Evidence Based Scheduling - Joel on Software

Now this is superb! I know it shows off how geeky I have become in software product management, but so be it.

One item I note probably pushes me even further into geekdom: I notice that Joel Spolsky recommends
"As estimators gain more experience, their estimating skills improve. So throw away any velocities older than, say, six months."
Now we in the Business Intelligence business cannot bear to throw data away, and those out there with an interest in Data Mining and Operations research (you know you're out there), will remind us that one could analyze the increasing accuracy of estimating skills, and then one might choose to reward it. Now if I could only remember the classification of analysis that would do the job...
Either way, bravo Mr. Spolsky, bravo.

Evidence Based Scheduling - Joel on Software