[extropy-chat] silent night

[Robin Hanson]

On 12/22/2004, Harvey Newstrom wrote:
>>>My experience with future prediction in general is that it is wrong
>>>more often than not.
>>
>>I predict the sun will come up tomorrow.  I will keep making this
>>prediction everyday for the next year.  Go ahead, collect stats on how
>>well I do.  Betcha I'm right more often than not.
>
>If you think that is really a good counter-example to my complaints, I
>must not be making myself very clear.  I am not looking for obvious
>predictions that anybody can make, or predictions that add no new
>information above what everyone knows or assumes.  What I want is new
>information, or predictive value of something that is not obvious.
>This would be a useful prediction.  The history of such "useful"
>predictions of the future are more often wrong than not.

Make up your mind what standard you want to impose.  If all you want is
new information, then all you need is a source that predicts slightly
better than other sources.  It need not be very accurate, just more
accurate than the alternatives.  The cites I mentioned show that these
markets meet this standard.

>I meant models that can predict individual stocks ... I did not mean
>to say that there are no models in economic theory or of the market.

What you said before is:
>Someone needs to set up a working model that produces verifiable results.
>Then people will pay attention.

The whole point of having the best possible estimate is that you can't
predict future values better with anything else.  If you had a model
that predicted stock prices better than the current price, that would
mean the current price wasn't the best estimate.  Such a model would
go against my core suggestion of using prices as our best available
estimates.


Robin Hanson  rhanson at gmu.edu  http://hanson.gmu.edu
Assistant Professor of Economics, George Mason University
MSN 1D3, Carow Hall, Fairfax VA 22030-4444
703-993-2326  FAX: 703-993-2323