Theoretical Loss and Gambling Intensity (Revisited): A Response to Braverman et al. (2013)

In this paper, we provide a brief response to Braverman et al. (J Gambl Stud. doi:10.​1007/​s10899-013-9428-z, 2013b) critique of our ‘Theoretical Loss’ metric as a measure of monetary gambling intensity (Auer and Griffiths in J Gambl Stud. doi:10.​1007/​s10899-013-9376-7, 2013a; Auer et al. in Gaming Law Rev Econ 16:269–273, 2012). We argue that ‘gambling intensity’ and ‘gambling involvement’ are essentially the same construct as descriptors of monetary gambling activity. Additionally, we acknowledge that playing duration (i.e., the amount of time—as opposed to money—actually spent gambling) is clearly another important indicator of gambling involvement—something that we have consistently noted in our previous studies including our empirical studies on gambling using behavioural tracking data. Braverman and colleagues claim that the concept of Theoretical Loss is nullified when statistical analysis focuses solely on one game type as the house edge is constant across all games. In fact, they state, the correlation between total amount wagered and Theoretical Loss is perfect. Unfortunately, this is incorrect. To disprove the claim made, we demonstrate that in sports betting (i.e., a single game type), the amount wagered does not reflect monetary gambling involvement using actual payout percentage data (based on 52,500 independent bets provided to us by an online European bookmaker). After reviewing the arguments presented by Braverman and colleagues, we are still of the view that when it comes to purely monetary measures of ‘gambling intensity’, the Theoretical Loss metric is a more robust and accurate measure than other financial proxy measures such as ‘amount wagered’ (i.e., bet size) as a measure of what players are prepared to financially risk while gambling.