56307Re: [scrumdevelopment] How to increase velocity
- Nov 30, 2012I've observed something that supports Cass's theory somewhat. In the case of learning something new, what I've seen a couple of times is that the team has say four very similar stories, but they are using a new technology. They will size the stories something like 8, 3, 3, 2, stating that the first story will take longer because they'll have to get up to speed on the new tech. Then, the remaining stories should come faster. Is the right thing from a user story perspective to size them all like a 5? In that case, you would see an increase in velocity.
OTOH, I've certainly seen cases where velocity has improved and even skyrocketed when some large impediment is removed.
I have mixed feelings here -- while velocity can sometimes indicate increases in productivity, it's also pretty worthless IMO for indicating that at other times.
I mean, Ron, in your "count the stories" pattern, you indicate stories should be sized to be about 2-3 days in size. If a team doubles their productivity, then they can theoretically get twice as much done in 2-3 days --- so the amount of work in their stories will increase, but their velocity will not increase.
I return back to what I've learned from most folks on this list -- velocity is a planning tool, and trying to use it for any other purpose is very suspect.
From: RonJeffries <ronjeffries@...>
Sent: Thursday, November 29, 2012 2:04 PM
Subject: Re: [scrumdevelopment] How to increase velocity
Cass,On Nov 28, 2012, at 8:46 AM, Cass Dalton <cassdalton73@...> wrote:
Trying to increase the velocity means that you've fallen into the illusion of control mind trap. You can't change velocity. It is inherent in the team and their ability to work the stories. As long as the team is estimating stories honestly and consistently and the team makeup does not change, the true team velocity will be effectively a constant.What if they learn something new? What if someone has a creative idea?
- << Previous post in topic Next post in topic >>