In recent years, there has been a rapidly growing interest in various aspects of sports forecasting. In part this has occurred because there is a large amount of data regarding the outcomes of sporting events and the factors that are assumed to contribute to those outcomes.
Part of the motivation has been the existence of betting markets where financial winners and losers are determined by the outcomes of those sporting events and it is, thus, possible to undertake significant research about the financial aspects of the forecasts of those events.
Yet despite the 40,000 entries in JSTOR and 3700 in Econ Lit that refer to sports, few papers have focused exclusively on the characteristics of sports forecasts. While many analyses have focused on the efficiency of betting markets, there are a large number of additional topics that can be explored further in order to enhance the sports forecasting literature. The huge number of observations makes it possible to obtain robust tests of various forecasting hypotheses. It is not necessary to base findings on laboratory experiments with a small number of observations that may not replicate real-world conditions.
For example, the types of forecast that are made differ from sport to sport. In some sports the forecast is merely intended to determine the winner of an event. If betting markets are involved, individuals predict whether/or not the favored team wins by x points (called the point spread). What are the characteristics of the various types of forecasts? How are they modelled? Are they biased? Are there differences between statistical and expert forecasts? This involves a comparison of alternative forecasting methods.
Another topic involves the procedures that are used to evaluate the forecasts. What are appropriate benchmarks for evaluating each type of forecast? Thus the field of sports forecasting is much broader than assessing betting markets and determining whether there were any betting strategies that could "beat the market", in the sense of being profitable. Even so, research in the field must encompass the widest range of topics, including papers focusing on the characteristics, operation and efficiency of sports betting markets.
CALL FOR PAPERS: The International Journal of Forecasting (IJF) will publish a special issue devoted to all aspects of sports forecasting. Examples of topics include the accuracy of the forecasts, the biases of the forecasts, models of forecasting events, betting markets and their characteristics, comparisons of alternative forecasting methods including differences between experts and models, etc. All topics related to sports forecasting will be considered for this special issue. Each submitted paper will be peer-reviewed according to the standards of the International Journal of Forecasting. The deadline for submitting papers for the special issue is DEC 31, 2008. The special issue is planned to appear at the end of 2009.
SUBMISSIONS: Papers can be submitted via the IJF online submission system at http://mc.manuscriptcentral.com/ijf. Please include a note saying that the paper is intended for the special issue.
SPECIAL CONFERENCE: A conference on gambling and prediction markets, which will include a special session on sports forecasting, will be held in association with the Southern Economic Journal and the International Journal of Forecasting, in Nottingham, England, in September of this year. Details at: http://www.ntu.ac.uk/nbs/news_and_events/events/62462.html
If you would like to attend, please send an abstract (maximum of 200 words) as soon as possible to Prof. Leighton Vaughan Williams (address below).
GUEST EDITORS: All questions should be addressed to the co-editors of this special issue: