Diplom Dissertation, Information and Knowledge Management, Hochschule Darmstadt, June 2007. Supervisor: Prof. Dr. Bernd Joers

The latest findings on experimental economics and decision making almost leave no doubt that when it comes to forecasts, we have to rethink. Forecast errors can be expensive and the search for alternatives has not been exhausted. This dissertation presents the topic of prediction markets, and thereby keeps a constant detailed scientific evidence guide.

After an introduction on the issue and its problems, an explanation of fundamental theories and methods concerning the complex subject follows. Including forecasting, fundamentals of the decision theory and decision-support systems. Consequently, the author then focuses entirely on prediction markets, their fundamentals and performance. Prediction markets are successfully expanding in the areas of culture and entertainment, politics, sports and business. Compared to traditional market research studies prediction markets convince with great potential.

A detailed elaboration of the conditions to be observed in the implementation of virtual markets is developed. The conditions are addressed to market participants, the system and its design.

Three experimental implemented test markets, which showed positive results and an outlook that prediction markets can be classified as a future trend, complete the work.