So considering my last post on t-shirt sizes and burn-downs, how do you decide on how many t-shirts to take on in the next iteration? If you assign some value to each size I guess it is easy since you have your velocity to use. But I heard another interesting thing from a guy I worked with the other day. He had been on a team where they did not assign any number to their t-shirt sizes. They just expressed their velocity in terms of how many user stories of each size they could make in each iteration. For example they might have had a velocity of “2L, 1M, 1S” meaning that in each iteration they tried to complete two large, one medium and one small task. I like this idea since it is simple. And easy to adjust. When you realize you don’t have that many large user stories left but a bunch of medium ones, just try a new mix and adjust until you know your new velocity (which might be 1L, 3M, 2S).