Any explicit measures for quality using pair programming? I'd be interested in seeing the actual changes in escapes. A related question is what it does to the productivity compared to a more traditional personal review followed by peer inspection. .

The missing ingredient in the "thought experiment" introducing a SCRUM like process is natural variation from two sources. First, the task variance from estimate, second the actual effort variance from the plan or "average". Much like normal planning, we get a best case scenario that is good enough as a demonstration of plausibility and a target we could steer towards. In practice variance pushes things out, probably by 10-20%.

Proud of you Aayush and Aditi.

The title is very promissing. The content is just the comparison of TSP and Scrum:( Thats all.