The thing about useful stuff is: Wisdom of crowds

Wisdom of crowds - reality or virtuality Saturday, March 24, 2007

On Friday I met with COO of one of Europe's largest brokers, who is an old friend. During the "hellos" we reminisced about how in the 1990's it had been possible to analyse retail broking transactions and identify clusters of investors that might have appetite to participate in IPOs.

I mentioned how similar crowd "knowledge" was being generated amongst communities online now and gave him a few examples. In each case people are acting for personal gratification but their actions/behaviour is providing useful indicators when aggregated with other individuals.

PicksPal which is a sporting event prediction site. Its' users vote on amateur and professional sporting events but do so for fun rather than money. You to spend points to try and beat the odds makers - you win or lose points depending on the accuracy of your picks and your risk appetite. PicksPal does award prizes to the top performers and says they've over 100,000 registered users, albeit not that many playing regularly.

Like Betfair and Sporting Index, you can bet on a lot more than just the outcome of the game. such as who will be winning at various points in a game, or which player will score the most runs.

The value add though is that PicksPal sells "expert" picks, based on the selections of their top performing members, on which people can bet real money. They claim to have a 52% win rate for their last 25 picks, with a 63% win rate overall.

In similar fashion, Marketocracy looks to monitor the best investors and aggregate their collective knowledge to create a virtual fund. Users are allocated an initial $1 million when they sign-up and 60,000 users have joined to date.

However, Marketocracy went a stage further in November 2001 in launching a real mutual fund based on the virtual investments of its 100 most successful members (as determined by a computer ranking). The Masters 100 Index fund, has $44 million in assets and has outperformed the S&P 500 Index with an average annual return of 11.4% since its inception.

The only issue with these two system is that the users don't face any moral hazard or financial impact of participating. Consequently, users can "afford" to take greater risks than they otherwise would with their own money - it's not clear they would follow their own forecasts. In the case of Marketocracy, the output assumes that all investors have the same risk profile, when clearly that is not the case. In the "game" the experts are deemed to be those generate the highest returns. Yet the real world doesn't work like that. Certainly everyone wants to see their wealth grow but people temper that by the amount they are prepared to lose in an investment strategy.

It is for this reason that one has look a little deeper when looking at the results of "crowd knowledge" - are people really behaving as they would in the real world and will they suffer the consequences. If not, it's likely to be unreliable as a guide on which to base real world actions.

Labels: marketocracy, pickpal, wisdom of crowds