Is WAR (baseball) utter nonsense?

[Milestones]

I have never seen a baseball stat so completely silly and so completely fabricated as WAR.  I am sick of seeing it come up in baseball articles--for example, an article today arguing that 5 great players don't belong in Cooperstown--principally because of WAR.  I will never be convinced that there is any logic to how WAR is calculated or that it has a shred of validity.   

Comments?

[Scott Dolan]

I think most "next gen" stats are laughable. WAR being the leader in the clubhouse. 

[GA Russell]

Milestones, who were the five players?

[JSngry]

It's a metric. If the input is accurate, then you have an accurate metric. The question is not does it have merit in terms of accuracy, the question is what does it really mean/measure, and I think people are still figuring that out. "Value Added" is a real thing, and WAR is one way that people are trying to more specifically quantify it. I'm still a fan of "intangibles", but these days, everybody wants "proof", and to have proof, you gotta have metrics. Or so it seems. And so the hunt begins.

As far as WAR itself goes, I do think it means something in terms of probable long-term performance. The "problem" (such as it is) is that what people remember are situation-specific outcomes, not overall general performance. Some generally lackluster guy can have a set of "memorable moments" and hey, Beloved Figure Forever, perception becomes reality becomes legend becomes mythology. This is life.

But that's the way shit goes, right? You can work your ass of at your job every day and some scrub pops up with a little bit of ability, a little more luck, and a whole lot of fortunate time/space placement, BAM, off they go.

Still, if you're a Sabermetric Sensitive GM, and if you have a Sabermetric Sensitive manager/organization, I think you can look at WAR along with a lot of other metrics and look at is Player X likely to give us more shots more often than Player Y, and if so, how, and then proceed accordingly, realizing - one would hope -  that it really is a game that truly exists only in the moment. "Gut instinct", I still like that, but an educated gut is probably going to have better instinct over the long haul. The problem is that people want immediate gratification, one championship means more than decades of non-championship excellency. I don't buy into that in the least, but that's  just me.

[BFrank]

Utter nonsense or not (stat-wise), people in baseball are paying a lot of attention to it. THAT makes it important.

[felser]

Part of the problem is that there are still a lot of difficulties  in measuring fielding performance, though they've come a long way.   I place credence in WAR (though it is far from perfect, of course), also believe there are definitely players in HOF who shouldn't be (and ones not there who should be).  Context considerations are very important yet very difficult for baseball stats.  

 

[JSngry]

I look at it as a work in progress, really. The fact that there's different methodologies is telling, but so is the notion that they're all trying to evaluate the "worth" of a player in a more objective (or at least, less subjective/emotional) manner.

What remains to be seen is if anybody really wants that, or if indeed it will ever be possible to be totally objective about anything involving "value". But even if eliminating the subjective proves to be neither possible nor desirable, I don't think you face eternal death by fine-tuning perceptions about what "is" really is. Just, you know...allow that moving forward sometimes involves getting it wrong and getting it right. Like the man said, you'll know when you get there, and really, do you ever?

1 hour ago, felser said:

... also believe there are definitely players in HOF who shouldn't be (and ones not there who should be).  Context considerations are very important yet very difficult for baseball stats.  

 

yeah, if Phil Rizzuto had spent his career with the St. Louis Browns, would we today even know who he was?

Of course, he didn't and of course we do. But...

Perceptions is a tricky little imp, that it are.

[Hardbopjazz]

WAR, what is it good for?

Absolutely nothing.

Say it again

 

I couldn't resist. 

[JSngry]

1 hour ago, Hardbopjazz said:

WAR, what is it good for?

Absolutely nothing.

Say it again

 

I couldn't resist. 

That's more true when you're 25 miles from home than when you're just 90 feet.

[Jim R]

3 hours ago, JSngry said:

BAM

bunt attempts missed?

batter attacks mound?

balks after midnight?

Hey, you never know these days...

[felser]

2 hours ago, JSngry said:

That's more true when you're 25 miles from home than when you're just 90 feet.

I'm just known as Agent Double O Stat.

[GA Russell]

I'm going to start a new thread about Cooperstown.

[Scott Dolan]

5 hours ago, BFrank said:

Utter nonsense or not (stat-wise), people in baseball are paying a lot of attention to it. THAT makes it important.

And then there's that...

There isn't a team in MLB that doesn't have a dedicated sabermetrician in their front office. 

My only question remains: if Moneyball is all it's cracked up to be, why haven't the A's won a World Series since it became popular? 

[felser]

1 hour ago, Scott Dolan said:

And then there's that...

There isn't a team in MLB that doesn't have a dedicated sabermetrician in their front office. 

My only question remains: if Moneyball is all it's cracked up to be, why haven't the A's won a World Series since it became popular? 

In the early 2000's, just bad luck.  They had great records, just didn't have the luck in the short series.  Run a binomial distribution, and you will see that there is a huge chance of the best team not coming out on top in a short series.  Make it two or three short series in a row, and the odds are stacked against any team, no matter how good.   That is the nature of baseball, where the greatest teams of all-time lose 1/3 of their games.  Different than other  major sports.   Since then, other teams have caught up, and have more money to invest wisely.  You think the Yankees and the Red Sox weren't paying attention?  Boston hired Bill James himself.  And in recent years, Billy Beane has gone a little wacko, going against his own credos in things like the  Addison Russell trade and the Josh Donaldson trade (and just now the  Rajai Davis signing).  BTW, if we knew then what we know now,  we would realize that the incredible individual season of 1961 wasn't Roger Maris nearly as much as Mickey Mantle, Norm Cash, and Jim Gentile.

