Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Using the Random Forest Algorithm on Customer Gambling Data for Predicting Gambling Freezes in an Online Gambling Platform
Stockholm University, Faculty of Social Sciences, Department of Psychology, Clinical psychology.
Stockholm University, Faculty of Social Sciences, Department of Psychology, Clinical psychology.ORCID iD: 0000-0002-3061-501X
Stockholm University, Faculty of Social Sciences, Department of Psychology, Clinical psychology.
(English)Manuscript (preprint) (Other academic)
Abstract [en]

Background. Data on gambling behaviors routinely collected on online gambling platforms can be used to detect individuals at risk of developing or having gambling problems. As only data on gambling activity is available on gambling platforms, it is important to find a proxy measure for gambling problems. Temporarily freezing one or several gambling categories has potential to serve this purpose. Aim. To predict gambling freeze in a sample of active users of an online gambling platform one week before the freeze, based on one week of behavioral data tracked on the platform. Method. N = 105 predictors were created, covering total values, frequencies, variations, and trajectories of monetary and time-related gambling involvement, number and type of games played, point in time when gambling occurred, age, and gender. The random forest algorithm was applied to a sample of N = 2618 gamblers (of which N = 1309 freezers), with the sample divided 70/30 into a training and testing data set. Results. The accuracy of random forest applied to the testing data set was 0.615, with sensitivity of 0.543 and specificity of 0.686. The five most predictive variables were current age, age on registration date, average session length, average sum of winnings per session, and total session length. Discussion. The predictive accuracy of the algorithm in the current study was relatively low, suggesting the need for a more suitable target variable. Also, analyzing data collected during a longer period might be needed to create a tool that could be used to identify at-risk gamblers.

National Category
Applied Psychology
Research subject
Psychology
Identifiers
URN: urn:nbn:se:su:diva-176229OAI: oai:DiVA.org:su-176229DiVA, id: diva2:1372739
Available from: 2019-11-25 Created: 2019-11-25 Last updated: 2019-12-09Bibliographically approved
In thesis
1. Responsible provision of online gambling: Effects, usability and gamblers’ experiences of protective measures implemented in online gambling environments
Open this publication in new window or tab >>Responsible provision of online gambling: Effects, usability and gamblers’ experiences of protective measures implemented in online gambling environments
2020 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Problem gambling is considered a public health problem in many countries and is associated with serious financial and health-related harms for both problem gamblers and significant others. It is possible to create gambling environments that would promote sustainable gambling behaviors and prevent excessive gambling. However, research on the effectiveness of tools for responsible provision of gambling is scarce and the quality of the research is low. Also, there exists a conflict of interest between making a profit when providing gambling and protecting vulnerable customers. The general aim of the project was to study the effects, usability and gamblers’ experiences of tools for responsible provision of online gambling. Study I evaluated the effects of a prompt to set voluntary deposit-limit of optional size among 4,328 customers of an online gambling platform. During the data collection period, all customers from Finland registering an account on the gambling platform were randomized into being prompted to set a deposit-limit either 1) at-registration, 2) before their first deposit, 3) after their first deposit or 4) to an unprompted control group. Gambling intensity, measured with aggregated net loss, was tracked during 90 days after registration. No differences in gambling intensity between the intervention and control groups were found neither on the whole-group level (B (95% CI) =-0.080 (-0.229-0.069), p=.291), nor in the subgroup of the most involved gamblers (B (95% CI) =0.042(-0.359-0.442), p=.838). Study II aimed at predicting gaming freeze (as a proxy parameter for problem gambling) in online gamblers. For the sample of N=2,618 (N=1,309 freezers), a total of 105 predictors were created based on the data tracked by the gambling platform. The analysis was carried out using the machine learning method Random Forest. The predictive accuracy of the model applied to the dataset was 0.615, with a specificity of 0.686 and a sensitivity of 0.543. Study III aimed at investigating non-problem gamblers’ experiences of protective measures. A total of N=10,200 active customers of an online gambling platform were asked to rate their previous experiences of protective tools, their inclination to abandon a gambling service due to perceived overexposure to protective measures and answer questions on their symptoms of problem gambling. N=1,223 responded to the questionnaire, with the majority of the sample being moderate-risk gamblers (38.5%), followed by low-risk gamblers (26.8%), non-problem gamblers (18.9%) and problem gamblers (15.8%). In general, non-problem gamblers were not more disturbed by protective measures than other categories of gamblers. More problem gamblers have previously abandoned a gambling service due to perceived overexposure to protective measures compared to non-problem gamblers (OR(95% CI)= 7.17(3.61-14.23), p<.001). In conclusion, a prompt to set a voluntary deposit-limit of optional size did not appear to be effective in decreasing gambling intensity in online gamblers, indicating the need of evaluating alternative designs. Predicting gaming freezes in the current project resulted in a low accuracy, indicating that gaming freeze is not suitable as a proxy measurement for problem gambling and suggesting the need for collecting subjective data on symptoms of problem gambling. The results of Study III suggest that protective measures can be tested and implemented without the risk of disturbing recreational gamblers.

Place, publisher, year, edition, pages
Stockholm: Department of Psychology, Stockholm University, 2020. p. 74
Keywords
Problem gambling, responsible gambling, responsible provision of gambling, deposit limit, prediction of gambling problems, experiences of responsible gambling tools, attitudes towards responsible gambling tools, online gambling
National Category
Applied Psychology
Research subject
Psychology
Identifiers
urn:nbn:se:su:diva-176230 (URN)978-91-7797-921-0 (ISBN)978-91-7797-922-7 (ISBN)
Public defence
2020-01-13, David Magnussonsalen (U31), Frescati Hagväg 8, Stockholm, 09:30 (English)
Opponent
Supervisors
Note

At the time of the doctoral defense, the following paper was unpublished and had a status as follows: Paper 2: Manuscript.

Available from: 2019-12-18 Created: 2019-11-27 Last updated: 2020-01-28Bibliographically approved

Open Access in DiVA

No full text in DiVA

Search in DiVA

By author/editor
Ivanova, EkaterinaLindner, PhilipCarlbring, Per
By organisation
Clinical psychology
Applied Psychology

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 85 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf