Management by Baseball

In Baseball Prospectus' recent discussion of the Houston Astros, the commentator complemented team manager Jimy Williams on his handling of his mostly-young mostly-promising starting pitchers. Handling starting pitching, and the blood-drenched debate between the two established schools of belief, is a key area where baseball has a lot to teach non-baseball management. That debate parallels the two sides in Douglas McGregor's Theory X/Theory Y schools of management.

Tersely, Theory X posits people at work should be made to perform as closely to machines as possible, by control and coercion. If you haven't read Lewis Mumford on how this world view came into being, it's worthwhile (Technics & Civilization, Myth of the Machine are the most focused on this). Lots of people think Theory X is passé in North America, but it's made a big comeback in the last four or five years in industry, and has strong support in the technology world among those who argue the laughable delusion that "the brain is like a computer". Theory Y posits people work hard when they're given incentives and because they naturally want to be good at what they do and that control is counter-productive.

Bitgods and Bitmonsters

In baseball, Theory X is the "Back In The Good Old Days" (BITGOD) theory of using a starting pitcher. There's compelling evidence that starters on average in today's environment are most continually effective if they're limited to throwing 100-115 pitches per start. But the BITGOD adherents, usually of the generation that had pitchers who pitched a lot of complete games and no one was counting pitches except some scorekeepers, measure effort by innings. They also tend to conflate masculinity with complete games.

In baseball, Theory Y is Pitcher Abuse analysis. Rany Jazayerli argues:

The injury rate of pitchers, in particular young pitchers, is astonishing. Pitchers are several times more likely to get injured than hitters, and for every prospect that becomes a successful major league pitcher, a dozen more have their careers stalled or ended by injury. This is a reality of baseball that has persisted since the game was invented; the act of throwing a ball overhand is inherently unnatural, and the repetition of throwing, even with excellent mechanics, can lead to inflammation or injury to the muscles of the rotator cuff, or in the ligaments that hold the elbow in place.

Hard to argue with. Basically true, not only in baseball, but by analogy in the non-baseball workplace, too. I once supervised a big multi-city project that had over 20 people producing white-collar analysis work that was measureable both quantitatively (how much, how many, how complex) and qualitatively (how accurate). Once we had a good bit of track record, we analysed the history and found two things. One, no two people were generally alike in their patterns, every individual had important differences. Two, we found that that no-one was at least half as effective in their 53rd hour as they were in their first 36 (which were pretty close throughout them for effectiveness). Worse, someone who broke the 52-hour barrier for work in a week was diminished from their previous week's performance for the whole next week, with some individuals lugging for two weeks.

The modern "More With Less" cult loves to exhort employees to work beyond 52 hours (in most cases, they aren't paying overtime, so in the instant benefit/cost thumbnail, they feel they're getting more) even though they're getting less than half an hour of work per hour applied, and degrading the next week's performance, too. Bad management. Strip-mining.

In baseball, scores of young arms have been maimed for life, careers broken or just diminished by BITGOD pitcher management. Others have just been made less impressive because most pitchers who throw a 130 pitch start and come back to their next start on normal rest underperform their own norm in the next start. Any minimally-competent project manager would tell you in almost all circumstances they'd rather get a little less out of a team member today if it increased the assurance that it would return them a lot tomorrow.

But there's one shortcoming in the Pitcher Abuse school, an unnecessary one. In Jazayerli's seminal essay on the subject, he recognized right up front that each pitcher was an individual, with different patterns. But a lot of the current Pitcher Abuse work tends to view "pitchers" as a class that can be averaged. So the argument about Jimy Williams management of Astro hurlers' arms may or may no be valid. It's valid as a general, average case. Is it valid for the 2003 Astros, and more to the point, is it valid for each of the individuals who are Astro pitchers?

Some pitchers are currently being overused on a regular basis (compared to the 100-115 pitch average standard). Most suffer immediate diminution. For example, Mark Redman of the Marlins looks to be a guy suffering from immediate effects of outings of over 115 pitches. He's allowing more baserunners and apparently shifted from being a pitcher who induces a lot of ground balls (impossible to loft home runs, harder to hit doubles and triples) to one who is throwing pitches that end up as fly balls (easier and easier). But in contrast, the Cubs' Mark Prior seems to have little short-term diminution from his blood-curdling overuse. He broke down earlier this summer, but his high set point for next-few-starts effectiveness seems closer to 120 pitches per game than to 100. Other guys can't go 100 without it degrading them. In the long term, there will undoubtedly be health effects. These, too, we can guarantee two things about...the same two things we can guarantee about non-baseball work.

In the Non-Baseball World...

First, the overused will more likely fail to maintain effectiveness, and second, there will be many individual patterns, many of which will defy the averages.

TIP: Remember when you collect and apply measures that averages, once grouped, may disguise very different individual patterns. And don't be a BITGOD.