Self-exclusion practices typically refer to the possibility for gamblers to voluntarily ban themselves from playing all (a selection of) games over a predetermined period. The period of exclusion can typically be chosen by the gambler although some operators have non-negotiable self-exclusion periods. Self-exclusion in both online sites and offline venues has become an important responsible gambling practice that is widely used by socially responsible operators. Auer, Littler and Griffiths.¹ briefly overviewed self-exclusion practices in both online and land-based environments. They argued that empirical research on the effectiveness of VSE in online gambling is rare. Hayer and Mayer² investigated a sample of 256 online gamblers who self-excluded. who were then surveyed six & twelve months later. They found that VSE can have favourable psycho-social effects for the gambler. For instance, players showed a marked decrease in the willingness to gamble online shortly after they self-excluded Haefeli et al.,³ tried to predict future self-exclusion by analyzing written player correspondence with a gaming company from 150 self-excluders (compared to 150 controls). They were able to correctly predict 76.6% of future self-exclusions. based on written communications. Important indicators extracted from customer emails were increasing amount of interaction with customer services over a six-month period prior to self-exclusion, doubts about the results of games, and issues concerning account administration and financial transactions. Self-excluders were also different from controls with respect to the tonality of the email (i.e., they used threats and were more abusive in written communications). Based in previous empirical research, Gainsbury⁴ claimed that VSE programs are under-utilized by problem gamblers. In general, it is known that individuals do not seek help for problem gambling until they reach serious crisis.¹⁵

Reasons for players to self-exclude are manifold. In a study by Hayer and Meyer,² players frequently reported excluding as a preventive measure and annoyance with the operator as reasons for VSE. Furthermore, about one-fifth of self-excluders reported to be problem. gamblers (21.2%). Using the DSM-IV criteria for pathological gambling,⁶ Lischer⁷ reported that in a study of three Swiss Casinos, 29% of self-excluders were pathological gamblers, 33% were problem gamblers, and 38% were recreational gamblers. Given that many voluntary self-excluders do not exclude themselves for gambling-related problems, Lischer⁷ concluded that self-exclusion is not a good indicator of gambling-related problems. In line with these results Dragecevic and colleagues⁸ compared self-excluders with other online players and reported no differences in the (i) mean number of gambling hours per month or (ii) minutes per gambling session. Similar to Hayer and Meyer, [2] who report self-exclusion to be rather spontaneous, Dragicevic and colleagues, Dragicevic and colleagues⁸ reported that 25% of players self-excluded within one day of their registration with the online operator. This could also be due to the fact that online players can self-exclude with just a few mouse-clicks.

The aforementioned studies report that the majority of voluntary self-excluders tend to be non-problem gamblers. Additionally, the Australian Productivity Commission⁹ reported 15,000 active voluntary self-exclusions from 2002 to 2009 and that this represented only 10-20% of the population of problem gamblers. This means that in addition to most self-excluders being non-problem gamblers, that most problem gamblers are not self-excluders. This leads to the conclusion that there is little overlap between problem gambling and self-excluding.

Over the decade, analytical approaches to harm minimization have become popular. This has led to the development of various tracking tools such as PlayScan (developed by Svenska Spel), Observer (developed by 888.com), and mentor (developed by neccton). Furthermore, regulators are increasingly recognizing the importance of early risk detection via behavioral tracking systems.¹ VSE also plays an important role in this context. However, some approaches use VSE as a proxy of at risk or problem gambling. In relation to the prediction of online problematic gambling, Doctor reported¹⁰ a system designed by the company Bet Buddy with Dr. Artur Garcez (City University, UK) that identifies whether an individual's gambling pattern is already showing indicators of risk, and outlined how such detection works. Current players are regarded as problem gamblers if their behavior begins to match those of past players who requested a self-exclusion. This is a two-step approach which first identifies typical playing patterns of self-excluders. It might be gambling behavior, types of games played, time spent gambling, etc. The system then constantly searches for these patterns that are nothing else than numerical formulas in the player base. If players show such a pattern they are highlighted as problem gamblers. However, such an approach is wrong and does not detect “at-risk” or problem players at all. Using a risky driving analogy, this approach would basically mean that wearing a seat belt while driving a car is a sign and predictor of reckless driving. For obvious reasons this does not make sense.

