French game site Overgame has a pertinent interview of Roman Campos Oriola, a game design from Ubisoft who is working on a game for the Nintendo Wii. There are some good thoughts about the game controller and the potential interactions (I rougly translated the interesting excerpts):

- We had to reinvent control methods because there are no standards, we worked on that with Nintendo - fight with the sabers are achieved through motion detection: movements are detected and compared with a set of known movements, if there is a matching, the program trigger animations (so it's not a a real spatial positioning) - the challenge was to find the most natural movements for the players. Typically, for doors, we first though a movement (like a wrist rotation) would be ok because we are used to open doors like that. We actually noticed that nobody does the same movement do to that. And finally the most natural way to do it was to explain to the players that he simply had to push the door, and this is the movement we kept. That's where the difficult part is: trying to know what will be understood by players in terms of movements. - the problem is then to know what will be obvious for the player but there are also other issues, for instance we had to cut out the different actions: it's not possible to ask the players to perform simultaneous movement with both pads as if he were playing drums

Why do I blog this? This kind of issues are very important and empirical testing would be great to understand the grammar of interactions that would be affordable/understood by players using such tangible interfaces.