Can a market beat the market?

Marketocracy is a game where investors are given $1 million in virtual money to build a mutual fund, much like fantasy baseball. Currently some 55,000 players manage 65,000 fantasy mutual funds. But at Marketocracy fantasy becomes reality (I ought to be paid for that line) because the sponsoring firm monitors and ranks the performance of all the traders using an algorithm incorporating short and long-term analysis, market sector, risk and so forth. The stock picks of the top 100 portfolios are then used to create a real portfolio of stocks, the M100 (MOFQX). In addition to fame, the players have an incentive to trade carefully because top managers are paid a percentage of the assets in the fund!

Efficient markets theory says that the idea shouldn't work but it also says that if transactions costs are low (the fund does have relatively low costs) then it shouldn't hurt either. In fact, the M100 mutual fund has beaten the S&P 500 over the past several years and it has done so at lower risk.

Addendum: My friend Dan Klein recommended this mutual fund to me when it first appeared. I demurred based on efficient market theory and bought Webvan instead. Ugh.