Estimating by Story Point (SP) is important for a project that is following Scrum. It helps Product Owner (PO) and Development Team (DT) quickly imagine the workload of the upcoming Sprints and vision the release roadmap. Good estimation helps so much; the better accuracy is, the better Sprint goes. Then choosing the “unit” User Story (US) or the baseline of SP is important step.
Some of the Scrum guideline chose the baseline of SP by following:
- Choose the “smallest” US and set it is 1 SP, call it a baseline.
- Estimate other USs by comparing them with the baseline.
But this method isn’t good enough by 2 reasons at least:
- What does it come between 1 or 2 if DT think a US is larger than 1 but less than 2 SP? We don’t have 1 and 1/2 in single card.
- How does DT maintain the consistent compare of each US estimation? How they can make sure all 5-SP US are the same?
As my observer, following guideline above leads the new DT to less accurate estimation. One signal of not-good estimation is that SPs are distributed around a number like 1, 2, 1, 2, 1, 1… as it shows that the DT doesn’t know how to compare the difference between USs so that they choose the same number just because of the feeling.
Another way to choose the baseline is:
- Choose the “smallest” US and set it 2 SP.
- Find a US that is good to set 5 SP.
- Call 2 and 5 baseline, estimate others.
The idea of the above method is:
- DT can reserve 1 SP for the future US that maybe less than today smallest – 2.
- DT understand how to compare US because they did it when finding 5-SP US.
- When estimating a US, DT can choose to compare it with 2 or 5 instead of single unit for better accuracy.
The idea of choosing 2 SP as baseline is getting DT understood about US comparison and setup more anchor for the next estimation. Then choosing 2 & 5, 3 & 8, 1 & 2… doesn’t matter but my recommend is 2 & 5 because they are small enough in both absolute value and their difference.
