Occassional thoughts about orienteering
Saturday, May 17, 2003
Predicting future performanceI spent a good nine hours yesterday in airplanes and airports. I had a lot of time to read and think. I spent some of that time thinking about how to predict performance.
I was thinking about it in the context of hiring performance auditors and scouting baseball players. But, the basic question is just as interesting in orienteering.
How do you hire a good performance auditor? Performance auditing isn't a widespread job. You can't just hire people who have been good performance auditors at their last job. As a result, a lot of performance auditors come straight from school or from another career. Figuring out how someone will perform on the job is tough.
I read a story from Sports Illustrated about the process the Oakland A's used to select a player in the draft. Oakland employs a bunch of scouts who go around the country watching prospects and taking notes. Oakland also employs a Harvard-trained economist. The economist spends time looking at baseball statistics trying to find ways to predict performance and, in particular, take advantage of opportunities that other teams miss.
I won't go into the details, but the article (published in the May 12 edition of the magazine) makes for good reading.
Can you tell which top juniors are going to be the best in a few years? What would you look at to figure out who will perform best? If you could design a series of tests, what would you test?
When I was a junior, the top three in my age group were Doug Hollowell, James Baker and me. I'd say I've had a better set of results over the years as an M21 than Doug or James. But, as a junior, I didn't do as well. In 1984 (when all three of us were just about to turn 21), could someone have predicted who'd perform better in the next 5 or 10 years?
A few general thoughts about predicting performance
In just about any context, I think there are a few things to keep in mind when you try to predict future performance....
1. Past performance predicts future performance, but not perfectly.
2. It is easy to over-emphasize the most recent performance.
3. Is is easy to over-emphasize your own personal experience.
4. It is easy to over-emphasize the easy to measure stuff.
5. The hard to measure stuff is very important. It is probably worth trying to figure out how to measure it. posted by Michael | 5:02 PM
Comments: Post a Comment