September 22, 2006

...Learn TDD with Codemanship

Product Management Game

Today I'm recovering from the excesses of last night, which involved Belgian food and Belgian beer: a lethal combination. With my remaining brain cells, which tend to be very overworked this days, I think I may have made some progress on the idea of a Nonlinear Product Management game. In the basic Agile Governance game, the value associated with each planned move is known in advance. In reality, this is rarely if ever the case. We might think we know what could bring more value, but we are never really 100% sure.

To incorporate this into the game in a very simple way, I need to make value unpredictable. But at the same time I need to retain the skill in estimating potential value. So instead of having points randomly generated in each square in the game grid, what if I randomly specify a type of dice to be thrown when landing on that square? Let's say for example we plan to land on a square that contains a dice with 6 sides, for example. I have a 1/6 chance of scoring between 1 and 6. If the dice specified has 10 sides, then I have a 1/10 chance of throwing any number between 1 and 10 (or a 3/10 chance of throwing a number greater than 6).

So instead of creating a plan that delivers the most points, we create a plan that delivers the best chance of the most points. This is more like real software development, in so much as we earn chances of value rather than directly earning value by delivering features. We "deal ourselves into the higher game" by delivering the software, just as we deal ourselves into a higher program of change by delivering software. Delivering the software in no way guarantees success in the wider context, but it buys us a shot at wider success.
Posted 15 years, 2 months ago on September 22, 2006