Based on the findings from empirical research, self-exclusion is a poor proxy measure for categorizing at-risk or problem gamblers and VSE should not be used in early problem gambling detection systems.

The reasons for this are evident:

1. There is no evidence of a direct relationship between self-exclusion and problem gambling. As argued above, self-excluders are not necessarily problem gamblers and thus cannot be used for early risk detection.
2. There are various reasons for self-exclusion that have nothing to do with problem gambling. Players exclude for different reasons and one of the most salient appears to be annoyance and frustration with the operator (i.e., VSE is used as a way of venting their unhappiness with the operator). In this case, an early detection model based on self-exclusion would basically identify unhappy players and be more useful to the marketing department than to those interested in harm minimization.
3. Problem gamblers who self-exclude are already actively changing their behavior. The transtheoretical model,¹¹ argues that behavioral change follows stages from pre-contemplation to action and maintenance. One could argue that the segment of players who self-exclude because they believe their gambling to be problematic are the ones who already past the stages where assistance is usually helpful in triggering action to cease gambling. These players are making use of a harm-minimization tool. The ones actually in need of detection and intervention are the ones who have not yet reached this stage of change yet and are not thinking about changing their behavior at all. This is one more argument for the inappropriateness of self-exclusion as a proxy for problem gambling.

But what could be done to prevent the development of gambling-related problems in the first place? For the reasons outlined above, we would argue that the attempt to identify problem gambling via playing patterns that are derived from self-excluders does not assist harm minimization. Firstly, this approach does not target problem gamblers, and secondly it does not provide any insights into the prevention of such problems.

It is evident that any gambling environment should strive to minimize gambling-related harm and reduce the amount of gambling among vulnerable groups. It is furthermore known that information that is given to individuals to enable behavioral change should encourage reflection because research has shown that self-monitoring can enable behavioral change in the desired direction.^12-14 Orford ¹⁵ also states that attempts to explain such disparate gambling types from a single theoretical perspective are essentially a fool‘s errand. This also complements the notion that problem gambling is not an homogenous phenomenon and there is not a single type of problem gambler.¹⁶ This also goes in line with our belief that gambling sites have to personalize communication and offer the right player the right assistance based on their individual playing history.^17,18 Research supports this line of thinking.

Studies¹⁹ have shown that dynamic feedback in the form of pop-up messages has a positive effect on gambling behavior and gambling-related thoughts. Kim and colleagues.¹⁹ found that animation-based information enhanced the effectiveness of a pop-up message related to gambling time limits. Auer and Griffiths²⁰ found that an enhanced pop-up message (that included self-appraisal and normative feedback) which reminded players they had played 1,000 consecutive games on an online-slot machine led to significantly greater number of players ending their session than a simple pop-up message. In a real-world study of online gamblers Auer and Griffiths¹⁸ found that personalized feedback had a significant effect in reducing the time and money spent gambling.

Personalized feedback is a player-centric approach and in addition to gambling-specific research, there is evidence from many other areas that shows the beneficial effects on behavioral change. For instance, personalized messages have shown to enable behavioral change in areas such as smoking cessation,^21,22 diabetes management^23,24 and fitness activity.²⁵ Contrary to the self-exclusion oriented detection approach, we conclude that personalized feedback aims to prevent and minimize harm in the first place and is a much better approach to the prevention of problem gambling than using data from those that self-exclude from gambling.

None.

The author declares no conflict of interest.

