Virtual markets for prediction have been around a while - HP was using an internal sweep-stake to forecast printer demand in the early 1990s, and found that it was much better than relying on sales peoples' hunches.
Inkling allows anyone to set up a virtual market in a few easy steps. You define your question and wait for others to start trading their $inkles - or find a question you have an opinion on and do some trading yourself.
The theory of predictive markets follows financial markets theory, in which financial instruments of different types will - in an ideal world, if not always in practice - reflect in them all the pertinent information available about the future value of an asset. In the case of an oil future, for example, the price combines every traders' dergrees of beliefs about the complex of events that influence the oil price - world demand, terrorism, Middle East politics, environmental regulation, etc. If I am an oil trader who believes that world demand is about to collapse, and believes that others do not believe this yet, I will take on obligations to deliver oil in the future up to the limit of my preference for risky living (or the limit of my wallet, which might kick in first).
Markets like the one you can set up on Inkling work in the same way, except ... 1. you play for play $s, and 2. there aren't a huge number of traders aggregating their beliefs. Nevertheless, it is quite engrossing.
It seemed to me that Blair was being given too high a chance of still being in office in June 2007, so I have been selling that contract heavily. Brown was being given too low a chance of becoming next Prime Minister, and I have been buying. Al Gore is slated as favourite for Democrat nomination, just ahead of Clinton - but the opinions are still pretty bunched here. I haven't traded.
To explore the possibilities of the system, I read George Schopflin here on oD on the Hungarian crisis and formulated the - admittedly not blindingly original, but at least suggested by Schopfln's very equivocal ending scenarios - contract that pays out $100inkles if Gyurcsány is still in power at Christmas, and the opposite contract that pays out $100inkles if he has been ousted by Christmas. I priced the original contracts at $50inkles each, indicating that I, at least, thought the question finely balanced. I am glad to see that there has been some trading activity on my contract, and Gyurcsány stays put is currently trading at $54inkles, indicating some degree of belief being expressed in the inkling community that he will stay on rather than go.
Inkling is a practical demonstration of the way in whcih financial markets are sophisticated polls, in which strength of belief, size of wallet and appetite for risk all factor in to the poll result. Maybe I should start running a book on the outcome of the Bad Democracy Poll ...
