ABSTRACT
Gambling markets have provided a fertile ground to study the efficient markets hypothesis. The use of prices in these markets – such as odds, point-spreads and totals – have served to test whether all available information is included in the current prices formed in these wagering markets. Given the simple nature of this market, including a clearly defined starting and end point, quick realization of returns and usual ample liquidity, the fundamental notions behind the efficient markets hypothesis are generally straightforward to test. In general, betting markets have offered support in favor of the efficient markets hypothesis, especially in the market for all games in a given sport over a long time horizon. In a market dominated by staunchly loyal fans and sometimes crazed enthusiasm, these results give a stamp of approval to the notion of efficiency and support the logic behind the wisdom of crowds. Along the way, in various subsets of betting market data, phenomenon such as the favorite-longshot bias (and its reverse) and behavioral biases such as preferences for good teams (manifested as road favorites and big home favorites) and scoring (more preferred to less) have led to some interesting questions in behavioral finance about how investors and consumers truly behave. One area of this study that is often overlooked – given the importance of the efficient markets hypothesis as it relates to fundamental theories in finance – is how prices formed in betting markets may have outside uses in economics and other areas related to business. Instead of being the sole proprietorship of finance research, point-spreads, odds and totals are now being linked to other areas of research. What was once thought of as only a study of gamblers for the ease of use of studying market efficiency is now finding new outlets where this information can be quite valuable to researchers and people working in the sports business world. These outside applications of gambling market data may not seem straightforward at first, but the usefulness of the information provided by these markets stems from what we understand about the concept of prices. Free and open markets do an incredible job of providing information. Wants and needs of consumers, desires and cost structures of producers, and expectations of all parties are amalgamated into a single number. Prices formed in financial markets, such as betting markets for sports, are assumed to include all available information at
the current moment. Prices in financial markets are expected to provide an optimal and unbiased forecast of future events; in the case of gambling markets, this future event is the outcome of the game. Even in cases where behavioral biases may result in a rejection of the efficient markets hypothesis, this too provides researchers with important information, because systematic deviations from unbiased forecasts may in fact be revealing more information about what bettors, and likely fans, enjoy about their particular sports and teams. Having some foresight into the outcome of a sporting event in terms of which team is likely to win and by how much is important information to many audiences. Obviously, fans of the teams care about this in the days and hours they spend thinking about the upcoming game for their team, and gamblers care about these prices as they attempt to place a winning wager. This information spreads well beyond these groups, however, as the teams themselves likely care a great deal about the expectations of their fans as it could easily affect ticket sales, concession sales (i.e., fans leave early in a blowout) and merchandise sales. Television networks are likely to find this information useful in deciding which games to broadcast. Large television entities such as ABC/ESPN have a number of networks where they broadcast many different games, often from the same league and sport. Having some insight into the most enticing match-ups of the day or week can be very important to their bottom line. Similarly, advertisers wanting to achieve a maximum return for their advertising dollar would like to place their ads where the most viewers will be watching. Prediction markets in the form of gambling markets provide this information and likely much more. Some lines of investigation using prices formed in gambling-related prediction markets to research a concept other than market efficiency have a substantial history, while others are just beginning to emerge. What we wish to illustrate in this chapter is some of the lines of research where these betting market prices are being used to answer other questions in economics in general and, more specifically, in sports-related businesses. To do this, we have decided to discuss past research in a few distinct areas that use odds, point-spreads, totals or other wagering-related prices as sources of useful information. One natural course of study has been in the realm of the uncertainty-ofoutcome hypothesis. If fans prefer games to have uncertain outcomes, this should be reflected in things such as attendance and television ratings. We outline and describe the research in this area along a few different lines. We first describe tests of the adequacy of betting market formed prices in being a proxy for uncertainty of outcome and its applications to the sport itself. We then outline the background and specific tests performed in past research to determine if prices formed in betting markets have any impact on attendance. We also investigate the role that wagering market prices play in relation to television ratings. Lastly, we describe some research that would fall into the “other” category, where these prices are used to investigate other areas of the sports world beyond the uncertainty-of-outcome hypothesis. Each section presents a summary of past research on these topics, with some natural overlap of categories for studies which pursued multiple conceptual goals.
