premature and expensive efforts to predict analytically

[David Brin]

The Defense Department wants new computer tools to analyze mounds of unstructured text, blogs and tweets as part of a coordinated push to help military analysts predict the future and make decisions faster. The search is part of the Office of Naval Research's "Data to Decisions" program, a series of three-to-10-year initiatives that will address the volume of information that threatens to overwhelm planners in the digital age. ONR is calling for computer algorithms to predict events, fuse different forms of information and offer context on unfolding events. The office expects to spend $500,000 each year in funding. http://mashable.com/2013/01/14/twitter-pentagon-future/

This is just one of many such endeavors currently underway, the most ambitious being a European Union plan to devote billions of Euros to developing predictive analytics that sound a lot like the "psychohistory" of Isaac Asimov's Foundation Universe and its genius prognosticator Hari Seldon.  (I know because I entered Seldon's mind!  In my novel set in Isaac's cosmos, FOUNDATION'S TRIUMPH.

Alas I consider all these efforts to be premature and likely doomed.  Instead, what would be far better is to lay the foundations by doing a broad-spectrum survey and finding out who - in human civilization - happens to be right a lot!  Can you believe that there has never been a systematic effort to do that one, simple thing?  Shine light on all the race touts and stock analysts and pundits and so on who claim to have a handle on things, and just appraise and score them, so that we find out who gets it more often than anyone else... perhaps anomalously often, far above chance.

I discuss this elsewhere. And there are many aspects, e.g. that the very best might not want scrutiny applied to their methods, even though we'd all have a chance for a better society if the best methods were discovered and studied and spread around.   and so on........... http://www.davidbrin.com/predictionsregistry.html

Yes, Prediction Markets are another approach.  Fine.  But a survey of who is right, how often could shine light on almost-unknown methods.