Team Velocity: synthetic/hypothetical/predicted values

Imagine looking at a developer in Tracker in the context of a specific project, and having Tracker tell you what velocity to expect from that person.

Imagine assigning teams to a project, and having Tracker tell you what resulting velocity to expect for the project.

Imagine having Tracker recommend a set of optimal team members and teams for a project based on the history of your organization members' past work, velocities, and types of problems.

This would be possible if Tracker kept track of velocity at the individual developer level (still measure it at the team level, but allocate it to each member of the team). And tags/labels for stories should be stored alongside points. If these lablels are skill-related (mySQL, backend, frontend, AJAX, security, analytics, graphics, etc.) then they can be mined to get a sense of developer velocity by type of problem. You might in the end get a dynamically updated view of a developer with hit-point by type of problem, much like a character in a D&D role playing game.

The problem could probably be broken down to solving a system of simultaneous linear equations, or pehaps just one "production function" subject to a budget constraint. If you get the problem down to one of a single "production function" and a single "budget constraint" then you might be able to present the problem as a Lagrangian equation, apply the envelope theorem and look at a ton of cool economic views of the problem: a supply function, a demand function, market clearing prices, implicit costs. You might also be able to do hedonic (skill based) pricing and come up with an internal "wage" for individual skills to properly allocate them in your own internal economy. This is the sort of stuff that would get Hal Varian at Google excited about the possibilities of Tracker in that company, or any large organization.