The Art of Prediction
summary
The talk examines why software time estimates are often wrong and how cognitive biases lead developers to substitute simpler questions (e.g., “how long to write code”) for the real project-duration question. It cites project-success statistics (only ~40% finish on time without quality loss) and discusses practical implications for improving estimation practices across Agile and Waterfall projects.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:06Project estimate problems
- 3:58Why estimates fail
- 8:19Underestimation vs overestimation
- 10:13Practices for accurate estimates
- 14:10Reducing project unknowns
- 17:24Team estimates and biases
- 20:04Downsides of detailed estimates
- 23:15Comparative estimation method
- 26:27Time tracking data usage
- 28:13Low and high double estimates
- 30:44Project risk reduction
- 32:34Documenting estimates
- 33:40Summary
- 34:23Q&A
related talks
46m¿Por qué estimamos tan rematadamente mal?
31mI want to grab the next 23 minutes and 11 seconds to discuss time estimations
54mEstimates – the Enfant Terrible of Software Development
30mSoftware Estimation Made Easy
46mEstimates and complexity: a tale in 42 and a half minutes
01h 32mNo Estimates?
40mEstimations: from guessing to weighing in grams
47mCreating efficient, accurate, software estimations