Comments on: Step 2: How To Estimate Your Product Backlog

By: Sam Gamaliyal

Sam Gamaliyal — Sat, 12 Feb 2011 12:32:54 +0000

Hi,

I have doubt whether the Fibonacci sequence numbers will be used only for estimating the size of the first simple story for a particular sprint. And this story would be used as a benchmark for comparing all other stories for that sprint.

For e.g. The team or scrum master will select a simple story. After that to find the size for that story, the team will play the poker by using the fibonacci numbers as story points sizes (like 1 SP, 1 SP, 2 SP, 3 SP, 5 SP etc). And after playing they will arrive at a size (lets says 3 SP) for that story.
And now the size for all other stories for the sprint will be estimated by comparing with the size of the first story mentioned above.

Please reply. Thanks in Advance.

By: tsietsi

tsietsi — Thu, 10 Sep 2009 13:28:41 +0000

Priorotizing backlog can be tricky though if one doesn't keep ones' mind open. Consider this: You have a list of priorotised backlog from the product owner (initial list). After estimating work item tasks usin Fabonacci numbers product owner decide to change priorities to have the small ones to go first. Now, what happens if the delivery of the items at the top of the list is dependent on the ones at the bottom of the list? Can we have a so called dependency backlog for the project?

By: Pinto Philip

Pinto Philip — Fri, 27 Feb 2009 11:14:00 +0000

What is kindof confusing here is that in agile projects all the teams :Dev,QA and Doc and other teams if involved will decide the size of the feature (Not Only Dev and Product Owner) it is quite unlikely that the size decided by each team (Planning poker) would logically be the same as each team will think from their perpective , like dev thinks how much time it will take for coding so they come out with a size XXL(in this case it gets 21 points) as multiple layers of code\design patterns may be involved and QA thinks about QAing which should not be much and so on..so estaimation of a feature might become like a share market instead and I have personally seen this.

By: Lunar Logic Polska Sp. z o.o.

Lunar Logic Polska Sp. z o.o. — Tue, 11 Dec 2007 13:44:00 +0000

Mike Cohn’s book, Agile Estimating and Planning, changed my life. We now do estimates in a fraction of the time, with more accuracy. Why? Because more minds uncover more risks. Perhaps the biggest advantage of planning poker is that the client (product owner) offers a very valuable overview of all of the features considered for an iteration up front, so everyone starts with the same vision. Then details, of course, are discussed between the product owner and the developer coding the feature.

By: mdediana

mdediana — Fri, 28 Sep 2007 14:23:00 +0000

I prefer the pseudo-Fibonacci set too because it suggests that the bigger the number, the less precision it has. I feel it makes more sense think about a 40 than a 34 or 55.

And to anyone interested in
learn more about planning poker I suggest ail.html" REL="nofollow">this link.

By: Artem Marchenko

Artem Marchenko — Fri, 28 Sep 2007 08:32:00 +0000

I prefer pseudo-Fibonacci set: 1, 2, 3, 5, 8, 13, 20, 40, 100. Somehow the big difference between 20 and 40, between 40 and 100 more clearly indicates that if item is that big, it probably has to be split up before actually starting to work on it