing and see that it is a sloppy explanation and attempt to correct for a very real difference in relative value between proficiecny at different positions. It's crap math. Crap science. I'm not going ot pretend I have the mathematical acumen to do better. I don't. But just because I can't bake a cake doesn't mean I can't tell you the one some else bakes tastes like shit.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: Oh My God it's Finally Happening!

Post by Ted » Sat Aug 11, 2018 3:42 pm

RonCo wrote:
Sat Aug 11, 2018 3:29 pm
Ted wrote:
Sat Aug 11, 2018 3:10 pm
A league average CF adds no more wins than a league average 1B. This is wrong. If they are both average, they have the same impact as each other.
The average shooting guard will score more points than the average center. If you're trying to project this counting stat using variance from average, you'd best make a baseline adjustment for position.
Also, the argument that an average CF is more hard to come by than an average 1B is ridiculous. That's not possible. He's literally defined as the average.
The selection process takes care of this. The average nurse is not "as good" at being a doctor as the average doctor. But when pressed into service a nurse will often save your life.

Regarding your first reply, that's fine, but the "baseline adjustment" cannot be a flat number. The average center used to score way more points than the average shooting guard. Things change. Don''t use a flat number. It's wrong. The adjustment will have to both measure and scale for the changing relative value of the proficiency of what it is adjusting for.


Regarding your second comment. I have no idea what your point is. The average nurse is an average nurse. The average doctor is an average doctor. Both have roles. Sometimes, one is more important than the other. The impact of a better or worse one might be more or less significant than the other at different times.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19815
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1982 times
Been thanked: 2902 times

Re: Oh My God it's Finally Happening!

Post by RonCo » Sat Aug 11, 2018 3:51 pm

Ted wrote:
Sat Aug 11, 2018 3:34 pm
Ron, I'm not arguing that their shouldn't be positional adjustments. I'm arguing that the method is BAD. Pujols does NOT add value by playing in the field. This is not true. He is WORSE than an average 1B. He can't freaking move anymore. Any metric that gives him more value simply for having a certain position penciled next to his name is needs improvement.
Don't confuse "Average" with "replacement."

It is 100% possible Albert Pujols can be below average and still provide value above replacement (which is positive WAR).
I've read that fangraphs link you posted many many times are their are so many glaring issues with the explanation that it even ADMITS.

First, there's this

"In general, we want to add runs for players who play tough positions and subtract runs for players who play easier positions to account for the fact that average at one does not equal average at the other in terms of total run prevention. For example, for a shortstop, the adjustment is +7.5 runs per full season. For a left fielder, it’s -7.5 runs per full season."

WHAT????? Read that again. Average at one positions HAS to equal average at another position in terms of relative run prevention. It has to. It's mathematically impossible for it to not. In terms of ABSOLUTE run prevention, of course it doesn't, which is what that paragraph says of course. But why should I care around total or absolute run prevention when I'm trying to use a metric to compare a player to a theoretical replacement level in terms of RELATIVE value?
You've replaced the word "absolute" in the original discussion with "relative" in yours. You have to understand absolute baselines to determine WAR because WAR is based on Replacement, not Average.

If you want to ignore the adjustment, use WAA (wins above average).
It goes on "That means the difference between an average left fielder and an average shortstop is about 15 runs per season. That’s a sizable gap."

Now we're just defining ourselves in terms of ourselves. That's just bad science.
I have no idea what you mean here.
Then there's this

"The actual numbers we use are based on some calculations that were done about a decade ago that used the performance of players who moved positions. It’s certainly reasonable to suggest that those numbers have changed as the game has changed, so use the adjustments as guides more than as firm rules. The DH adjustment might be too negative because it’s harder to hit when you’re not playing in the field, and the catcher adjustment might be a bit too large. There’s lots of room to disagree on the precise decimals and if you’re so inclined, I’d invite you to come up with a more accurate rendering of the numbers. Here is one recent example of such work."

