Early Detection Items and Responsible Gambling Features ... · Slide 6 Identification of at-risk...
Transcript of Early Detection Items and Responsible Gambling Features ... · Slide 6 Identification of at-risk...
Lucerne
Prof. Jörg Häfeli, Projectmanagerlic. rer. soc. Suzanne Lischer, Research Associate
T direct +41 41 367 48 [email protected]
Early Detection Items and Responsible Gambling Features for Online Gambling
EASG-Conference, Vienna, 14 – 17 September 2010
Slide 2
Acknowledgments
This research was supported by a grant from the European Gaming and Betting Association. Funding bodies had no influence over design and conduct of the study, and analysis and interpretation of the data.
We would like to thank for interviews and data (in alphabetical order):bwin Interactive Entertainment AGPartyGaming PlcUnibet Group Plc
Slide 3
The specific risks of Internet Gambling
Not long after gambling was brought to the Internet, first assumptions about the addictive nature of the new medium were published.
Griffiths (1999):
• High availability (everywhere) and accessibility (24/7)• Lacking social protection (underage gambling or gambling while intoxicated)• Usage of electronic cash• Risk of fraud
These concerns are often been repeated or slightly modified manifold (Hayer & Meyer, 2004; Griffiths et al., 2006; Williams, West & Simpson, 2007; Wood & Williams, 2007).
Still there was a remarkable lack of empirical evidence: Until 2007 there is no published research, based on actual internet gambling behavior (c.f. Peller et al., 2008).
Slide 4
Actual Internet Gambling Behavior
Evaluation of actual Internet Gambling behaviour showed, that despite the speculated risks, the gambling behaviour of the vast majority is very moderate (LaBrie et al., 2007; LaBrie et al., 2008; LaPlante et al.; 2009).
As well in population based prevalence studies no increased risk for Internet gambling could be found (Welte et al., 2009; LaPlante et al., 2010).
Considering the epidemiological triangle, potential explanations for these effects could lie in protective properties only the technology of the Internet can offer for the time being:
• Pre-commitment methods (Nelson et al., 2008)• Higher transparency of losses (Productivity Commission, 2010)• Different (earlier) usage of responsible gaming tools (Meyer & Hayer, 2010)
Slide 5
RG Measures in land-based and Internet GamblingProtective measures for gamers Land-based Gambling Internet Gambling
Exclusion
Partial exclusion from single types of games not possible common practice
Self-exclusion common practice common practice
Prescribed exclusion common practice common practice
Pre-Commitment / Limitation
Limit to gaming volume not possible (except for a minority applying smart-cards) common practice
Limit to gaming time not possible (except for a minority applying smart-cards) possible
Limit to gaming frequency possible possible
Transparency
Succinct presentation of the gaming time possible (restricted to the time spent on a single EGM) common practice
Succinct presentation of the gaming volume not possible common practice
Succinct presentation of the gaming frequency possible common practice
Information offering
Awareness material and responsible gaming advice common practice common practice
Self tests possible common practice
Interactive Self-help / eHealth tools not possible possible
Contact with qualified healthcare structure common practice common practice
Under-age protection
Access limitations possible (but with many forms of land-based gambling not implemented)
common practice
Handling credit
No award of credit common practice common practice
Slide 6
Identification of at-risk and Problem Gamblers
In the light of an individualized protection of gamblers(c.f. Blasczcynski, Ladouceur & Shaffer, 2004) early detection of at-risk and problem gamblers is a central requirement for any responsible gaming framework.
Land-based Gambling Internet Gambling
Observation of gambling behavior
Typically not observable in an objective way(comprehensive usage of smart-cards may offer first approaches)
Stored and readily availiable for longitudinal analysis(Xuan & Shaffer, 2009; Braverman & Shaffer, 2010)
Observation of social interaction behavior
Availiable – quality of observations depends on standardiation of protocols and training of staff(Allcock, 2002;Schellinck & Schrans, 2004; Hafeli & Schneider, 2005)
?
Slide 7 7, 30 September 2010
Social behavior as a predictor of gambling-related problems in land-based gambling (assortment)
Allcock (2002) Schellinck &Schrans (2004)
Häfeli &Schneider (2005)
Delfabbro et al. (2007)
Repeated visits to an ATM; borrowing money on sites; trying to cash cheques; disorderly behaviour; family enquires; long sessions, etc.
