Estimation is a broad topic that provokes much discussion in the software engineering community. In the past decades many estimation models, theories and techniques emerged. However, in my opinion, the accuracy and quality of estimates in software engineering does not seem to be much improved.

The purpose of this presentation is to share some of my personal lessons learned with respect to estimation, some models I've used (i.e. PERT, UUCP, COCOMO II, analogous), and how can you improve the reliability of the estimates in a software development project.