game theory

BARTELBYVQF at cs.com BARTELBYVQF at cs.com
Tue Mar 19 17:27:01 PST 2002


In my non-expert opinion, game theory seems reductionist. The establishment of such a theory requires reducing the number of variables until the choices left are the ones deemed 'essential' or 'most revealing' then the results are extrapolated. The process of deciding which factors are to be subject to choice and the extrapolation are, in a relative sense, subjective, ultimately may have a material/objective basis . Could the decisions implicit in these components of the process be measured through a 'game theory' scenario?

I will stop short of deconstructing the scientific method but I will further give examples from real live games. I gather in this instance games are more analagous. In chess the first move can be only one of 18 , while in go the first move can be one of 161 and it goes from there. The number of variables in go decreases by at least 1 per player/turn more or less depending upon the style of play. In chess (and I should add that I lose at chess *every* time, Kasparov v. deep blue) as far as I can tell the number of possibilities never gets much higher than 18 and is often significantly lower. As per my previous post-machines can consistently beat humans in chess, while this is not the case with go. By defining the rules in a game theory scenario to the point that there are a small number of variables less is revealed, not more.

mcapri



More information about the lbo-talk mailing list