this problem was designed to serve as an abstract representation of some of the very real decisions we all face in our lives, ones that force a choice between the stability of the status quo (and relatively little gain) and the uncertainty of a new situation (with higher overall potential, but also containing an element of risk that the choice will in time prove to be a bad one with a highly negative outcome)...
the problem as i presented it is flawed, however... it seems to me that there are some important variable factors that i stated as constants, and the values i picked skew the problem...
Relative Risk:so, given these variables, how then are we to determine an optimal strategy? what is the most rational way to evaluate one of these situations when we are presented with them, and make a good decision?
i chose "75%" (3 in 4) for the risk factor in my hypothetical situation... and i think this reflects my overall pessimism, and moves the thought experiment from a generalized and widely applicable one to a more specific and less useful one... what i SHOULD have said is "an indeterminant, non-zero chance" that things will turn to shit on you if you choose "box b"...
Relative Return:
the second variable that affects the decision-making process on this problem is "just how much better is the potential upside of box b compared to box a"... clearly if the difference were very small it would be insanity to choose B over A... and conversely if the difference was very great, it would be equally foolish to choose A over B, because you stood to lose very little by choosing B and having it blow up on you.
Participant's Preference:
this is the big one, really... people are almost never entirely rational when making decisions, and their preference for one thing over another will likely have the strongest impact on the final decision... it is a sort of "multiplier" that weights the two options... ie, if the participant STRONGLY prefers stability, that preference would proportionately skew the decision towards A, *even if* the relative risk were very small, or the relative return very large... or on the other hand, if the participant STRONGLY desired whatever the potential upside of B entailed, they may well choose B even if the relative return was miniscule over A, and failure was almost assured... in my mind this kind of behavior is absolutely nuts, but i see it all the time, and its effect cant be discounted when considering the problem...
mmmm game theory :)