Read other letters about this article

• Tuesday, December 13, 2005 07:26 AM

  Accuracy metrics

  I'm disappointed to not see any firm information on the accuracy of predictions in the Intrade market. As a previous poster noted, the market price is derived by a correlative effect based on the number of people who agree that something is/isn't likely to happen. The significant information therefore isn't trade price, or even movement in the trading price, but an overall predictive ratio.

  Intrade markets should really be evaluated very similarly to weathermen. If a weatherman (weatherperson, whatever) estimates a 50% chance of rain tomorrow, 50% of the time that should be the result. In the long term, if it only rains 40% of the days the prediction is 50%, then it's more interesting and useful to realize that the weatherman is only 80% accurate.

  Clearly the article is more intended to explore the ethical issues of Intrade markets, and I wonder how much of this is relevance vs. how much is laziness. How "dead on" are the markets?

  The last trade on AIRSTRIKE.IRAN.JUN06 this morning was at 16.5. After reading this article, I should have a good feel for whether I should rationally buy or short this contract without basing my decision on some kind of vote about how I feel about airstrikes on Iran. If anything, predictive measurements might give journalists a jump on what questions they should be asking at press briefings prior to events unfolding.

  —anon for now Read anon for now's other letters
  Permalink
  Flag
  Click "Submit Flag" if you feel this post is inappropriate. Explain why below if you wish.

  Cancel