On average (and very much more than just average) there is a negative correlation between trading frequency and portfolio performance (you are not an HFT firm, don't compare yourself to one).
Set up a system like that wrong, and it doesn't matter how smart the crowd is, you're going to lose all your money. It would be interesting to play with the parameters of the game and see how results changed. (how often trades are made, max portfolio % per trade, voting mechanics, etc.)
