Wednesday, March 31, 2004

The Business Intelligence of
the Chicago White Sox  

Organizations that use "business intelligence" (BI) tools (and most do, to some degree, organized or otherwise) tend to become "lumpy" in their strengths. They tend to become much better where their BI tools and staff have high aptitudes, and this makes their weaknesses appear even larger.

It's not that lumpy strengths make their weaknesses worse (though it can happen...look at Enron's brilliance at creative accounting and playing Wall Street and how ultimately, it overwhelmed everything else in the organization to the point it that this toolkit of simple tricks became the corporate mission).

Anthony Giacalone recently produced for Baseball Primer a sharp preview of this year's Chicago White Sox, and he detected one of these BI tools making strengths lumpy trends in a way that's illustrative for managers all kinds of organizations. From his preview (emphasis mine):

(White Sox G.M. Ken) Williams’ horrible beginning as a GM highlights his shortcomings. Williams is just not a very astute judge of major league talent. Sure, he can understand that Bartolo Colon has a lot of ability but the overwhelming majority of his major league acquisitions have been failures. Basically, he has dumped major leaguers Keith Foulke, Chad Bradford, Tony Graffanino, Chris Singleton, Mike Sirotka, Kip Wells, Rocky Biddle, Jeff Liefer and Mark Johnson for Bartolo Colon, Tom Gordon, Esteban Loaiza, Royce Clayton, Juan Uribe, David Wells, Alan Embree, Jose Canseco, Scott Schoeneweis, Roberto Alomar, Carl Everett, Todd Ritchie, Cliff Politte, Armando Rios and Brian Daubach. Anything pattern in these moves? For the most part the players that he parted with were young and unproven while those that he acquired were not only older but also had established reputations.

Of all the major leaguers that he has brought in, only Colon (trade), Loaiza (FA) and Canseco (trade) have been successful to this point. But there is another side of Kenny Williams' dealings. Under Williams, the White Sox have been tremendously successful at acquiring unproven and minor league talent. Since Williams has taken over, the Sox have brought aboard unproven players Willie Harris, Damaso Marte, D’Angelo Jimenez, Miguel Olivo, Mike Rivera, Jamie Burke, Ross Gload and prospects Ryan Meaux, Enemencio Pacheco, Neal Cotts, Felix Diaz, Ruddy Yan for very little. That’s not to say that the White Sox won’t occasionally decide to bring in a bad player like Kelly Dransfeldt, but in sum that’s an impressive list. Further, not one single minor league player that the White Sox have traded away has been successful in the major leagues yet. In fact, in the last 10 years only Ron Coomer and Josh Fogg have been traded away by the White Sox and have later had any success at all. That’s a pretty good record.

It just seems very unlikely to me that a GM can be so bad at picking major league players but so good at analyzing his own minor league talent and acquiring it in other organizations.

Lumpy. Extra lumpy. A very high strength concurrent with a very marked weakness.

What many people don't know, even in the sabermetric community, is that the White Sox were an early adopter of sabermetric methods. Manager Tony LaRussa and executive Jack Gould embraced the STATS event-tracking technology around 1982, according to Dick Cramer, one of the big brains behind it. That organization didn't just run the system; they liked analysis-support technology so much, Gould & G.M. Roland Hemond had Cramer build them a custom scouting-support system. Practical and logical: use technology where the benefit/cost ratio of the status quo truly blows chunks, where the options are so multitudinous that current systems are destined to get overwhelmed, and where your competitors are unlikely to have the proper cognitive underpinnings to follow so that your comparative advantage can persist a while.

Where virtually everyone else was using whiteboards and index cards, scouting the Mongol Horde of potential playing talent, the Chisox applied special focus. And BI technology to back it up. (I think you can see where the discussion is leading).

The White Sox, because they invested in this strength, appear from the clear duality Giacalone brought to our attention, appear to be really awful at at their weakness. And they may be that awful, but it seems to me a majority of baseball organizations make these apparently foolish decisions with veteran players. I think, based on skimming through Retrosheet transactions, the White Sox are little worse than anyone else at their veteran-selection weakness, but that it looks worse because this strength is such a counter-example.


It's a natural gravitational field. As an organization invests resources (time, money congitive commitment) in one thing, it reduces the resources available for other things (See Enron example above) unless you're an organization like the Yankees (mucho dinero, obsession with excellence).

Sometimes it's the technology that creates strengths or weaknesses. Usually it's not. Usually it's the human talent involved that makes this happen. Like Frank Thomas taking extra batting practice when he really should be working on his footwork in the field, it's easy to enter a cycle where you do something well, get rewarded for it, and focus your resources in a narrow way on what the organization is rewarding you for. It's a universal temptation. I succumb to it myself; it's hard not to.

┬┐So what do you do about it?

I counsel organizations that do BI to run analyses not unlike the one Giacalone did for the White Sox. What are you finding out? Where are you successful with your BI? What are you really bad at? Could BI you don't currently do if it magically was perfect make any difference to that really bad attribute? And if you threw investment at it, would you still be good (maybe not quite as, but good still) at the BI you're good at now?

Preston's Law states if you want to improve your net performance, it's generally much better to turn your '1's and '2's into '5's and '6's rather than try to turn your '5's and '6's into '8's. It makes sense mathematically, ecologically, and in baseball, too.

Lumpy strengths & weaknesses may not just look like a worse curse than they are. They might just be the limiting factor that is both putting a ceiling on your organization's success and that is relatively easy to break through.

I suspect that's the case for the Chisox.

This page is powered by Blogger. Isn't yours?

free website counter