Nausea;depression; headaches; gambler plays longer that 3 hours; borrowing money; shaking; feeling edgy; etc.
Frequency of visits; duration of visits; guest borrows money from other guests; level of bets per visit; guest gambles almost uninterruptedly, etc.
Gamblers plays longer than 3 hours; loose track of what is going on around them, play quickly without a proper break; favour gaming machines; etc.
Slide 8 8, 30 September 2010
Goals of the project
Considering the fact, that the lack of social interaction was typically named one of the risks of Internet gambling, one would expect the analysis thereof will not be feasible in the Internet.
However, online gambling operators communicate with their customers as well; typically via email or telephone - amounting up to 150,000 customer contacts per month per operator.
The aim of this study is to generate a basic theoretical and empirical guideline which permits the development, implementation and validation of objective protocols for early detection of gambling issues based on customer communication behavior.
Slide 9
Study I - Semi-structured interviews with senior customer-services staff
Sample:• 8 senior staff members from 3 private internet gambling operators• Interview duration between 45 and 60 minutes
Learnings:• Customer communication does contain indicators for future gambling
problems• Cannot be solely based on discrete key-words; the problem is defined by
the context
Risk Indicators identified:• Chasing losses• Financial situation / financial requests• Loss of control• Family or social situation• Heavy complaining / allegation of fraud• Criminal Activities / threats• Health issues
Slide 10
Study II – Prospective Analysis of Customer Communication
The second part of the study should be understood as a confirmatory investigation with the goal to investigate how far the indicators, identified in the previous study are able to predict manifest gambling related problems in a prospective empirical design.
Criterion Definition:As a problem criterion, gamblers were selected who excluded themselves from gambling because of gambling-related issues. Customers who closed their account for any other reasons (e.g. not satisfied) were not selected.
Sample:150 randomized self-excluders; 150 randomized controls Independent of the type of gamedue to feasibility reduced to customers communicating in English or German
All communication of both groups was analyzed:1008 mails (observer-blinded design, 2 independent raters)Inter-rater reliability: 0.78
Slide 11
Hypothetizised risk-indicators
Content-based Indicators:
• Chasing losses
• Financial problems
• Loss of control
• Social situation
• Criminal acts
• Health issues
• Doubts about results of games/bets
• Request for an increase of betting limits
• Request for lower limits
• Request for partial blocking
• Request for account reopening
• Technical problems
• Account administration
• Financial transaction
• Request for bonus
• Announcement / threat of account closure
Tonality-based Indicators:
• Complaining
• Threatening
Other Indicators:
• Frequency of customer contacts
• Immediate repeats
Slide 12
Description of the Sample
Self-Excluders
Controls
Gender 94.6% Men 91.3% Men
Age 31.5 32.9
Self-Excluders
Controls
Communi-cationavailable
52.7% 39.3%
Nr. of mails 8.3 3.3
Socio-demographics: Communication
While Self-Excluders and random controls do not differ in their socio-demographics, they do differ in their communication behavior.
Slide 13
Density of communication in relation to the date of self-exclusion
43% of all communication of self-excluders happens during the final 6 months prior to self-exclusion.
Slide 14
Descriptives: Content Predictor Frequencies
Predictor Frequency
Chasing losses 0 %
Financial problems 0 %
Loss of control 0 %
Social situation 0 %
Criminal acts 0 %
Health issues 0 %
Results of games / bets 25.5%
Increase limits 2.0%
Predictor Frequency
Lower limits 0.2 %
Partial blocking 0 %
Account reopening 6.0 %
Account administration 13.5 %
Technical problems 3.6 %
Financial transaction 34.3 %
Request for bonus 4.8 %
Threat of account closure 0.6 %
Slide 15
Descriptives: Tonality Predictor Frequencies
Predictor Frequency
Neutral 58.0%
Complaining 36.7%
Threatening 5.3%
Slide 16
Bivariate Analysis: Spearman Correlations
Predictor r
Results of games / bets 0.194**
Increase limits 0.096
Lower Limits 0.058
Account reopening 0.232**
Account administration 0.180**
Technical problems 0.150**
Financial transaction 0.252**
Request for bonus 0.037
Threat of account closure 0.130*
Predictor r
Complaining 0.268**
Threatening 0.239**
Predictor r
Age -0.081
Frequency 0.442**
Immediate repeats 0.405**
Several predictors do correlate moderately with the criterion. Especially the frequency of customer contacts seems to be an important indicator.
