Scrum Trainers Gathering (4/4): Affinity Estimating

Update: Since writing this post, I’ve become aware of two other descriptions of Affinity Estimating. Jukka Lindström posted his experiences on ScrumDevelopment, and Chris Sterling has written a great Step-by-step description.

Update: part 1, part 2, part 3 and part 4.

The last session that I’d like to report on is Lowell’s “Affinity Estimating.” Affinity Estimating is a technique many teams use to quickly and easily estimate (in Story Points) a large number of user stories. This is a great technique if you’re just starting a project and have a backlog that hasn’t been estimated yet.

We started by reading out each Story to the entire team. Lowell then asked us to arrange the stories horizontally on a wall in order of size, without talking. We placed the largest stories on the left and the smallest stories on the right. This only took a few minutes. We were then given a final opportunity to make adjustments to the ordering, again without talking.

Lowell then placed some numbers above the list of stories. In our case, he used the Fibonacci numbers as described by Mike Cohn in “Agile Estimating and Planning“. He asked us to group the user stories around the nearest number. After only a few minutes, our team had estimated about 30 User Stories.

I loved this estimating technique for a number of reasons: It’s quick and easy; it feels very natural; and, the entire decision making process is made very visible. Finally, “Affinity Estimating” helps make estimating a positive experience rather than a confrontational one. The next time you have a need to estimate a large number of User Stories, consider trying “Affinity Estimating.”

Bookmark and Share

14 thoughts on “Scrum Trainers Gathering (4/4): Affinity Estimating

  1. I like it. Real simple and highly interactive. I especially like the “no talking” rule. I think it makes for some very interesting communication. I shall try this as soon as I get the opportunity.

    One question: how were disagreements resolved, or were there none?

  2. Hi Tobias,

    >how were disagreements resolved, or were there none?

    Great question. We didn’t have any disagreements during the exercise. I can think of a number of reasons why this may have been so: The example Stories may have been too abstract; The initial discussion may have driven out all disagreements; or, the visibility permitted by this approach enable disagreements to be made transparent and resolved.

    I suspect (gut feel) it’s that final point (i.e. visibility allowed disagreements to be made transparent and resolved) because *none* of the teams had serious disagreements. I’d been interested in your feedback from practical experience.

  3. I can see how this would be a quick way to get relative estimates. I use planning poker now with the the delphi model of having the first “bids” be silent, followed by discussion and a re-bid. The discussion can take a while, especially for the first few stories.

    I always point out afterwards that a fair amount of story clarification and design happen during estimation. In this method there is no talking so those benefits are missing. That seems to me to a drawback. Are those recovered in some other activity? Or does your “??” mean that we are not sure about talking on the second round?

  4. There is some degree of discussion at the very start, but it cannot be a replacement for a detailed discussion that you’d normally have while playing planning poker.

    I think there are some techniques that are appropriate to use with new teams, and some techniques that are better used with experienced teams. “Affinity Estimating” falls into the latter category, in my personal opinion. I’d only try this with a team that has already bonded.

Comments are closed.