[EM] another reason to avoid strategic motivations

[Abd ul-Rahman Lomax]

At 08:36 PM 12/1/2008, Terry Bouricius wrote:
>But seriously, it would be important to extend the experiment to find out
>if the respondents would ACT on that pessimistic statement that flies in
>the face of probability, or if they were making a mildly humorous
>statement as in "I forgot my umbrella, so it will probably rain."

The paper really reads like a joke. There is an ad on the radio with 
questions like, "If it's Tuesday, how many inches of rain will fall 
in Brooklyn?" The respondent immediately and confidently gives some 
answer, and the announcer always says "Correct." Then, the last 
question is something like, "If you rent a movie from [internet video 
rental company], how much will you spend on gas?"

Do people really believe that if they will get a reward from one 
outcome of a coin flip, that this will influence the results? I mean, 
people use coin flips to decide matters where they want a 
no-influence method of making a choice!

There are definitely common assumptions that are inaccurate with 
probabilities, one of the most common being the "gambler's fallacy," 
the belief that if one has had a string of bad luck, if the gambler 
perseveres, the luck is bound to change.

But this is a product of the artificiality of "gambling," with games 
designed to have invariable odds. Repetition does not improve them. 
That's quite contrary to what is normal in nature. When we repeat 
behavior seeking reward in "real life," we tend to get better at it. 
I.e., we expect our "luck" to turn. It isn't luck. It's increased 
skill, conscious or unconscious.

The point has been made about von Neumann-Morganstern utilities, 
i.e., expected value of an outcome modified by the probability of 
that outcome, is how we make decisions. We do not simply "invest" 
according to our preferences, we invest in possible outcomes that are 
also reasonably likely to succeed. Gambling and lotteries and the 
like generally take advantage of artificially fixed probabilities, 
flying in the face of our instincts.

But voting isn't gambling. It's, like politics on the other side, a 
game of skill. Happens to be very low return, though. Individual 
utilities might suggest not voting at all. But it's a collective 
ritual, and thus it isn't just one vote, it's a collective action 
that does have weight.

I'm glad I voted for Obama, even if it made no difference at all. On 
the other hand, if something had happened that day that prevented me 
from voting, I wouldn't have been crushed. It's just one vote, a 
pretty small thing. Might be worth about $20. (Which is more 
powerful, to send $20 to a candidate or party, or to vote for the 
candidate? I can say which would have been more powerful for Nader in 
2000. If he'd asked people to send some money to the Green Party in 
his name, and vote for Gore, he'd have made history. As it is, 
history will not treat him kindly. He very badly damaged the U.S. Greens.