When I do original estimates, they're usually pretty bad. Once I get into the code, I have a much easier time estimating how long something is going to take. But I'm told that even if I update my estimates, FogBugz's EBS uses the original. I don't want FogBugz to use the bad estimates. I want it to use the good ones!
|
3
1
|
||||
|
|
3
|
It's natural to want to do this. Seeing our old, ugly estimates makes us think we're doing a bad job. We naturally want to correct them. But, EBS is designed to help you look into the future based on limited, gut-level information. If we used revised estimates to predict how you would do in the future, then we would need you to revise your estimates before we made any prediction. That sort of defeats the purpose of prediction. So, for instance, let's say you have a dedicated user who makes all the estimates. You sign in as that user for your group meetings and run through the case list, giving a consensus estimate to each pending case. When we run the EBS simulations, we care primarily about how good that consensus estimate has proven to be in past releases, and we relate that to your current estimates and your current release. Revising estimates as you go along is still good practice, as it will help you to see how the release is developing, but the original estimate actually allows you to see into the future. |
|||
|
