Question About Scrum Velocity

A peer of mine is working on a major software development effort. He is a Software Architect in the role of "ScrumMaster" and is just learning the ropes of Scrum. He works for a very major consulting organization. He sent me this e-mail (excerpt):

From: MN
Sent: Monday, February 25, 2008 8:21 AM
To: Santiago, John
Subject: Question about SCRUM velocity

We've been measuring velocity for three days and, as you suggested, the numbers are relatively low. They are heading in the right direction. Now that we're moving, do you think velocity will improve naturally or do you think I should be doing things to encourage increased performance? If so, what sorts of things might help? The team isn't reporting much in the way of barriers; its more like we underestimated things.

MN

After reviewing his backlog, I had some suggestions about individual challenges some folks might be having, and I also suggested that he check out this great post about Sprint Retrospectives so that he can examine in more detail the Sprint with his team:

http://pragmaticintegration.blogspot.com/2007/01/first-scrum-sprint-retrospective.html

He responded back with the following:

I'm not stuck but vastly underestimated the amount of work. I thought my estimate was pessimistic when I originally made it. Sometimes when you turn over a rock, you find something you wish you hadn't found. As far as I can tell, it absolutely needed to be dealt with and I can't think of a good way to cut the work short.

I think that acknowledgment is critical to understanding how Scrum operates so I sent him the following e-mail:

From: Santiago, John [mailto:John.Santiago@emc.com]
Sent: Wednesday, February 27, 2008 10:52 AM
To: MN
Subject: RE: Question about SCRUM velocity

“There are known knowns. These are things we know that we know.

There are known unknowns. That is to say, there are things we know we don't know. But, there are also unknown unknowns. These are things we don't know we don't know.”

-- Donald Rumsfeld, befuddling the press at a Pentagon briefing

A huge part of software development is learning to recognize what you don’t know. What you will find is that your velocity will tend to account for this over the course of the project… allowing you to fairly accurately predict the impact of the unknown. Assuming that there aren’t monumental roadblocks in your way of course, which is why project managers still need to actively do aggressive risk identification and management even in a Scrum environment to steer around potential roadblocks.

What you must be doing right now to maximize your team’s velocity is to be very aggressive about removing any gold-plating from the project and to make sure the team is focused only on the minimum functionality needed to go-live.