Identifying high-risk online gamblers: a comparison of data mining procedures

Abstract

Using play data from a sample of virtual live action sports betting gamblers, this study evaluates a set of classification and regression algorithms to determine which techniques are more effective in identifying probable disordered gamblers. This study identifies a clear need for validating results using players not appearing in the original sample, as even methods that use in-sample cross-validation can show substantial differences in performance from one data set to another. Many methods are found to be quite accurate in correctly identifying player types in training data, but perform poorly when used on new samples. Artificial neural networks appear to be the most reliable classification method overall, but still fail to identify a large group of likely problem gamblers. Bet intensity, variability, frequency and trajectory, as well as age and gender are noted to be insufficient variables to classify probable disordered gamblers with arbitrarily reasonable accuracy.

Predicting online gambling self-exclusion: an analysis of the performance of supervised machine learning models

“…Three of these are used by Philander (2014); Bayesian networks is an additional method reviewed in this article. We add Bayesian methods as the ability to unpick the conditional probabilities linking each variable to self-exclusion is potentially helpful from an interpretation/communication perspective.…”

“…As Gainsbury (2011) and Philander (2014) note, the data collection made possible in account-based Internet gambling revolutionizes the kind of analysis of gambling behaviour that is possible and opens up new ways to identify early warning signs of potentially harmful behaviour. However, the quantity of data simultaneously opens up questions of ABSTRACT As gambling operators become increasingly sophisticated in their analysis of individual gambling behaviour, this study evaluates the potential for using machine learning techniques to identify individuals who used self-exclusion tools out of a sample of 845 online gamblers, based on analysing trends in their gambling behaviour.…”

“…In this article, we build on Philander (2014) and Dragičević, Percy, Kudic, and Parker (2013) to explore data analysis for a particular type of risk identification, that of self-exclusion. Through gambling platforms that permit individuals to self-exclude, it is intended that individuals might recognize that they are at risk of losing control during gambling sessions and instruct the gambling platform to deactivate or block their account for a certain period of time.…”

“…(3) To demonstrate accuracy performance on test data-sets in the 72-87% range, significantly higher than the 62-67% range in Philander (2014), pointing towards a set of additional gambling behaviour variables that are worth assessing further.…”

Telecommunications Network Measurements of Online Gambling Behavior in Switzerland: A Feasibility Study

“…With the emergence of online-gambling and data-mining procedures, it is now possible to capture, analyze, and process large data sets on gambling behavior (e.g., gambling time, gambling frequency) to extract reliable and objective information [10,11]. A series of studies, for example, has analyzed data sets of gambling accounts derived from bwin Interactive Entertainment AG (bwin), a large European Internet gambling operator [12,13,14].…”