[JSngry]

4 hours ago, Jim R said:

bunt attempts missed?

batter attacks mound?

balks after midnight?

Hey, you never know these days...

BOOM!

[Scott Dolan]

2 hours ago, felser said:

In the early 2000's, just bad luck.  They had great records, just didn't have the luck in the short series.  Run a binomial distribution, and you will see that there is a huge chance of the best team not coming out on top in a short series.  Make it two or three short series in a row, and the odds are stacked against any team, no matter how good.   That is the nature of baseball, where the greatest teams of all-time lose 1/3 of their games.  Different than other  major sports.   Since then, other teams have caught up, and have more money to invest wisely.  You think the Yankees and the Red Sox weren't paying attention?  Boston hired Bill James himself.  And in recent years, Billy Beane has gone a little wacko, going against his own credos in things like the  Addison Russell trade and the Josh Donaldson trade (and just now the  Rajai Davis signing).  BTW, if we knew then what we know now,  we would realize that the incredible individual season of 1961 wasn't Roger Maris nearly as much as Mickey Mantle, Norm Cash, and Jim Gentile.

Yeah, but the best teams win more three game series than they lose.

So that doesn't count when the series are 5 and 7 games? 

I don't buy into the best team rarely wins it all thing. 

[Milestones]

For one thing, I don't get who the replacement is.  Is it just some generic schmuck who can't hold a candle to, say, Mike Trout?

 

[GA Russell]

1 hour ago, Milestones said:

For one thing, I don't get who the replacement is.  Is it just some generic schmuck who can't hold a candle to, say, Mike Trout?

 

That's my understanding of it.  They're talking about a Mendoza Line player.

[JSngry]

https://en.m.wikipedia.org/wiki/Value_over_replacement_player

[felser]

3 hours ago, GA Russell said:

That's my understanding of it.  They're talking about a Mendoza Line player.

They are talking about a guy you could go get from an AAA team and plug in.

[Milestones]

To my mind, the 15 or so paragraphs of explanation (plus some equations almost no one can understand) seals the absurdity.  

 

[Scott Dolan]

Agreed. Give me average stats. That's all that's necessary. 

[JSngry]

That works when the talent level of an "average" player stays the same relative to the entire talent pool every year/decade/whatever, which is does not. Skill sets evolve, talent pools strengthen and weaken, somebody who could kick ass today could take a time capsule either forward or backward and find themselves either even more or even less kickassable.

WAR is probably not a good "fan stat" at this point, and may never be. But it is not "utter nonsense". It's an imperfect - and hopefully still evolving - performance metric of an individual relative to a general population. Most stats just measure the individual alone, and then the results are compared. With WAR, the general population is a factor at the beginning of the calculation. That's different, and that has the potential to be quite meaningful.

Consider driving records. If you're not all that great a driver but drive in a low impact traffic area, your driving record could make you look like the best and safest driver ever. But take your driving habits in relation to all traffic areas, not just yours, and maybe you're a bad driver after all, maybe it's your environment that's kept you out of the hospital, maybe your driving is really not that good at all. I've not had experience with these little driving monitors the insurance companies will try to give you to see what your driving behavior is and how "safe" a driver you are, nor do I know what their baseline is for "safe driving" when they look at your driving data. But I could easily see scenarios where, once enough data, refined data, is collected nationally, a national minimum of "replacement level driver" is established. How sharp are your turns? What is your average stop time? How soon do you signal before turning? What is the average distance you leave between cars? Etc.

Of course, driving is not baseball, but the notion of comparing an individual's performance relative to a collective performance instead of a set of individual outcomes is not without meaning, even if the methodology is far from perfected.

And personally, I very much like that something as "rigid" as metrics can be so fluid in interpretation. Metrics really don't mean a whole hell of a lot with context, and the deeper and broader the context, the more - or less - meaningful is the metric. As an evolutionary conceit, I think it's inevitable - and hopefully wonderful - that we are essentially attempting to expand our consciousness with the tools at hand. Of such things are progress made, except when they aren't, but that's what happens when you've yet to determine a Global Replacement Level Reality...so...strap it on and Science Up, World!

 

[Scott Dolan]

I meant average stats as in numbers derived from averages. 

[JSngry]

One thing I think that WAR is addressing is the pursuit of a more accurate performance metric in the face of the vanishing balanced schedule. Now more than ever, performance needs to be evaluated from a global rather an individual context. Used to be that if a guy played 10 years in the American league and had a .300 average, that gave you a good idea of his ability in that specific context, and that could then be used to ascribe value (although based wholly on past performance). Now, the same guy over 10 years, he might still hit .300 for those years, but .300 relative to what? Sure, relative to all players, but not all .300s are created equally these days, sometimes a .295 hitter will serve you better under certain conditions than a .300 hitter, and that's stats like WAR is trying to get at, a portrait of more objective, real-time/fluid global look at performance rather than just looking at the back of a baseball card and then negotiating a contract.

And yeah, make no mistake, data is power, so don't think that management and labor both need to stay on top of this lest one get the upper hand over the other for no real good reason. Nothing wrong with calling bullshit where bullshit can be rightfully called, and nothing wrong with standing firm on firm ground. Too much money not to.