Monday, May 09, 2005

The Nats Blog Book Review

A couple of weeks ago I finished two books that might interest you, and which have some tangential relation to baseball. The first was Freakonomics: A Rogue Economist Explains the Hidden Side of Everything, which is getting a lot of buzz these days, especially on the blogs. It is about Steven Levitt, an young economics professor at the University of Chicago, and his wide-ranging, interesting research. The authors have a blog discussing the book.

This was a quick, enjoyable read, and that's both the positive and negative review. This book is more like an extended NYT magazine article (where the authors first collaborated). So, if you want an introduction to cutting-edge views in economics (which purports to be more accessible to the non-economist), this is a very good book for that purpose.

I personally got a lot more out of The Wisdom of Crowds, which I've mentioned before on this blog. It is a much more in-depth treatment of the subject -- examining the economics principle that a group has more collective intelligence than any one individual in the group -- than Freakonomics, so therefore I found it more rewarding. The latter third of the book devolves a bit into more of a business management book, but the first 2/3 make up for it.

The baseball connection in these books? Levitt has posted on his blog about Billy Beane, making a simple but controversial point that although Moneyball's premise is that Beane used OBP to find guys who could help him win on the cheap, the stats show that the A's don't have any better OBP as a team than other clubs. Levitt thinks they won because of pitching, and will suffer mightily now that Hudson and Mulder are gone.

As for Wisdom of Crowds, it is my view that fielding statistics would become a lot more meaningful if we turned them over to the crowd.

P.S. To help you understand where I'm coming from, I have no formal training in economics whatsoever and for years thought its appellation as the "dismal science" was too good for the subject. That attitude has changed somewhat dramatically over the past months, due in part to blogs like Marginal Revolution, which tipped me to the Levitt book and which I strongly recommend for anyone who is curious about anything, literally.


At 8:06 AM, Blogger Josh Crockett said...

Here's the thing, though -- at the tail end of Moneyball, Michael Lewis (the author) mentioned that the market was starting to correct the undervaluation of OBP that Beane et al. had taken advantage of in building the '00-'02 teams. People paid (and still pay) lots of attention to the OBP case because they understood it, but the generic case of Moneyball was to determine which stats (and therefore which players) were monetarily undervalued relative to their ability to make wins, and use those -- whether they be OBP, defensive stats (although the conventional ones are damned near useless), or whatever.

The A's were already moving on to other methods of analysis and creating new weird stats to try to catch market inefficiencies in 2002, so it's no surprise that a couple years later, they don't show much difference in OBP relative to the rest of the league. Even the stat-dumb GMs (like our Jim Bowden, I'm afraid) know OBP's value now.

At 4:19 PM, Blogger DM said...

You may be right, but I haven't looked at the numbers on Oakland's OBP relative to the league five years ago.

I do think Levitt overlooks that Moneyball also talked about how to scout pitchers (e.g. don't take high schoolers, you can use college stats), and that's why they got guys like Mulder and Hudson and Zito for less. So Beane should get that credit.

Note that Oakland's payroll is now middle of pack in MLB, so it is hard to keep referring to them as a "small-market" team in that sense.

At 4:40 PM, Blogger tmk67 said...

Thank's for the Freakonomics review...I saw this referenced in The Economist over the weekend and was thinking of buying it. I think I'll wait for the paperback version!

You may be onto something about utilizing "crowds" for some pesky sabermetric issues, like fielding. Particularly as it relates to the subjective aspects of ERV Boxscores.

For instance -- utilize collective input from a crowd to determine whether the throw that missed a
cut-off man resulted in a player advancing from second to third (and later scoring). Or determining whether a player's defensive reputation caused the runner to stay at second after a single -- that might never show up in an ERV Boxscore but certainly has an impact on the game (and the outfielder with the reputation should be rewarded.)

At 4:49 PM, Blogger DM said...

Hi, Backward K! Good to see (or is that "read"?) you again. Yes, I think you can wait for the paperback on Freakonomics.

On the wisdom of crowds, that is exactly what I am thinking. Note that you could get into DIPS-like issues too, such as, "Who was really responsible for that ground out, the pitcher or the fielders?"

Essentially you could set up a program like Project Scoresheet and have people ERV Score games (past and present), then aggregate their scoresheets and see what the average score was for each play.

At 4:33 PM, Blogger tmk67 said...

DM, yeah, I've been away for awhile. Mostly tucked away in a fetal position given the way the Cubs have started the season.

You could also do it with a nifty applet that would appeal to folks watching the game on ("Would Neifi Perez had scored? Vote here.")

I thought of ERV Boxscore when I saw the Post snippet on the Red Sox game last night -- their lead was, "Kevin Millar made up for two costly errors that lead to two runs with a walk-off home run in the ninth." I wondered if the ERV of that game would really show that the walk-off home run really did that or not!


Post a Comment

<< Home