Comment author:Houshalter
07 July 2015 03:27:45AM
*
5 points
[-]
That's not really true. You need to make correct predictions many times on many different things to get significant returns from your initial seed investment. There are also a lot of details like finding someone to take your short, and getting it at just the right time before the market crashes.
The big surprise has been the support for the unabashedly elitist “super-forecaster” hypothesis. The top 2% of forecasters in Year 1 showed that there is more than luck at play. If it were just luck, the “supers” would regress to the mean: yesterday’s champs would be today’s chumps. But they actually got better. When we randomly assigned “supers” into elite teams, they blew the lid off IARPA’s performance goals. They beat the unweighted average (wisdom-of-overall-crowd) by 65%; beat the best algorithms of four competitor institutions by 35-60%; and beat two prediction markets by 20-35%.
There is also this guy. I remember him from the book Automate This, that he was supposedly able to predict exactly how Iran would act back when they were developing nuclear weapons. His method, as best I understand it, is to simply list everyone involved that has influence, and predict they will do exactly what benefits them individually the most.
Comment author:sohois
11 July 2015 07:48:36AM
1 point
[-]
Perhaps I'm misunderstanding the study that was performed, but from the articles it seems that this study has only been going on for 3 years now? In which case, any one sitting at the top of the heap is still pretty likely to have gotten there largely through luck. With a large sample size it's entirely possible for at least a couple of people to 'beat the odds' and get a number of questions correct again and again, without necessarily being any better than those who did poorly.
Even with a fairly significant number of questions being asked and rated, it does not appear to be a long enough study to start suggesting those at the top have better skills as opposed to better luck.
= 27d567bc2d48d9f40e9ccc20ea370b15)
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d)
That's not really true. You need to make correct predictions many times on many different things to get significant returns from your initial seed investment. There are also a lot of details like finding someone to take your short, and getting it at just the right time before the market crashes.
There also is such a methodology to make really good predictions. Train a bunch of people through practice to be good at forecasting the future and learn from their mistakes, and take the top performers from that group: http://www.economist.com/news/21589145-how-sort-best-rest-whos-good-forecasts
They even beat CIA analysts: http://www.npr.org/sections/parallels/2014/04/02/297839429/-so-you-think-youre-smarter-than-a-cia-agent
There is also this guy. I remember him from the book Automate This, that he was supposedly able to predict exactly how Iran would act back when they were developing nuclear weapons. His method, as best I understand it, is to simply list everyone involved that has influence, and predict they will do exactly what benefits them individually the most.
Perhaps I'm misunderstanding the study that was performed, but from the articles it seems that this study has only been going on for 3 years now? In which case, any one sitting at the top of the heap is still pretty likely to have gotten there largely through luck. With a large sample size it's entirely possible for at least a couple of people to 'beat the odds' and get a number of questions correct again and again, without necessarily being any better than those who did poorly.
Even with a fairly significant number of questions being asked and rated, it does not appear to be a long enough study to start suggesting those at the top have better skills as opposed to better luck.