Interesting article at Prospectus about different WAR systems.

[dougdirt]

Originally Posted by lidspinner

I agree but one thing almost all scouts have in common is thst they all use stats as a heartbeat to their system.....he is one of the few who don't. So many of the advanced metric stats are the talk of the scout community yet a few of the old farts refuse to use them.....I am not saying that's good or bad cause I fall somewhere in the middle. I think you need to use the stats but you need to trust your eyes more than anything. From what I have been told the Cubs new brass is not letting ANYTHING go uncovered.

It certainly depends on the set of eyes.

[RedsManRick]

Originally Posted by lidspinner

I agree but one thing almost all scouts have in common is thst they all use stats as a heartbeat to their system.....he is one of the few who don't. So many of the advanced metric stats are the talk of the scout community yet a few of the old farts refuse to use them.....I am not saying that's good or bad cause I fall somewhere in the middle. I think you need to use the stats but you need to trust your eyes more than anything. From what I have been told the Cubs new brass is not letting ANYTHING go uncovered.

Stats and scouts don't really measure the same thing.

Stats count what a player has done and can place a run value on it. In general they help us better understand how much the various on-field events contribute to overall run scoring and prevention.

A scout assesses a player's ability to play, current and future.

Sure, stats be useful as a proxy for scouting. And scouting be used to estimate future stats. But are their core they aren't doing the same thing.

What I'd love to see more of in the public domain is attempts to convert true scouting information into statistical projections. That is, I don't want a scout telling me a guy is a .280/20-30 HR type of guy. I want that raw information about what he's seeing. Then I want to see a body of evidence that says "guys who have X swing plane, who have good instincts, who don't jump at good breaking pitches and who get their hands through on inside fastballs or simply have "good makeup", as judged by a scout, end up putting up X, Y, Z kind of performances.

A large part of my day job is program evaluation for non-profit organizations. We used mixed methods approaches and there's no way we could do our job well without a mix of quantitative and qualitative information. But they aren't simply different ways of measuring the same thing. That would be silly. They are ways measure different aspects of things in the most effective way.

And while the use of the quantitative and qualitative vary depending the question and the balance shift over time based on what we already know, they never fully supplant the other because at the end of the day, they both have domains where they remain the most accurate and effective way of capturing information.

All that said, I'm curious. The question WAR attempts to answer is this: Of a team's wins, how many wins can be attributed to X player's on-field production. That's it. We might try to use (and find it useful) for other things, but for that question, I'm not sure scouting adds anything other than perhaps some detail on why the player was able to do what he did (of failed to do).