No more WAR!

[Moose Milligan]

http://espn.go.com/mlb/story/_/id/8900693/call-moderation-use-wins-replacement-stat

[weams]

http://espn.go.com/mlb/story/_/id/8900693/call-moderation-use-wins-replacement-stat

If we can't figure a stat out on our own, then how do we verify whether it is accurate?

"Mike Trout had a 10.7 WAR and Miguel Cabrera had a 6.9 WAR, so anyone who thinks Cabrera deserved to be the American League MVP should be strip-searched, tied to an anthill and forced to rely on dial-up for his Internet connection for the remainder of his pathetic life."

[OFFNY]

No more WAR!

In 2013, there will be no more WAR, and we'll give Pearce a chance.

[skanar]

Writer has two main complaints about WAR, both reasonably valid:

1) It's tough to calculate because it relies on a large number of components.

2) It relies on defensive statistics that are unreliable.

Most of the article is in relatively measured tones, unlike the headline.

[The Wedge]

Obligatory "WAR, huh, what is it good for?" post.

[DrungoHazewood]

If we can't figure a stat out on our own, then how do we verify whether it is accurate?

Why can't you do that? In fact, bb-ref and Fangraphs both break it down by components so you can add up the final tally yourself. You can investigate each component to see if you agree or disagree or want to dig deeper. This isn't rocket science. The only part of WAR you can't figure out yourself is the fielding component, because that's the compilation of hundreds of uniquely observed and recorded individual plays. If you want, you can just take everything else and throw in your own subjective measurement.

All WAR does is consistently and in an unbiased manner add up all the stuff the MVP voters (or fans or whomever) do in their heads in a mostly haphazard way.

Everyone uses some form of WAR when making judgments about players. The only debate is how carefully you want to add up the components.

"Mike Trout had a 10.7 WAR and Miguel Cabrera had a 6.9 WAR, so anyone who thinks Cabrera deserved to be the American League MVP should be strip-searched, tied to an anthill and forced to rely on dial-up for his Internet connection for the remainder of his pathetic life."

I would have just stopped at "shook head at, and laughed a little" but everyone is entitled to a little hyperbole for effect.

[DrungoHazewood]

Almost no one knows how to calculate WAR.

Almost no one knows how to calculate ANYTHING. My God, if we only trusted calculations that we could personally figure out on the back of a napkin, we'd be sitting in a corner trembling in fear at almost every aspect of modern society. Everything from stocks to cars to your cell phone to Google Maps to your interest rate on your house... Every day you trust a million calculations that are not anywhere near as transparent as WAR.

This is basically the dumbest argument in the history of the world. (Yes wildcard, that hyperbole was just for you.)

[DrungoHazewood]

I don't know why I still get so spun up over yet another garden variety random sportswriter spewing strawmen and outliers to prove that some metric is not 100% accurate in measuring the past and predicting the future. Haters will hate, people need to fill column inches, and old newspapermen don't usually change their stripes. But life will move forward without them.

[The Wedge]

I don't know why I still get so spun up over yet another garden variety random sportswriter spewing strawmen and outliers to prove that some metric is not 100% accurate in measuring the past and predicting the future. Haters will hate, people need to fill column inches, and old newspapermen don't usually change their stripes. But life will move forward without them.

Probably because the tyranny of willfully ignorant men will glom on to those articles like the gospel.

[Nigel Tufnel]

Almost no one knows how to calculate ANYTHING.

Yeah, seriously. I would bet that a huge percentage of baseball fans doesn't know how to calculate On Base Percentage. Would Caple use that to mean OBP isn't a relevant stat?

[crissfan172]

There are only two things about WAR that really bug me:

1. I hate that it's calculated based on the position you play. If a SS and a 1B put up the same numbers, their oWAR should be the same. If you want to add dWAR for defensive difficulty, I get that, but it just bothers me when we say that such and such players is a better offensive player, when what really should be said is that the difference between him and the average player at his position is greater than the difference between the other player and the average player at his position. I understand that this is what it's trying to measure, but it just seems like many people who use it don't understand that, and it just drives me crazy.

