June 21, 2003

New National Post Column Fr...

New National Post Column From Me -- 'Turning Baseball on its Ear': Here's the lede:

Can a ragtag, decentralized army of passionate outsiders and unpaid enthusiasts use their brains to break down 140 years of tradition and transform a multi-billion-dollar industry?

Hell yes, it can.

Posted by at June 21, 2003 11:21 AM
Comments

And "moneyball" teams have won exactly how many World Series'? The Minnesota Twins, a non-"moneyball" team does just as much as the Oakland A's, and with just as small of a payroll. Big deal. Once "moneyball" teams are winning the World Series on a conistent basis then the theory will be proven. Until then it's just another pointless thing to argue about down at the bar after 5 or 6 beers.

Posted by: David Crawford at June 21, 2003 03:21 PM

I guess I must have missed those 296 wins and three consecutive playoff appearances....

Posted by: Matt Welch at June 21, 2003 04:01 PM

Matt:
I've been a regular visitor (and user) of the Retrosheet information and know several members. Great - but a little geeky - bunch of folks.

It seems to me that the Jamesian method (if we can label it that) works at the "margins", that if you apply it here and there you can pick up a terrific setup reliever (Chad Bradford) or an excellent reserve (Catalanotto). These are key supplemental players around which you build a team.

Still need a Hudson (first round pick), a Carlos Delgado, a Miguel Tejada, a Zito, et cetera.

SMG

Posted by: SMGalbraith at June 21, 2003 05:02 PM

Just re-read my post. I visit the Retrosheet site (along with its companion site - SABR) and I'm saying that THEY (the members) ARE geeky?

Okay, I have met the geeky and it is us (me?).

SMG

Posted by: SMGalbraith at June 22, 2003 09:02 AM

Okay Matt:
So how would James rate the following players? Game in 1922:

8/22/1922 (Indians at Yankees) - In the top of the sixth, pitcher George Uhle singled to center. The Yankees then agreed to let Les Nunamaker run for Uhle while the latter had a shoe repaired. Nunamaker was out rounding 2b on a strange play when Joe Dugan threw poorly to 2b trying for a force on Nunamaker. Les started for 3b but retreated towards 2b and was tagged before he got back to the bag. The inning ended when Nunamaker, now coaching 3b for the Tribe, grabbed Bill Wambsganss as he rounded 3b to hold him there on a single to lf. Uhle returned to the mound in the bottom of the inning and finished the game, beating the Yanks 6-2.

Put all that in the computer, hit enter, and watch the dammed thing blow up.

SMG

Posted by: SMGalbraith at June 22, 2003 09:43 AM

Great article !! I would add that the moment I began reading my first Abstract, on a long trip to the Bay Area in the summer of 1982, was tranformative in my life as well. What I found most captivating about Bill James was his fearlessness in attacking sacred cows, and the seriousness he approached the subject of baseball. Every March, I would obsessively visit Moe's Bookstore in Berkeley each day, waiting for that shipment to come in, feeling like I was part of some secret club of the knowing.

The world needs a Bill James blog...although I heard he was working on a book about college basketball, of all things.

Posted by: Steve Smith at June 22, 2003 10:45 AM

Congratulations on the award! I haven't been reading your blog for very long, but anyone who can write a piece like "Turning Baseball on its Ear" deserves such adulation and more. I know little about baseball but was moved almost to tears by the dynamics of this story. And on a practical level, where else would I have learned the true origins of "Fenway" and "Wrigley" -- the names of these two dogs in my neighborhood? (For years I assumed they were just goofy names the kids gave them, and the dogs never filled me in. Please no spatula!)

Posted by: Eric Scheie at June 22, 2003 09:05 PM

I'm not sure the NBA is ripe for a sabermetric revolution, at least without some major, qualitative advances in statistical measurement. I may sound like the baseball Luddites who laughed at Bill James, but in basketball, there's just no substitute for seeing the guy play.

James addresses the difficulty of measuring baseball defense (fielding % and errors are overrated, etc.). But at least he found a better way (total number of chances). In basketball, the data just isn't there. There's no way to measure defense, apart from blocks and steals, and those numbers aren't all that useful. I suppose an official scorer could sit there and keep track of every time a guy makes a stop or gets burned on defense, but that sounds pretty unlikely.

That's a big problem, because defense is half of every basketball player's value. Not having a good way to measure defense doesn't hurt your evaluation of a first baseman or a left fielder too much.

The other reason basketball can't be distilled to raw numbers is that its a true team game. All the parts have to fit. You can't stock up on whatever the equivalent of "fat guys who draw lots of walks" would be.

I know one of James' theories is that statisical performance at one level (minor leagues) is a good indicator of performance at the next level (big leagues). Not true in basketball. The difference in athletic ability between even big-time college ball and the NBA is huge. Now imagine scouting a high school player. How the hell do you evaluate a 6-8 center by looking at his numbers? I saw Mark Madsen play in high school, and he looked dominant. But it didn't take a genius to see he was too small and too slow for the NBA.

I've read a couple of those guys who call themselves "the Bill James of basketball." The best I can say is that they're interesting, but they're not going to change anybody's entire outlook on the game.

