« Effective Scenarios - Part I: Diverse Inputs | Main | Effective Scenarios - Part II: Orthogonality »

25 July 2005


Mike Linksvayer

That was Cass Sunstein guest blogging at Lessig's.

I strongly agree that prediction markets ought to be more useful when relevant knowledge is distributed (many of the claims on tradesports are nearly worse case for PMs, not to mention plain non-interesting to me -- I want science and policy outcome bets), with one caveat, which is itself comes with several caveats: PMs provide an incentive for those holding relevant knowledge to distribute it via PM participation. The Roberts nomination may have been an example -- he became the favorite at tradesports (very) shortly before the announcement, see http://www.volokh.com/archives/archive_2005_07_17-2005_07_23.shtml#1121820182

Chris. F. Masse

Hello guys,

The world is small. Howdy?

Be sure to also read Cass Sunstein's follow up (and my PAM rebuttal as the third comment):

Markets, Prediction Markets, and OSS


Best regards,

Chris. F. Masse

Dave Scotese

PMs tend to find the result the market participants think is the most likely outcome given all the information that those participants have to process. Although those with relevant knowledge are encouraged to add it by participating themselves, the influence of their knowledge through that mechanism is limited to their willingness to bet the kind resources in which that PM trades (usually money).

When the resources traded by the PM is reputation and social collateral itself, the claims that A) the kowledge is relevant and B) the knowledge is accurate become the representation of the bet. But how do you set up such a market? Every group conversation is such a market. This blog is such a market. The next step is to allow the participants to reward or punish each other depending on their judgments of the quality of those claims A and B. My website attempts to do exactly that using the Condorcet election method. It uses a "like/dislike" judgment and a favor-the-most-recent strategy to resolve preference cycles. It creates an overall consensus-based ordering of comments.

I'm working on a more elaborate implementation in which each successive piece of a conversation thread can be composed of several competing comments.

The comments to this entry are closed.

My Photo

June 2009

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30        

Sites, People, Blogs