Wednesday, October 4, 2017

The Power of Prediction Markets
Prediction markets can be uncannily accurate — sometimes. Scientists have begun to understand why they work, and how they can fail.

Summary and Critique by Oddinigwe Onyemenem


In 2012, a group of international psychologists embarked on a project dubbed the “Reproducibility Project” in an effort to repeat dozens of psychology experiments to see which held up. One of the participants, Ann Dreber, who leads a team of behavioral economists at the Stockholm School of Economics, viewed it as an avenue to mix science with gambling and thought it would be fantastic to bet on the outcome. The team was specifically interested to see whether scientists could make good use of prediction markets: mini Wall Streets in which participants buy and sell ‘shares’ in a future event at a price that reflects their collective wisdom about the chance of the event happening. As a control, Dreber and her colleagues first asked a group of psychologists to estimate the odds of replication for each study on the project’s list. Then the researchers set up a prediction market for each study, and gave the same psychologists USD 100 apiece to invest. In 2015, the project had replicated fewer than half of the studies examined and Dreber found that her experts hadn’t done much better than chance with their individual predictions. But working collectively through the markets, they had correctly guessed the outcome 71% of the time. According to Mann, experiments of this nature depict the power of prediction markets to turn individuals’ guesses into forecasts of sometimes startling accuracy.

Mann points out that prediction markets are increasingly being used to make various kinds of forecasts such as the outcomes of sporting events and business decisions. Prediction markets advocates claim that it allows people to aggregate information without the biases that affect traditional forecasting methods such as polls or expert analysis. The application in science was shown by Dreber and her team by giving researchers a fast and low-cost way to identify potential problems with replicated studies. On the other hand, skeptics point out that prediction markets are far from being perfect. This is due to an incorrect notion that a great prediction is almost, always guaranteed when a market is set up. According to Eric Zitzewitz, an economist at Dartmouth College, it is an area of active research to determine the best designs for prediction markets and the limitations. Nevertheless, advocates of prediction markets argue that even imperfect forecasts can be beneficial. For instance, hearing there’s an 80 or 90% chance of rain can make an individual take an umbrella.

The prediction-market idea was revived by the spread of the Internet, which dramatically lowered the entry barriers for creating and participating in prediction markets. In 1988, the University of Iowa’s Tippie College of Business launched the not-for-profit Iowa Electronics Market (IEM) as a network- based teaching and research tool. Over the years, IEM has set up several markets to predict election outcomes which a 2008 study found that the its predictions across five presidential elections were more accurate than the polls 74% of the time. The success of the IEM helped to inspire the creation of dozens of other prediction markets.

The article also addresses the fact that prediction markets have also missed the mark by a long shot in several cases. In the Brexit case, the prediction market gave the odds of a stay vote as 85% on the day of the referendum, whereas the outcome was narrowly in favor to leave. Also, prediction markets were off the mark in predicting the outcome of the 2016 US presidential election, which elected Donald Trump instead of the highly-favored to win, Hillary Clinton. These sorts of instances have caused academics to probe prediction markets about why they work so well, their limits and reasons for failures? Mann points out that if prediction markets offer a way to update guesses considering new information, they will do as well or better than other forecasting methods. Prediction markets in general still need to deal with challenges such as how to limit manipulation and overcome biases.


As rightly stated in the article, prediction markets are used in a wide array of markets such as sports, politics, movie, business, technology, etc. While some fully embrace it, others remain highly skeptical due to the possibility of manipulation or infusion of biases.  Research into improving it still needs to be done to ensure, as best as possible, the integrity of the process. In the next decade, prediction markets can be a major driving force behind significantly improving decision-making as more tools are implemented.



  1. I share your concern that prediction markets in general still need to deal with challenges including manipulation and bias. The article addresses the fact that prediction markets have been inaccurate in several cases, specifically the 2016 US presidential election. Given the nature of prediction markets, traders are especially vulnerable to the bandwagon effect and following the wisdom of the crowd. With that being said, Do you think prediction markets are more or less accurate than the delphi method?

  2. I think that prediction markets can really aide in identifying large moving players in an environment. I think that they can be similar to the beginner analysts that Professor Wheaton talked about. Yes nuance would help but even at a basic level I think that they can identify major moving points thus being pretty ok at forecasting.

  3. I also agree there needs to be some nuance. What do you think would be some ways to cut the possible biases and add some nuance to the prediction markets?