Posted by: Tony Biasotti at June 23, 2003 12:22 PM

The problem with most attempts to use sabermetric principles to the study of basketball is that there is not yet a basketball version of Bill James. The important thing to remember is that there had been efforts to rigorously interpret numbers in baseball going back to Branch Rickey in the '50's, but until James came along, there was no one who made the numbers sing. I am of the opinion, in fact, that his statistical analysis is the least interesting aspect of James' writing; after all, he seems to change his mind every five years about whether the career numbers of Don Drysdale or Joe Tinker merit their inclusion in the H.O.F.

As far as your point about basketball defense, the question any hoopsmetrician would have to ask is the same questions James (and other sabermetricians) would ask about baseball: what does the individual player do to prevent points from being scored? and what does he do to enable his team to score more points than the other team? Steals and defensive rebounds are clearly important in that context, since each takes a possession away from the opponent. Fouls are important, since a good deal of the time they directly lead to the opponent getting free throws. Blocked shots, while interesting for the fans, is of marginal value, since it only lessens the other teams FG%, and does not assure a change of possession. A +/- chart, indicating the margin of points the teams scores in relation to their opponent when a player is on the court, would also be useful, since it allows us to compare individual players to their teammates and to other players around the league.

Posted by: Steve Smith at June 24, 2003 11:27 AM

It would be more useful to keep track of the point differential between a player and the guy he's matched up against. Of course, with zones, that would be tough. And even in a man-to-man defense, you're not always guarding the guy who's guarding you. If the Lakers play the Spurs, Duncan is often guarding Shaq, but Shaq is rarely guarding Duncan, because Shaq is too fat and slow.

Blocks, steals and fouls are all sort-of useful (though I don't think fouls in themselves are a sign of bad defense). The real thing no one is measuring is actual man-to-man defense -- are you stopping your guy from scoring? That's a lot more important (and often unrelated to) getting lots of blocks or steals.

Offense is much more measurable. You help your team score by making shots and getting rebounds and assists. Missing shots and turning it over hurts your team. We've got stats for all of those things. The only thing that's missing is setting screens and making good, non-assist passes, but I can live with those shortcomings.

Anyway, I don't think imperfect statistics are a shortcoming. I'm a maniacal professional basketball fan -- I watch every game I can without driving the wife completely nuts -- and I think that makes me pretty informed. I can watch a guya "3", the official scorekeeper? Wouldn't that impose a subjective burden not unlike the one imposed in deciding whether a fielder has committed an "error"?

Speaking of "errors", what the h*** has happened to the Angels?

Posted by: Steve Smith at June 25, 2003 08:45 AM

Steve -- Oh, it'd only be a best guess, and I wouldn't take it too seriously. Still, in any unit of five, you can get a crude idea. The most important concept, though, is to measure the *unit*, and the *team*, and work back from there to the individual. Least that's what I think.

As for the Angels: Mebbe Eck & Spez were over their heads last year, and whatever magic dust Bud Black had been sprinkling on Lackey & even Mickey Calloway has run out quick. Terrible work from the starters, and we're too slow to recognize that Sele hasn't been a major league pitcher for several years now.

Posted by: Matt Welch at June 25, 2003 06:47 PM
Post a comment









Remember personal info?






ear the top of the league in what I call "adjusted field goal percentage" (field goal points divided by field goal attempts), and "total adjusted shooting percentage" (the latter, blended with the percentage & value of free-throws). Why is that? Well, they run an especially efficient offense, with two of the best offensive players ever. What does that do? It makes stars out of unlikely role players -- Donyell Marshall, Shannon Anderson, Howard Eisley. What do these people do once they leave John Stockton? Suck eggs in the night, 9 times out of 10. Oh, I could go on for days.

Posted by: Matt Welch at June 24, 2003 03:29 PM

Matt W.:

I'm in agreeance with your post, although I have some question as to how you would incorporate "position" into your analysis. That would seem to be more of a baseball issue; after all, a baseball team will have only one shortstop, one catcher, one left fielder playing at any particular time, whereas basketball teams will go with two shooting guards, or two centers, depending on the game situation (and the talent available). Who would make the determination of whether Kobe is playing a "3", the official scorekeeper? Wouldn't that impose a subjective burden not unlike the one imposed in deciding whether a fielder has committed an "error"?

Speaking of "errors", what the h*** has happened to the Angels?

Posted by: Steve Smith at June 25, 2003 08:45 AM

Steve -- Oh, it'd only be a best guess, and I wouldn't take it too seriously. Still, in any unit of five, you can get a crude idea. The most important concept, though, is to measure the *unit*, and the *team*, and work back from there to the individual. Least that's what I think.

As for the Angels: Mebbe Eck & Spez were over their heads last year, and whatever magic dust Bud Black had been sprinkling on Lackey & even Mickey Calloway has run out quick. Terrible work from the starters, and we're too slow to recognize that Sele hasn't been a major league pitcher for several years now.

Posted by: Matt Welch at June 25, 2003 06:47 PM
Post a comment









Remember personal info?