Guesstimate - Some Thoughts on Estimation in Software Development
3 likes1,320 views
This document discusses techniques for estimating project parameters and outcomes. It describes expert estimation, where an expert in the domain is asked to provide estimates. Formal modeling approaches are also discussed, like COCOMO 2 and function point analysis. Good practices for creating estimates are outlined, like providing a range instead of a single value, stating assumptions, and reviewing estimates over time. Estimation techniques like affinity mapping and analogy to past projects are presented as well. The document encourages focusing on delivering value, working iteratively, and collaborating with customers.
More Related Content
Similar to Guesstimate - Some Thoughts on Estimation in Software Development (20)
More from Project Management Berlin Meetup (6)
Recently uploaded (20)
![SketchUp Pro Crack [2025]-Free Download?](https://cdn.slidesharecdn.com/ss_thumbnails/tighteningeverybolt-fosdem2025-250224092147-cc175839-250305114942-10710f44-thumbnail.jpg?width=560&fit=bounds)
Guesstimate - Some Thoughts on Estimation in Software Development
- 1. Photo by gordon (TD8316) -Creative Commons Attribution-NonCommercial-ShareAlike License https://www.flickr.com/photos/22934398@N05 Created with Haiku Deck Guesstimate Peter Rubarth, 2014
- 2. Prediction is very hard, especially if it¡®s about the future Nils Bohr
- 3. Exercise 1 ¨CExpert estimation Write down your estimates for the following facts: 1.Length of German border (in km) 2.Amount of water in Bodensee lake (in l) 3.Dollars (as notes) in circulations within the USA in 2004 (in $) 4.Birth year of Alexander the Great (year) 5.World wide revenue with movie ¡°Titanic¡± (in $) 6.Latitude of Shanghai (in degrees) 7.Temperature on the surface of the sun (in ¡ãC) 8.Number of bacteria in human intestines 9.Number of works of Ludwig van Beethoven 10.Expanse of Asia (in km^2) 11.Number of stars in Milky Way 12.Number newly published books in Germany 2004-2007 13.Area covered by Hamburger Hafencity including water area (in ha) Thankyou@projectzone-http://yourprojectzone.com/2014/03/30/uebung-team-estimation-game
- 4. Why do we estimate?
- 5. ¡ñPrepare project bid ¡ñPlan time, resources, ¡ ¡ñFoster discussion ¡ñDetermine economic feasability ¡ñSomeone told us so ¡ñ¡
- 6. What do we estimate?
- 8. How do we estimate?
- 9. Expert estimation ¡ñAsk someone who should know
- 10. Formal model ?COCOMO 2, Function Points, IFPUG FPA, LOC, COSMIC Full Function Point (FFP) ¡
- 11. What could go wrong?
- 13. Negotiation ¡ñNow cut that in half and we have a deal
- 14. Estimates become targets ¡ñBut you¡¯ve said it will take only so long!
- 16. Work expands so as to fill the time available for its completion. ¡ñParkinson¡®s Law
- 17. This is why buffers (or the Project Managers¡® multiplier) don¡®t work.
- 18. Cycle of Mistrust Deliver the wrong thing, or late Trust breaks down Commitments of scope and time Focus on schedule ? Copyright Neil Killick, Iterative, 2013
- 19. Good Practise
- 20. Clarify objectives ?For which purpose? ?What are the risks and mitigation strategies? ?Awareness of the nature of estimations?
- 21. Provide ranges not a single value Likely effort from ¡ to ¡. Helps to communicate that an estimate is not a guarantee
- 22. Use indirection, avoid direct time estimates ?Abstract units such as Story Points, Ideal Days or T-Shirt sizes help to communicate uncertainty ?Support empiric forecasting
- 23. Explicitly state assumptions Provide a disclaimer: ?Which assumptions ?Use approach and data ?Risks and possible impact At least you have told `em
- 24. Review and update your estimates ?Use what you have learned ?Do not hide unwelcome news ?Burnt hours are a bad progress indicator
- 25. The evil of Fixed Price Contracts Knowing it will take longer does not help, because everything is fixed (unless people really idled before) Strong incentive to cut corners and damage quality
- 27. ?FOCUS ON VALUE ?DELIVER SMALL SLICES ?DELIVER EARLY & FREQUENTLY ?CUSTOMER COLLABORATION ? Copyright Neil Killick, Iterative, 2013
- 28. EMPIRICISM ?Use actual results to predict progress
- 29. Useful Methods
- 30. Group Estimation Don¡®t rely on a single opinion Methods: ?Wideband Delphi ?Planning Poker ?Affinity Mapping
- 31. Analogy Method ?Compare project to previous experience ?Use checklist to compare and classify on different dimensions ?Only useful if sufficient similarity
- 32. Affinity Estimation ?Silently order items relative to each other ?Build clusters ?Apply appropriate scale
- 33. Now it¡®s your turn!
- 34. Exercise 2 ¨CAffinity Estimation 1.Take first item and place it in the center 2.Take turns and position subsequent item relative to existing items 3.Reorder if you disagree 4.Do not discuss!
- 35. Exercise 2 ¨CAffinity estimation Write down your estimates for the following facts: 1.Length of German border (in km) 2.Amount of water in Bodensee lake (in l) 3.Dollars (as notes) in circulations within the USA in 2004 (in $) 4.Birth year of Alexander the Great (year) 5.World wide revenue with movie ¡°Titanic¡± (in $) 6.Latitude of Shanghai (in degrees) 7.Temperature on the surface of the sun (in ¡ãC) 8.Number of bacteria in human intestines 9.Number of works of Ludwig van Beethoven 10.Expanse of Asia (in km^2) 11.Number of stars in Milky Way 12.Number newly published books in Germany 2004-2007 13.Area covered by Hamburger Hafencity including water area (in ha) Thankyou@projectzone-http://yourprojectzone.com/2014/03/30/uebung-team-estimation-game
- 36. Exercise 2 ¨CSolution Find the solution at http://yourprojectzone.com/2014/03/30/ uebung-team-estimation-game/
- 37. Thank you! Twitter: #PRubarth