Bob Ryan on baseball's WAR

I'm just looking for something that explains fielding better than what we have now. OPS I can live with and is easy to understand.

 

Basically, you're saying that ESPN and MLB.com and CBS Sportsline should use it more frequently, right?

 

I know. And it's a noble goal. But I don't think "gloved balls" accomplishes it because, like I said, it captures too much noise.

 

Spoken like someone who is happy to abandon all logic in a desperate attempt to justify statistics that don't measure what they claim to measure and rely fair too heavily on subjectivity.

If you can't understand why a baseline being consistent rather than constantly changing is a huge difference, you are truly beyond help.

This is exactly why people like you fail to make a compelling argument for these statistics. You refuse to acknowledge the flaws and you refuse to even try to do it without unjustly attacking the intelligence of those who question them.

 

Gee said it on Page 1, but I think fielding is going to prove to be mostly not measurable. Teams with full scouting departments and huge budgets will be able to evaluate things according to their specific needs and fits and acquire players that way, but a catch-all number that measures everyone in the league against each other for the now-completed season will be too tough. It's like the misguided attempts to apply the approach to football, where there's too much else happening to isolate an individual's performance.

 

I can't believe I'm letting myself get sucked into this conversation, but yet again:

WAR is not a statistic, and there certainly isn't only one of it. It is an idea, a way to design a scale.

Saying "WAR showed X" doesn't make any sense. One metric, trying to measure player value and using the WAR scale, had a defensive component that wasn't accounting for shifts in any way and caused Brett Lawrie to be severely overvalued.

 

It doesn't constantly change. It's the same within every position. It would be flawed if it did change. Someone suggested this. Someone suggested that it should change for each team.

I certainly think WAR is probably a flawed statistic, mostly because of what you said: It doesn't measure what it purports to measure. But it's not flawed because it has a base line. The base line doesn't matter. It affects accuracy, not precision. Name it EWAR (Estimated Wins Above Replacement) and we're solid.

 

I understand that, Rick.

But WAR was used almost exclusively to mount the case for Trout as MVP. Not by you, but it is absolutely correct to say that's the kind of influence WAR has in a certain very large and loud segment of the online public.

 

I just can't get behind temperature. They can't even make their minds up about it, and it's totally arbitrary. Those Fahrenheit guys say that zero degrees is really damn cold, but the Celcius guys say that zero is sort of cold but tolerable. They can't even make up their minds about that, and we're supposed to believe that they have a good system for measuring the atomic activity of objects? Bullshit.

I even heard that there was this one temperature scale that said that zero was so cold it couldn't possibly exist in the universe. What more proof do we need that these people aren't living in the real world?

 

This post comes like one page after I explained why I like Win Value Added better than Wins Above Replacement.

 

Cool beans. Like I said, I'm tired of WAR. I wash my hands of it. We can burn it at the stake or whatever. Even if we find it necessary to use total value stats, we can just as easily use "Bases above average" or "Runs above zero" or "Outs below perfection" or some other unit/index combination.

It's really dawned on me in the last few years how interesting baseball is and how pointless these arguments are. "Stats vs. scouts" is really just "I typed his name into Fangraphs" vs. "I have a gut feeling that I want to justify and saw several baseball games."

Baseball is really fucking complicated. The people who really understand it are the ones who spend hours watching it live, then more hours breaking down tape, then more hours pouring over meticulously prepared statistical reports. Every day for years and years on end.

The rest of us are just wanking ourselves if we think we really understand it. I don't need to understand it. I enjoy it.

 

After trying that ridiculous argument that replacement works as well as a baseline as zero.