THE BOOK--Playing The Percentages In Baseball

It’d be nice to get some demographic information on the fans who fill out your survey regarding what outside information (like the defensive metrics published elsewhere) they have knowledge of.  I would be somewhat surprised if the people filling out the form here don’t have a fair degree of familiarity with some of the more advanced defensive metrics out there.  The problem of distinguishing between people’s actual opinions derived from their personal observation and their internal bias incorporated from other sources of data seems quite difficult.  I suppose I would expect a fairly high degree of correlation among all these simply because of that.

Most people who fill out the ballot are NOT readers of this blog.  They come from SBNation or Fanhome or Primer and whatnot.

As for prior knowledge of metrics, the best test-case would be to look at players with fewer than 100 or 300 or whatever innings played entering 2008.  They are basically a clean slate, in terms of fielding metrics.  Evan Longoria is a case in point.  He had no metrics, and fans think he is the equivalent of Beltre and Rolen.  High praise indeed.  Let’s see what UZR says about him in 2008/09.  Same deal with Fukudome, and Asbdrubal Cabrera.  Fans think these 3 guys are among the best fielders in all of baseball.

I won’t deny that it exists, but I think the design Tango uses minimizes that.  He’s not asking fans if a guy is an above or below average fielder, but specific tools.

I remember in his rookie year Nick Swisher had a ridiculously high UZR rating.  Fans might have read that here or on BTF, and have been biased to giving him a high overall rating.  Opposite for Ichiro, who has not been impressive by UZR.

But the fans were asked to evaluate the running speed, throwing accuracy, etc.  And regardless of who is the statistically better outfielder, I don’t think anyone who watches games is going to tell you that Swisher runs faster than or throws better than Ichiro.

An aside, no comments on my blog about this yet, despite being one of the better (IMO) pieces I’ve done in awhile, at least one that took a long time to bring the data together and invent some methodology.

Yet I mention Moneyball in a simple, analysis free, random coincidence sort of post, and I get a dozen comments from people who still want to debate what Moneyball really meant.  For some reason I find that a little depressing.

FWIW, I read it and didn’t really have anything to add. Agree on DER - it’s basically all we have, but it’s simply not as rich as, say, team run scoring is for offense. It’s harder to validate defensive value systems at the team level than, say, run estimators.

Also FWIW, Rally, I read it this morning and saved it in bloglines for a subsequent link dump (whenever I get around to doing one).  It’s nice work, and would give me an opportunity to plug your essentially one of a kind defensive projections as well.

As for the comments on your moneyball piece, I think you can blame studes and his link at THT for that!
-j

I hear you Rally.  I feel that sometimes as well…

Thanks guys.

I don’t normally pay attention to comments, I don’t do it to be popular, just write what I have to say.  But it makes me wonder, if I just write about Billy Beane and Michael Lewis, that is the way to appeal to a wide audience?

That’s very depressing for an Angel fan.

Thanks for the info Tango.  I hadn’t seen either of those links.

I have also wondered about the problem of survey participants being biased by stats that they have seen.  There is probably some of that but I think the survey design helps to minimize it.  One thing I did for the Tigers was deliberately recruit a good number of people who I know are not into fielding stats but know the game well. They enjoy participating as well. 

Lee