Posted by Jan-Eric Nilsson on September 24, 1997 at 08:47:22:
I like this experiment, but our experience is that it is tricky to drive home the point about long-term equilibrium that we want to make. The problem that we have had is that they don't intrepret losses as a signal about too many firms. I guess that the information is simply so rich so that they cannot see through what is happening. Indeed, several of them also have problems to realise the short-term optimal behaviour.
My suggestion is that - as soon as there has been a round with too many producers - you simply go in and use 'licenses', for instance saying that no more than two could do business. During further rounds this experiment is, never the less, vulnerable to 'outliers'; it is really sufficient that one or two persons don't realise the importance of number of restaurants for the whole thing to continue with losses for several rounds.