!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: MIT on Prediction Markets

Wednesday, November 28, 2007

MIT on Prediction Markets

The MIT Center for Collective Intelligence has as one of its projects development of a comprehensive Handbook of Collective Intelligence.

The Handbook's page on prediction markets provides a helpful summary of types of prediction markets, research on their nature and reliability, and ideas for further research.

The list of benefits of prediction markets is a helpful reminder of why the topic is important. As summarized by the (typo-prone and grammar-challenged) collective authorship of the Handbook, prediction markets are valuable because:
  • Traders can correct their own biases, assuming they can see how others are voting.

  • Traders get a bigger picture that has a high signal-to-noise ratio. The aggregated prediction is a reasonably good summary statistic of many people's reading of the situation.

  • Organizations can be more agile, since agility depends partly on being able to better anticipate the future.

  • Availability of internal prices (shadow prices) leads to more precise asset allocation. For example, using a prediction market at a firm could lead to more individualized service if it informs sales staff about how much it would cost to accelerate orders to satisfy a particularly important customer.

  • Contingent contracts can aid decision making. (However, one needs to be cautious about interpreting contract prices as probabilities for contingencies, since it is easy to mistake correlation for causation.)
You can read here about what several of MIT's collective intelligence researchers — Thomas W. Malone, Alex (Sandy) Pentland, Tomaso Poggio, Drazen Prelec, and Josh Tenenbaum — have to say concerning "prediction economies," a step beyond individual prediction markets.