The players used were primarily utility players who do NOT represent typical players. And here the author basically admits that the adjustments are wonky. And THIS is where my biggest problem with them lies. Baseball changes. We should not be using data from nearly two decades ago to be making a flat numerical adjustment to a value that is otherwise calculated relative to other current players. Period. Ever. This is bad. It should always be relative to current players.
The author admits not that the adjustments are wonky, but that the adjustments can be argued over, and should likely be changed as better data is developed. Seems like a reasonable position to me.
You can literally just go paragraph by paragraph through this thing and see that it is a sloppy explanation and attempt to correct for a very real difference in relative value between proficiecny at different positions. It's crap math. Crap science. I'm not going ot pretend I have the mathematical acumen to do better. I don't. But just because I can't bake a cake doesn't mean I can't tell you the one some else bakes tastes like shit.
The issue seems to me that you're looking for something absolute out of WAR that it can't give you. That's fine. But the concept of WAR is pretty starighforward and it mathematically consistent when you realize Replacement does not equal Average, and that lots of the measurements are based off Average, so must be adjusted for. The point here is that replacement level performance can be said to change year over year...another element of the conversation that makes people go frizzy. It's all good.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

scottsdale_joe
Ex-GM
Posts: 3407
Joined: Tue Jun 29, 2010 12:55 pm
Location: scottsdale, az
Has thanked: 68 times
Been thanked: 121 times

Re: Oh My God it's Finally Happening!

Post by scottsdale_joe » Sat Aug 11, 2018 3:55 pm

I can't read anymore. I sorta get what both of you are saying and I suspect you are both partially correct.
For sure I know no one will "win" this debate.

[edit}
uhm:
someone edited my entry. the following line is not from me.
:headscratch:
odd
{edit}

Thanks for following along as well as you did. Really it's not about winning. I learned something through debate. Good enough for me.
Last edited by scottsdale_joe on Sat Aug 11, 2018 6:40 pm, edited 1 time in total.
Joe - GM UMEBA CAIRO PHARAOHS (2047-xxxx); Vancouver Mounties (1996-2009; 2035-2036); Halifax Hawks (2023-2026)Image LINKS:ImageImageImageImageImage

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19815
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1982 times
Been thanked: 2902 times

Re: Oh My God it's Finally Happening!

Post by RonCo » Sat Aug 11, 2018 3:59 pm

I often find it helpful to say that a team of 0 WAR players will create something in the range of 45 absolute wins. That's replacement level (about the least number of games a team that's trying to field a real team will win). If that's true, a replacement level player will create 1.8 absolute wins (45/25).

If we say a team of league average players would win 81 games (which may or may not be true), than the win value of an average player, taking all positions into account (so, being wrong!) is 1.44 WAR (36/25), or 3.24 absolute wins.

That's very rugged, pretend estimation--and it needs to be adjusted for positional skillsets in order to be useful in making roster decisions, but that's the framework for using WAR as something valuable (IMHO).
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: Oh My God it's Finally Happening!

Post by Ted » Sat Aug 11, 2018 4:07 pm

What's interesting is that you may have gotten to the crux of my problem. I just realized I don't care about WAR much at all. Your point about me asking it to do something it isn't designed to do now makes sense. I know you've said this ot me before, but it never clicked until we started really getting at it here. I've forgotten the name this whole time. It's measuring against a flat theoretical baseline, not a moving actual average. Yup, don't care about that. There's so much room for moving error in there that's it's not that useful.

I would make the next argument. The only reason for a metric like WAR to exist is to be used to measure performance. That's what sabermetrics is all about. And I would much rather have one that measures performance against average than some arbitrary point that may or may not be accurate. IN a way, this whole thing comes down to looking at relative performance versus absolute. Most of the time, you get way more bang out of relative measurements.

Also, I think I know now why WAR, particularly and it's positional adjustments get me irritated so much. It's kind of a bastard abomination of relative measurements AND absolute measurements crammed together in one noisy metric.

Anyway, long boring arguments aside, I'm just thrilled someone's taking a hard look at how we have come to view the defensive spectrum, and the way it is potentially affecting the way we value players. Even if wrong, maybe this will snowball us to a better version.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19815
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1982 times
Been thanked: 2902 times

Re: Oh My God it's Finally Happening!