2. Using it is just lazy. I don't need to see WAR numbers to know that Trout had a better year than Cabrera. Both had similar OBP, while Cabrera's ISO was somewhat higher. Add in speed, plus much superior defense, and it should be obvious that Trout had the better year. I guess I just don't like the shorthand, when it's not that difficult to analyze the numbers on our own.

I suppose my complaints aren't with the statistic, so much as they're with how people use them.

[Loophole]

1. I hate that it's calculated based on the position you play. If a SS and a 1B put up the same numbers, their oWAR should be the same. If you want to add dWAR for defensive difficulty, I get that, but it just bothers me when we say that such and such players is a better offensive player, when what really should be said is that the difference between him and the average player at his position is greater than the difference between the other player and the average player at his position. I understand that this is what it's trying to measure, but it just seems like many people who use it don't understand that, and it just drives me crazy.

This seems like a weird preference to have. For one thing, with Fangraphs' system at least, it doesn't work like that, and I like fWAR because everything is spelled out very plainly at the bottom of every player page. In 2011, JJ Hardy and Edwin Encarnacion both had a .344 wOBA, which isn't position-adjusted, in a similar # of PAs. They both contributed 10.9 batting runs to their WAR despite playing completely different positions. Positional adjustment is an entirely different thing. Now, it looks like oWAR is adjusted which really just means they combined the batting runs and positional adjustment to make things quicker. But when you're using the term "wins above replacement" you're specifically referring to the number of wins over a replacement player at that position. oWAR doesn't take defense into context, but you can't rightly call it "WAR" if it isn't adjusted for position. Thing is, how often do people specifically refer to oWAR anyway? And if you're comparing hitting ability, are you really gonna say that Cal Ripken (.346 wOBA) was worse offensively than Don Baylor (.347)? I'm gonna go ahead on call Cal the better hitter because position context does matter.

TL;DR: position, batting and defense are three completely different factors in WAR, oWAR just clumps two of them together because position matters in WAR.

2. Using it is just lazy. I don't need to see WAR numbers to know that Trout had a better year than Cabrera. Both had similar OBP, while Cabrera's ISO was somewhat higher. Add in speed, plus much superior defense, and it should be obvious that Trout had the better year. I guess I just don't like the shorthand, when it's not that difficult to analyze the numbers on our own.

This, I actually agree with. For some people, the fact that WAR is all-inclusive makes it a lot harder for them to actually pay attention to all the specific factors it's including, and their analysis becomes short, boring and too easy.

[The Epic]

A lot of people (and maybe this includes me) don't necessarily use WAR as the be-all-end-all, but as a support for the obvious. I don't think that people say that Player X trumps Player Y because their WAR is 0.3 better. But if they are -so- much better (like with Cabrera/Trout), then maybe we're missing something here.

[ledzepp8]

Almost no one knows how to calculate ANYTHING. My God, if we only trusted calculations that we could personally figure out on the back of a napkin, we'd be sitting in a corner trembling in fear at almost every aspect of modern society. Everything from stocks to cars to your cell phone to Google Maps to your interest rate on your house... Every day you trust a million calculations that are not anywhere near as transparent as WAR.

This is basically the dumbest argument in the history of the world. (Yes wildcard, that hyperbole was just for you.)

The worst thing about his ranting on not knowing how to calculate WAR is that he works for ESPN. This is a company that is still trying to cram Total QB Rating down our throats even though they won't release info on the formula that derives that ****ty stat.

[DrungoHazewood]

This, I actually agree with. For some people, the fact that WAR is all-inclusive makes it a lot harder for them to actually pay attention to all the specific factors it's including, and their analysis becomes short, boring and too easy.

I'll agree that the practitioners of tWAR (Trea-WAR) are little better than the folks who rely on pitcher wins and saves and batting average. There's occasionally someone who will argue that Team X will win 88 games this year because they added up their favorite version of WAR for everyone on the roster, subtracted the negative-WAR guys, and *poof* there's your projection for next year. Or take their favorite first baseman, use their career-peak-age-27 WAR as their baseline for the next five years, and obviously any dummy would know he's worth a 5/200 contract.

You can't fix stupid. It's going to happen with any argument. Saying that's a reason to not use WAR is also an indictment of all analysis.