Learning agile techniques playing

I've just taken part to a very informative workshop lead by David Parson about choosing the right technique at the right time in your team. The main goal is to build a Human Powered Machine on A4 trasparent paper sheet, on which you draw upon as slices with colored pens as the technology. In the team, there is the stakeholder, who prioritizes user stories (already written, as in the XP game, with their own business values), the developers, and a tracker/scribe (of course, it's me). Moreover, a QA person who decides if the stories pass, i.e. he performs acceptance tests, will act for every team involved. There were two teams, with 4 developers and 1 stakeholder. The game lasts three iterations, 20 minutes each (more or less a pomodoro...): 5 minutes planning, 10 developing, 5 testing on the integration table. The interesting part is that you have a plateau of techniques you are (not) allowed to use. For example, if you do not choose continuous integration, you can't use the previous slices.

In my team, Peter Bell took the role of the leader at once, so that the other developers initially were shy to propose their ideas, and also pushed our stakeholder, Noura. But it was Thomas that posed the key question: "what kind of vehicle?" and the common decision was: a kind of a bike. On the contrary, Philippe was the first to start actually developing, with a lot of details, unlike the others (each should develop alone in the first iteration). The integration phase on the table was quite good, and Peter had the idea to write on the stories what each should develop. Great idea: Philippe and Peter refactored their own work, Karl choosed another user story why Thomas started again and designed an improved version 2 of his slice. Now the QA person should evaluate the demo, and everything got ok, save two stories about a luggage feature and a mirror.

Iteration two: regression testing was chosen by our team, while the other choosed continous integration. New stories A pair in each table was now allowed. Thomas and Karl exchanged a lot of ideas, while Peter and Philippe talked with the stakeholder, Noura. The QA got well! Fantastic! The team was excited and, during iteration three, they never losed the focus, i.e. to design a human-powered vehicle, unlike the other team. So they discarded absurd stories like flying. What stucked me is Philippe, who made a spike who proved to be effective, even if developers were allowed to work together for the final integration. We won the game! In any case, it was funny and informative. David told us that the materials will be available on his web page.