Xpree’-s Mat Fogarty (responding to Jed Christiansen, even though Jed didn’-t talk to him but to Emile Servan-Schreiber —-argh, kids, today, interrupting adults’- conversations ):
We have had success combining voting to rank the ideas, then prediction markets to analyze the potential of the top ranked ideas. The phrasing in the prediction market needs to be quite specific, if we invested in idea A, how long would it take to get to market? how much would we sell in the first year? If the company does not invest in idea A, then the money bet in the market is returned to the user.
With long development cycles this can be challenging as it requires keeping the market active until ship, or for the sales estimate, one year after ship.
Of course, you could use a preference market – but this has issues of information cascades and rewarding of group think.
Here’-s the Xpree stuff which Mat is talking about.
Previous blog posts by Chris F. Masse:
- Since YooPick opened their door, Midas Oracle has been getting, daily, 2 or 3 dozens referrals from FaceBook.
- US presidential hopeful John McCain hates the Midas Oracle bloggers.
- If you have tried to contact Chris Masse thru the Midas Oracle Contact Form, I’m terribly sorry to inform you that your message was not delivered to the recipient.
- THE CFTC’s SECRET AGENDA —UNVEILED.
- “Over a ten-year period commencing on January 1, 2008, and ending on December 31, 2017, the S & P 500 will outperform a portfolio of funds of hedge funds, when performance is measured on a basis net of fees, costs and expenses.”
- Meet professor Thomas W. Malone (on the right), from the MIT’s Center for Collective Intelligence.
- Tom W. Bell rebuts the puritan and sterile petition organized by the American Enterprise Institute (which has on its payroll Paul Wolfowitz, the bright masterminder of the Iraq war).
I actually wasn’t talking to Emile. The New York Times article brought a few thoughts together that I’d been meaning to express for a while now. Namely, that prediction markets shouldn’t be used to rank in-progress or poorly-expressed ideas. Instead, they should be used to quantify a selected group of well-defined ideas.
Mat and the Xpree solution seems to be heading down the right path with this. To me what’s important about how they’re approaching it is that only a selected group of ideas move on to the forecasting stage, and ideally this is once the idea is better developed.
@Jed Christiansen: OK, thanks.
Jed, get yourself a gravatar.
@Jed Christiansen: Very interesting. Hope you’ll write other posts like that in the future.