Deadly sins of software estimation

This is so re­mark­ably sen­si­ble I think it de­serves its own ar­ti­cle.

It’s a pdf of the slides from a lec­ture, and should help with the plan­ning fal­lacy.

A few high­lights: Dist­in­guish be­tween tar­gets and es­ti­mates. Don’t make es­ti­mates be­fore you know very much about the pro­ject. Es­ti­mates are prob­a­bil­ity state­ments. Best as­sump­tion is that a new tool or method will lead to pro­duc­tivity loss.