Post by RonCo » Sat Aug 11, 2018 4:22 pm

Ted wrote:
Sat Aug 11, 2018 3:42 pm

Regarding your first reply, that's fine, but the "baseline adjustment" cannot be a flat number. The average center used to score way more points than the average shooting guard. Things change. Don''t use a flat number. It's wrong. The adjustment will have to both measure and scale for the changing relative value of the proficiency of what it is adjusting for.
I think there's a good likelihood that the flat number adjustments should be altered by era.
Regarding your second comment. I have no idea what your point is. The average nurse is an average nurse. The average doctor is an average doctor. Both have roles. Sometimes, one is more important than the other. The impact of a better or worse one might be more or less significant than the other at different times.
My point is that to come up with an absolute value (which is what WAR is), you can't just take relative values of the quality of one part of the process and add them together. I mean, sure, you can do it mathematically, but it won't get you what you care about because a surgeon's role in an operation is considerably more valuable during an operation than a nurse's. If, for example, you have to do without the doctor or without the nurse, you take your chances with just the doctor--because the doctor's base value in the success of the procedure is bigger than the nurses.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

Ted
Ex-GM
Posts: 5630
Joined: Thu Jun 04, 2015 2:50 pm
Has thanked: 368 times
Been thanked: 378 times

Re: Oh My God it's Finally Happening!

Post by Ted » Sat Aug 11, 2018 4:40 pm

Your comment about the absolute value is where I was failing. I kept forgetting that WAR is an absolute value. It just doesn't seem like it should be. And it is so ubiquitously used otherwise, that I (and probably others) keep making that mistake.

The real question I should be asking rather than getting irritated by positional adjustments, is "How good is the correlation between WAR's measurement of a value greater than a theoretically baseline and the relative value of a particular player above average?"

I would love to see this. Of course, we'd then have to figure out to accurately model relative value to average in terms of producing wins, but still this would be neat to see. I'm sure the intent is for WAR to model this reasonably well, but with the changes in the talent pool it may not.
Ted Schmidt
Twin Cities Typing Nightmares(2044-present)
California Crusaders (2021-2038)
Image

User avatar
RonCo
GB: JL Frontier Division Director
Posts: 19815
Joined: Sat Nov 14, 2015 10:48 pm
Has thanked: 1982 times
Been thanked: 2902 times

Re: Oh My God it's Finally Happening!

Post by RonCo » Sat Aug 11, 2018 5:11 pm

I think most WAR arguments are made because of this disagreement or misunderstanding on what it's trying to do. :)

On the whole I think we agree more than anything. The goal is to only use it when it's valuable. And a lot of people who don't think about it--or get lazy like I do sometimes--just use it as a single all-encompassing comparison of all players.

To make a statement I'm pretty sure you'll agree with here, though...
RonCo wrote:
Sat Aug 11, 2018 4:22 pm
I think there's a good likelihood that the flat number adjustments should be altered by era.
For example, a 3B may well be "more difficult" to play in an era where batters bunt a lot more. Catcher's ability to control the running game may be of bigger value in the speedy days of the 80s. Infield defense may be less valuable to absolute wins in today's "Three True Outcomes" viewpoints. Blah, blah, blah.

In each of these cases, the right answer for WAR would be to have era-specifc adjustments.
GM: Bikini Krill
Nothing Matters But the Pacific Pennant
Roster

agrudez
Ex-GM
Posts: 7681
Joined: Tue Nov 20, 2012 10:30 am
Has thanked: 21 times
Been thanked: 47 times

Re: Oh My God it's Finally Happening!

Post by agrudez » Mon Aug 13, 2018 9:25 am

I think we've had a dozen back and forths over WAR in the past so I won't add to the white noise here with repetition. I am very satisfied with how I use WAR and will continue to use it in that way. If they removed or changed positional adjustment factors it would have 0 bearing on how I use WAR. I am glad that your passion project is getting some wider exposure, though - that probable feels cool.
League Director: Kyle “agrudez” Stever*
*Also serves as chief muckraker
-Ron, 2025 media guide

Image

Post Reply Previous topicNext topic

Return to “Off Topic Chatter”

Who is online

Users browsing this forum: No registered users and 11 guests