* significant at p=0.05
** significant at p=0.01
Slide 17
Multivariate Prediction
Optimal Prediction rate was achieved by a logistic model consisting of:
• Frequency of Customer Contact
Relative share of communication topics:
• Account reopening
• Account administration
• Financial transactions
Tonality:
• Threatening tonality
Slide 18
Multivariate Prediction
Validity of the model: R = 0.57
Classification Rate: 76.6%
Slide 19
Discussion I
Customer communication does indeed contain information about whether or not gamblers are at-risk of developing gambling related problems.
However problems manifest primarily indirectly over high emotional involvement & distress, heavy complaining and failure to cope with arising problems.
Repeated closing / reopening of the account (for any reason) & frequent administrative changes in the gambler‘s account
Harsh tonality
Extensive communication about financial issues (credit card change, not having received the withdrawal yet,...)
Slide 20
Discussion II
Analysis of customer communication is possible for online gambling too. It appears to yield powerful indicators for at-risk gambling, which is – as typical for the internet – easily accessible and open for a rule-based objective evaluation.
Nevertheless communication behavior should not be the sole source of information, but instead be combined with other objective methods of behavioral analysis.
Detection of at-risk gamblers based on customer communication is only possible for those who did communicate. However there seems to be a sub-group of at-risk gamblers that does not communicate at all. In order to detect this group, other variables (e.g. objective gambling behaviour) need to be analyzed.
Slide 21
Evaluation of practical applicability (in a realisitic sample)
Feasible Risk Compromise:
Sensitivity: 53.7%
Specificity: 91.3%
Trying to achieve higher levels of sensitivity, will inevitably lead to a higher amount of false-positive detections.
Slide 22
Evaluation of practical applicability (in a realisitic sample)
Moderate gamblers At-risk gamblers
No Communication 57.7% 2.4%
Communication 37.3% 2.6%
Total 95% 5%
Not Detected
Users without communication
57.7% moderate gamblers2.4% at-risk gamblers
Users with communication and negative detection
34.1% moderate gambler1.2% at-risk gamblers
Detected
Users with communication and positive detection
3.2% moderate gamblers1.4% at-risk gamblers
Slide 23
Model Total Evaluation
Moderate gamblers At-risk gamblers
not detected 91.8% 3.6%
detected 3.2% 1.4%
Classification Rate: 93.2 %Sensitivity: 28.0 %Specificity: 96.6 %
While being able to detect roughly one third of all potential problem gamblers solely based on the analysis of correspondence, the impact of moderate gamblers falsely assumed to be at-risk is minimal.
The model would produce a rate of 93.2% correct classifications in practical application.
Slide 24
Take-home messages for the Industry
• Alongside with the analysis of objective gambling behaviour, customer communication analysis should be used as a further powerful method of identifying at-risk gamblers.
• In order to generalize these results over different online gambling operators with different products and different customer segments, replication studies should be initiated.
• Industry should not expect to be able to identify at-risk gamblers based on manifestation of DSM-IV criteria. Instead, looking for “difficult customers“ might be a more viable approach.
• Results underline the importance of having dedicated, well-trained staff in Customer Services, handling any suspicious communication.
• Industry should elaborate standardized, objective protocols for identification and handling of at-risk gamblers based on customer communication.
Slide 25
Thank you very much.
Slide 26
Ressources
Allcock, C. (2002). Current issues related to identifying the problem gambler in the gaming venue. Australian Gaming Council: Current issues. Australian Gaming Council, Melbourne.
Blaszczynski, A.; Ladouceur, R; Shaffer, H. J. (2004). A Science-Based Framework for Responsible Gambling: The Reno Model. Journal of Gambling Studies, 20, 301-317.
Braverman, J.; & Shaffer, H. J. (2010). How do gamblers start gambling: identifying behaviour markers for high-risk internet gambling. European Journal of Public Health.
