I Estimate this Talk will be 20 Minutes Long, Give or Take 10 Minutes

on YouTube ↗41m

" summary

Argues that points track relative complexity not hours and that useful early forecasts come from sizing stories, summing low and high point totals, then applying a fake velocity range to produce a budget or schedule. Uses Fibonacci sizing, needs a few iterations of data (2–6 weeks) to measure velocity, and often yields wide ranges like 100–120 points → 500–840 hours. Presents this as a working consultancy practice (Table XI) with emphasis on transparency and updating estimates as work proceeds.

AI-written summary · about

" chapters (32)

1. 0:00Start
2. 0:34How far do you live from an airport?
3. 4:20Ever done a home remodel?
4. 4:59Big Dig
5. 7:06What if we just did stuff and got paid?
6. 7:40Customers reasonably expect estimates
7. 9:54Bad Estimates communicate: Unrealistic expectations
8. 11:11Estimating at the beginning is the worst
9. 11:47How can we estimate Effectively?
10. 12:17A quick sidebar of agile terms
11. 12:48Iteration The unit of planning in a project
12. 13:09Point The size of a story
13. 13:21Velocity Points per interation
14. 16:09point story
15. 16:50What makes an estimate wrong?
16. 17:07Scope Creep
17. 17:32Rework
18. 19:03A lower estimate doesn't make it less complex
19. 19:28How to do better
20. 19:55What goes into the amount of calendar time a feature takes?
21. 20:33We are bad at estimating time on task
22. 23:07Estimate complexity and let time sort itself out
23. 23:19The humble point
24. 24:12Time on task and relative skill average out
25. 25:38Decomposing the problem RUBYCONF into similar size chunks
26. 26:10one-off estimates
27. 26:34beginning of a project?
28. 26:55Fake your velocity
29. 27:13Don't spend time chasing precision
30. 27:52Split the tasks as best you can
31. 28:24Don't spend a lot of time assigning points to a single task
32. 29:40Apply a load factor (5-7 hours per point)

" related talks

• 46mEstimates and complexity: a tale in 42 and a half minutesGary Fleming · Agile on the Beach 2023
• 01h 45mEstimates or NoEstimatesWoody Zuill · Agile on the Beach 2018
• 57mThe Science of not estimatingAnemari Fiser, Javier Sánchez · Software Crafters Barcelona 2018
• 20mHow long is a piece of string? Solving the conundrum of software estimationJonathan Rigby · UXDX Europe 2019
• 47mCreating efficient, accurate, software estimationsDominika Rogala, Amir Safavi, Yenny Cheung, Daiany Palacios · LeadDev Live Summer 2020
• 41mSo you think you can estimate? The lost art of useful estimatesMegan Folsom · Agile India 2012
• 44mAgile metrics for predicting the futureMattia Battiston · Agile on the Beach 2021
• 05mIgnite - How to Use Scrum Poker to Slay EstimationsLaura B. Janusek · devopsdays Portland 2019