Statcast and the future of WAR

[thewupk]

http://www.fangraphs.com/blogs/statcast-and-the-future-of-war/

Nice article by Dave Cameron yesterday.

Even with us getting to a point where we would know that one particular play is a hit 95% of the time but Player A gets robbed that we don't credit him even with him for WAR purposes doing what would be considered everything he can do to get a hit but having nothing to show for it.

WAR should be refined to what actually happened on the field regardless of the probability of it. However such information I think could lead to more stats like xOPS (expected OPS for example) that could take such information into account which would be better suited to try and predict future performance. Would really show us which players are actually getting super lucky via BABIP. For example I would feel something like this would show Ender was a lot more lucky than Freeman in the 2nd half last year despite both having super high BABIP's.

 

[Enscheff]

As long as they have a descriptive WAR and a predictive WAR, I am all for this and think it's great.

The Ethier HR referenced in the article was a HR, and it counted for 1 run whether the wind was blowing or not. That HR was not predictive of Ethier's future production vs LHers though, so I would like to see some stats developed that make that distinction.

I like your idea for xOPS, or xWAR...or whatever.

 

[smootness]

Seems like you could just have a WAR value based on what truly did happen on the field (meaning you also contextualize events in terms of what they actually produced, which is not done now) and a separate, more predictive WAR value based on what all available batted ball and context-neutral data tells you that players should have been worth.

 

[smootness]

As long as they have a descriptive WAR and a predictive WAR, I am all for this and think it's great.

The Ethier HR referenced in the article was a HR, and it counted for 1 run whether the wind was blowing or not. That HR was not predictive of Ethier's future production vs LHers though, so I would like to see some stats developed that make that distinction.

I like your idea for xOPS, or xWAR...or whatever.

Yes, this.

 

[thewupk]

Seems like you could just have a WAR value based on what truly did happen on the field (meaning you also contextualize events in terms of what they actually produced, which is not done now) and a separate, more predictive WAR value based on what all available batted ball and context-neutral data tells you that players should have been worth.

I think having both versions would be ideal. And whether MLBAM and Statcast want to do this I am not sure. But if they make the data publicly available then there would be enough crowd sourcing to get it done. There are a lot of nerds out there.

 

[Enscheff]

Descriptive WAR would be used for things like MVP discussions and arbitration hearings. Predictive WAR would be used to analyze surplus values when it comes to trades and FA signings.

Good luck trying to get folks that can't understand what the current incarnation of WAR represents to understand the distinction though. Sadly, it will most likely just widen the knowledge gap between dummies and smarties.

 

[jimsnores]

Descriptive WAR would be used for things like MVP discussions and arbitration hearings. Predictive WAR would be used to analyze surplus values when it comes to trades and FA signings.

Good luck trying to get folks that can't understand what the current incarnation of WAR represents to understand the distinction though. Sadly, it will most likely just widen the knowledge gap between dummies and smarties.

Encheff, I agree with your first paragraph. It makes logical sense.

I wish you'd not follow good information like that provided in the first paragraph with comments that downgrade others. This is not intended as an attack on you, but as a plea. You have an analytical mind that often allows you to disseminate thought provoking comments that I consider valuable, but for me, also including comments that put down others (even anonymous others) makes it hardly worth reading the stuff I enjoy. I realize I have the ability to choose to ignore, but I'd much prefer to have thought provoked without having to endure reading the other.

Thanks for your consideration.