## September 04, 2009

You can follow this conversation by subscribing to the comment feed for this post.

I'm trying to remember where I heard this, but I recall that "pick a number" picks are not uniformly distributed 1-100. I think it was in a sensometrics class... I'll have to find the source and put it up.

The book you're thinking of is "Fooled By Randomness" by Nassim Taleb.

Indeed it is. Thanks, Frederik.

I believe the reference I'm thinking about is a little older than 2008. He may have found the same source that I'm thinking about, however. I'm still looking...

This joke can became really dangerous when you're in the internet. Suppose you've 1000 e-mail address of financial brooker. At day one you send to 500 person the prevision that the google stock will raise. At day 2 you send to the 500 person with the correct answer a new prevision that the google stock will raise or fall. So in 4 day you have 125 person wich have received 4 corret prevision. So know you ask them for the new prevision 10\$.....

Numberpicking is not only not evenly distributed but it's also subject to manipulation.

You can also go on when your guess of 43 was wrong because you didn't say that you were guessing.

43
...(waiting for reaction)

@zerocold: I have read about people doing such a thing, but I can't remember the book at the moment.

>> "pick a number" picks are not uniformly distributed

The "37" mentalist trick works on this principle. The digits 0, 1, 5, are 9 are deemed special (ie, nonrandom) subconciously and are selected less frequently. And when you consider that a number with repeating digits, eg, 33, might also be considered nonrandom, well there just aren't a lot of numbers left to choose from.

For an interesting take on analysts guessing, check out article from Money Magazine, in particular search for the words "Yale undergrads" and "Norwegian rat"

The comments to this entry are closed.

## Search Revolutions Blog

Got comments or suggestions for the blog editor?
Email David Smith.