You are viewing the chat in desktop mode. Click here to switch to mobile view.
X
ALDS Games 4 Chat
powered byJotCast
Prev Page View Full Chat Next Page
Ben Clemens
5:56
I'm not sure what happened with Minor but something like that maybe?
I have a mildly hot take on WAR that we don't use it correctly
I don't think WAR necessarily measures who did the most to help their team win that year well
Particularly for pitchers
But hitters too
5:57
I think WAR is often better (with modifications) for forecasting
Paul Sporer
5:57
Interesting, I've tended to see it more useful for looking back at the season as a whole and not so much for forecasting, particularly w/hitters
Ben Clemens
5:58
I guess more what I mean is, the player who did the most to help their team win
Is not always, or even often, the player with the highest WAR
Paul Sporer
5:58
I'd say our pitching WAR is probably better for forecasting since it's FIP-based, but I think if ppl are gonna use it for awards, it should focus more on what happened, not what should've happened
Sam
5:58
What would that modification be out of curiosity? Or just a way of understanding it?
Ben Clemens
5:59
I think you need to regress defense more heavily than even we do (UZR is already regressed), and care a little bit less about batted ball outcomes
from the hitting side
That's essentially what projection systems do, in general
Paul Sporer
5:59
I'd agree to that more if it were a forecasting tool
6:00
But as an assessment of the season, why wouldn't you want to focus on what happened?
Ben Clemens
6:00
Right, but then what about a guy who hit a ton of dingers in games they were losing by 10
That's super helpful for forecasting his future
Because we know that outcomes are distributed somewhat randomly across PAs
Paul Sporer
6:00
Well that goes back to never using WAR or any stat as a catch-all
Ben Clemens
6:00
Yeah exactly
I think people over-use it in MVP debates and stuff
Paul Sporer
6:01
I think it's a great starting point to pare down the pool of candidates
Rarely if ever do I think someone who is like 3-4 WAR behind a league leader is deserving of the MVP
Ben Clemens
6:01
I agree with that
6:02
But 1-2? sure
Paul Sporer
6:02
But when people want to act like there's precision betweent 6.7 and 5.8, they lose me
The 5.8 guy could be JUST as deserving IMO
Ben Clemens
6:02
This might make me a little 'old man yells at cloud' but I kind of bought Bill James's argument that if a team wins 81 games but has the WAR of a 95-win team
You should discount the WAR on that team for the purposes of award voting
Because clearly players weren't THAT valuable
6:03
if by value you mean winning games
Paul Sporer
6:03
Does that happen with any regularity? (at all?)(
Ben Clemens
6:03
No but it does sometimes
I think his was about Judge vs. Altuve that year
Sam
6:03
Interesting, that makes sense. WAR is context neutral which would make it more useful as a predictor then?
Ben Clemens
6:03
Yeah
6:05
Houston threatening again
2 on 2 out
Connecting…