A new measure of risk-taking in gambling

ABSTRACT

This study uses a simple metric to assess the degree of risk an individual player takes on while gambling. The player’s average potential gross win per bet placed captures both the amount of money put at risk and the volatility of returns, each of which will be correlated with risk of gambling harm. To illustrate the potential usefulness of this metric, the study analyses individual account data relating to online bets on horse racing. Analysing total stakes, the traditional indicator of risk taking, and applying ordinary least squares regression, the greatest risk takers appear to be males in their mid-fifties. Using the new metric and applying quantile regression to allow a focus on heavier bettors, the highest risk takers are shown as substantially younger, males around 30. Choice of metric and estimation method therefore matter and the approach proposed here could be used by operators to identify which online players take greatest risks.

KEYWORDS:

• Internet gambling
• risk taking
• empirical
• statistical
• horse racing

Acknowledgements

I would like to thank David Forrest, Niko Suhonen and Jani Saastamoinen for their valuable comments and suggestions during the writing process. I also want to thank Veikkaus Ltd (formerly Fintoto Ltd) for providing the data, especially Pertti Koskenniemi and Reijo Anttila. A draft of this manuscript was presented at the 16th International Conference on Gambling and Risk Taking in June 2016. This research was supported by the Emil Aaltonen Foundation, the Foundation for Economic Education under Grant 8-3795 and the Finnish Foundation for Alcohol Studies.

Conflict of interestFunding sources

This study was funded by the Emil Aaltonen Foundation, the Foundation for Economic Education (grant number: 8-3795) and the Finnish Foundation for Alcohol Studies

Competing interests

The funding sources had no influence or involvement in any aspects of the research. The author acknowledges that he has no financial interests or benefits that would arise from the direct applications of this research.

Constraints on publishing

The author acknowledges that this work is not subject to any constraints on publishing.

Notes

1. For a more detailed description of the Finnish horse race betting environment and its development, see Raento and Härmälä (2014).

2. Formal ethical approval was not required because the bettors were anonymized (each bettor has only an identification number) in the data set supplied by the operator.

3. In addition, the minimum bet in Trifecta is €0.50 whereas in other game types it is €1.

4. The zip code of a racetrack was used to determine its location. The zip codes are available on the website of the Finnish Trotting and Breeding Association (http://www.hippos.fi/).

5. We categorized a municipality as a city if a municipality had classified itself as such in 2012.

6. The turning point of spending on betting with age for males is calculated from ${\hat{\beta }}_{age}/-2{\hat{\beta }}_{age\times age}$ and for females is calculated from ${\hat{\beta }}_{age}+{\hat{\beta }}_{age\times female}/-2{\hat{\beta }}_{age\times age}$.

7. For instance, in the lowest decile (10%), a gender difference in the total stakes for a 50-year-old bettor is calculated as ${\hat{\beta }}_{female}+{\hat{\beta }}_{age\times female}\times 50$ = −1.229 + 0.017 × 50 = −0.379.

Additional information

Funding

This study was funded by the Emil Aaltonen Foundation, the Foundation for Economic Education (grant number: 8-3795) and the Finnish Foundation for Alcohol Studies.

Notes on contributors

Tuomo Kainulainen

Tuomo Kainulainen (MSocSci) is a PhD student at the University of Eastern Finland (UEF) Business School. His research interests lie in the economics of gambling markets and in behavioural economics. He is also a member of an international research group whose main objective is to study risk behaviour in gambling markets.