1. Auer M, Littler A, Griffiths MD. Legal Aspects of Responsible Gaming Pre commitment and Personal Feedback Initiatives. Gaming Law Review and Economics. 2015;19(6):444–456.
2. Hayer T, Meyer G. Self–exclusion as a harm minimization strategy: Evidence for the casino sector from selected European countries. J Gambl Stud. 2011;27(4):685–700.
3. Haefeli J, Lischer S, Schwarx J. Early detection items and responsible gambling features for online gambling. International Gambling Studies. 2011;11(3):273–288.
4. Gainsbury SM. Review of self–exclusion from gambling venues as an intervention for problem gambling. J Gambl Stud. 2013;30(2):229–251.
5. Suurvali H, Hodgins DC, Cunningham JA. Motivators for resolving or seeking help for gambling problems: A review of the empirical literature. J Gambl Stud. 2010;26(1):1–33.
6. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. American Psychiatric Publishing, Washington, DC, USA; 2000.
7. Lischer S. Gambling–related problems of self–excluders in Swiss casinos. Paper presented at the 16th International Conference on Gambling & Risk Taking. Las Vegas, Neveda, USA, 2016. p. 1–56.
8. Dragicevic S, Percy C, Kudic A, et al. A descriptive analysis of demographic and behavioral data from Internet gamblers and those who self–exclude from online gambling platforms. J Gambl Stud. 2015;31(1):105–132.
9. Gary BA, Robert FA, Louise S, et al. Australian Productivity Commission. Gambling Report No. 50. Productivity Commission Inquiry Report. 2010;1(50):1–632.
10. Doctor RM. New system can predict if online gambling patterns exhibit signs of turning into addiction. Tech Times. 2015.
11. Di Clemente CC, Prochaska JO, Fairhurst SK, et al. The process of smoking cessation: an analysis of precontemplation, contemplation, and preparation stages of change. J Consult Clin Psychol. 1991;59(2):295–304.
12. Gilberts G, Agran M, Hughes C, et al. The effects of peer–reviewed self–monitoring strategies on the participation of students with severe disabilities in general education classrooms. JASH. 2001;26(1):25–36.
13. Hardeman W, Johnston M, Johnston D, et al. Applications of theory of planned behaviour in behaviour change interventions: A systematic review. Psychology and Health. 2002;17(2):123–158.
14. Schwedes U, Siebolds M, Mertes G. Meal related structured self–monitoring of blood glucose: effect on diabetes control in non–insulin–treated type 2 diabetic patients. Diabetes Care. 2002;25(11):1928–1932.
15. Orford J, Sproston K, Erens B, et al. Gambling and Problem Gambling in Britain. Hove: Brunner Routledge. 2003.
16. Griffiths MD. Adolescent gambling. London, Routledge, London, UK, 1995.
17. Auer M, Griffiths MD. Personalised feedback in the promotion of responsible gambling: A brief overview. Responsible Gambling Review. 2014;1(1):27–36.
18. Auer M, Griffiths MD. The use of personalized behavioral feedback for online gamblers: an empirical study. Frontiers in Psychology. 2015;6:1406 p.
19. Kim HS, Wohl MJ, Stewart MK, et al. Limit your time, gamble responsibly: setting a time limit (via pop–up message) on an electronic gaming machine reduces time on device. International Gambling Studies. 2014;14(2):266–278.
20. Auer M, Griffiths MD. Testing normative and self–appraisal feedback in an online slot–machine pop up in a real–world setting. Front Psychol. 2015;6:339 p.
21. Stotts AL, Groff JY, Velasquez MM, et al. Ultrasound feedback and motivational interviewing targeting smoking cessation in the second & third trimesters of pregnancy. Nicotine Tob Res. 2009;11(8):961–968.
22. Obermayer JL, Riley WT, Asif O, et al. College smoking cessation using cell phone text messaging. Journal of American College Health. 2004;53(2):71–78.
23. Cho JH, Lee HC, Lim DJ, et al. Mobile communication using a mobile phone with glucometer for glucose control in Type 2 patients with diabetes: as effective as an internet based glucose monitoring system. J Telemed Telecare. 2009;15(2):77–82.
24. Farmer A, Gibson O, Hayton P, et al. A real time, mobile phone–based telemedicine system to support young adults with type 1 diabetes. Inform Prim Care. 2005;13(1):171–177.
25. Buttussi F, Chittaro L, Nadalutti D. Bringing mobile guides and fitness activities together: a solution based on an embodied virtual trainer. Proceedings of MOBILE HCI 2006: 8th International Conference on Human–Computer Interaction with Mobile Devices and Services. New York: ACM Press; 2006. p. 29–36.