This SA article might be of interest here
See How These Stocks Chosen 3 Years Ago Are Performing Today
Mar. 17, 2015 3:39 PM ET
Summary
- How being scientific can help you improve your investing skills.
- The importance of sample size when testing performance.
- Performance data of 2 crowd sourced portfolios.
http://seekingalpha.com/article/3007986-see-how-these-stocks…
Denny Schlesinger
Don’t sell stocks too quickly just because it’s hit fair value.
On the other hand, many of those poorer performers you may have been better served to do just that. (Sell when they hit fair value)
Keep the good ones.
Really, as if anyone plans to keep the bad ones. :<)
- How being scientific can help you improve your investing skills.
The science he is talking about is only in stating a hypothesis that crowd sourcing would do better than the S&P.
It looks like there were no qualifications on the stocks picked. He just asked people to say which stock they wanted, but took no reasoning into the choices.
I think his results would be different if he had a 'scientific" system for picking the stocks.
tdonb
The Grand Adventure certainly espouses scientific principles.
The goal is to handily beat the S&P returns with a stretch goal of 15% returns per year.
Adam has picked the “best” dividend stock companies in the US based on exacting criteria:
3 years increasing revenues
3 years increasing income
Above industry average (and increasing) ROE, ROI,
etc…
This gave him a pool of 170+ companies.
He now has over 30 portfolios based on this pool, using different metrics and is tracking performance.
Decisions to buy or sell are entirely mechanical. Rules are at http://discussion.fool.com/i-cant-speak-for-adam-but-my-understa…
This reminds me of the various screened ports of the early AOL-based Fool 
We’ll see in 20 years if his hypothesis is correct.
The thing is scientific in that he is testing a hypothesis. The results are inconclusive, one sample beat and the other lost. I wonder how many samples would be needed to reach a reliable verdict.
Then there are more questions:
How large should a crowd be?
How do you make sure the crowd is random?
Denny Schlesinger