The Official Blog of Cogmap, the Org Chart Wiki



Are a majority of idiots smarter than a smart man?

I would need to think about this a little bit and I don’t have time unfortunately, but I have to blog my knee jerk reaction to this:

So the idea is that one thousand people voting on chess moves might be able to defeat a Grand Master.

This is an interesting case because theoretically the individuals might be able to avoid the big mistakes (?) that occasionally occur in games and maybe even see the big opportunities (!), but generally speaking, I suspect it is unlikely because one thousand people voting would probably blow the openings.  Your average person does not play chess well.  That is my hypothesis.  So your average person is going to vote for the wrong move.

Wikipedia’s example of the wisdom of crowds is guessing cattle weight.  This is a single variable decision and the bad decisions high are offset by the bad decisions low and things work out.  Chess is multi-variable, resulting in much more complex decision-making processes.  People voting to move the piece on the left aren’t offsetting people voting to move the piece on the right.  The center piece is the best piece to move, but you never get there.

Interesting idea, would love to see comments.

2 Responses to “Are a majority of idiots smarter than a smart man?”

  1. James Says:

    Nope, doesn’t work in this case.

    People can work extremely well together in large groups. One of the best books written in the past 20 years, that is still shockingly relevant is “Out of Control: The New Biology of Machines, Social Systems and the Economic World” by Kevin Kelly.* He gives some great examples in the beginning of how a large group of people can act rather surprisingly together as one. The catch is that it’s an instinctual reaction that people are able to cooperate on (in my opinion**).

    Chess is still lacking an optimum strategy. There are optimum moves at points, but no dominant strategy. Although researchers recently solved checkers, chess is still too complex. There are sets of moves that dominate, but no clear strategy to run the game.

    Regardless of which, even if an optimum strategy was apparent in chess, it requires a majority of parties involved to be aware of it. Given the mixed results of various computer opponents, I find it unlikely that a group of marginally cooperating people can truly take on a grand master.

    *I’m absurdly obsessed with emergent systems (the topic of the book). The book “The Long Tail” essentially repeated basic economics as I saw it, but will be ultimately worth it to me for having pointed out “Out of Control”.

    ** There’s a substantial set of research that demonstrates the difference between conscious reactions and instinctual reactions for people. One example being fear – in studies a person reacts to a snake, before they consciously register the snake in different parts of the brain.

  2. Jack Mayholffer Says:

    Absolutely not unless this is done in a controlled environment where the participants are engaged based on their knowledge of the game. Otherwise you might as well be asking if a studio audience could formulate a better response to an individual’s personal problems then say Dr. Phil (which would involve the unspoken “emotional IQ”).

    Individuals are smart, groups of people not so much. Just look at the Republican party, need I say more? :)