1. why estimate

   1. scheduling - how long

   2. selection - which ones

2. estimates are based on stories

3. how to estimate

   1. by comparison

      1. by size - what's the measure

      2. by effort

      3. don't forget context

   2. by expected effort

      1. perfect weeks

      2. points per perfect week

   3. by prototyping - spike solutions

      1. unknown problems, bad estimates

      2. throw-away prototype to answer questions - drive through the problem in one blow

      3. this needs to be fast and cheap, not accurate or successful

      4. and remember, throw it away

4. using estimates

   1. stories per iteration

   2. iterations per development

This page last modified on 3 July 2001.