Delfabbro, P.; Osborn, A.; Nevile, M.; Skelt, L.; & McMillen, J. (2007). Identifying Problem gamblers in gambling venues. Gambling Research Australia, Melbourne.
European Commitee for Standardisation (2010). Draft CWA – Responsible Remote Gambling Measures. CEN: Brussels. http://www.cen.eu/cen/Sectors/TechnicalCommitteesWorkshops/Workshops/Pages/WS58eGambling.aspx
Griffiths, M. (1999). Gambling Technologies: Prospects for Problem Gambling. Journal of Gambling Studies, 15, 265-283.
Griffiths, M.; Parke, A.; Wood, R.; & Parke, J. (2006). Internet gambling: An overview of psychosocial impacts. UNLV Gambling Research & Review Journal, 10, 13.
Häfeli, J.; & Schneider, C. (2005). Identifikation von Problemspielern im Kasino – ein Screeninginstrument (ID-PS). Luzern.
Hayer, T.; & Meyer, G. (2004). Sportwetten im Internet – Eine Herausforderung für suchtpräventive Handlungsstrategien. Suchtmagazin, 1, 33-41.
LaBrie, R. A.; LaPlante, D. A.; Nelson, S. E.; Schumann, A.; & Shaffer, H. J. (2007). Assessing the Playing Field: A Prospective Longitudinal Study of Internet Sports Gambling Behavior. Journal of Gambling Studies, 23,347-363.
LaBrie R. A.; Kaplan, S. A.; LaPlante, D. A.; Nelson, S. E.; & Shaffer, H. J. (2008). Inside the virtual casino: A prospective longitudinal study of actual Internet casino gambling. European Journal of Public Health, 18, 410-416.
LaPlante, D. A.; Kleschinsky, J. H.; LaBrie, R. A.; Nelson, S. E.; & Shaffer, H. J. (2009). Sitting at the virtual poker table: A prospective epidemiological study of actual Internet poker gambling behavior. Computers in Human Behavior, 25, 711-717.
LaPlante, D. A.; Nelson, S. E.; LaBrie, R. A.; & Shaffer, H. J. (2009). Disordered gambling, type of gambling and gambling involvement in the British Gambling Prevalence Survey 2007. European Journal of Public Health.
Slide 27
Ressources
Meyer, G.; & Hayer, T. (2010). Die Effektivität der Spielsperre als Maßnahme des Spielerschutzes. Frankfurt: Lang.
Nelson, S. E.; LaPlante, D. A.; Peller, A. J.; Schumann, A.; LaBrie, R. A.; & Shaffer, H. J. (2008). Real limits in the virtual world: Self-limiting behavior of Internet gamblers. Journal of Gambling Studies, 24, 463-477.
Peller, A. J.; LaPlante, D. A.; & Shaffer, H. J. (2008). Parameters for Safer Gambling Behavior: Examining the Empirical Research. Journal of Gambling Studies, 24, 519-534.
Productivity Commission (2010). Gambling, Report Nr. 50. Canberra.
Schellinck, T.; & Schrans, T. (2004). Identifying Problem Gamblers at the Gambling Venue: Finding Combinations of High Confidence Indicators. Gambling Research: Journal of the National Association for Gambling Studies (Australia); 16, 8-24.
Welte, J. W.; Barnes, G. M.; Tidwell, M. C.; & Hoffman, J. H. (2009). The association of form of gambling with problem gambling among American youth. Psychology of Addictive Behavior, 23, 105-112.
Williams, R. J.; West, B. L.; & Simpson, R. I. (2007). Prevention of Problem Gambling: A Comprehensive Review of the Evidence. Reportprepared for the Ontario Problem Gambling Research Centre, Guelph, Ontario, CANADA.
Wood, R.; & Williams, R. J. (2007). Problem gambling on the Internet: Implications for Internet gambling policy in North America. New Media & Society, 9, 520-542.
Xuan, Z.; & Shaffer, H. J. (2009). How Do Gamblers End Gambling: Longitudinal Analysis of Internet Gambling Behaviors Prior to Account Closure Due to Gambling Related Problems. Journal of Gambling Studies, 25, 239-252.