Scientific Director, Observatoire des Jeux (France)

Jean-Michel Costes is a Sociologist and Demographer, Scientific Director for hospital statistics in the Ministry of Health in 1986, leader for the Délégation générale  à la luttre contre la drogue et la toxicomanie in 1993, and director of Observatoire français des drogues et des toxicomanies from 1995 to 2011, Jean-Michel Costes is currently the director of Studies for Observatoire des jeux whose mission is to study gambling and its social impacts.

In his capacities, he designs and conducts a vast number of quantitative and qualitative studies of general and specific populations, notably the first nationwide survey on gambling.  He also contributed to the designing of the conceptual framework used to evaluate public policies concerning addiction as well as participating in many collective expertises in that domain. 

Abstract
Online gambling currently concerns 4% of the French population.  The use of internet represents an additional risk to gamblers. Online pathological gambling represents an important public health issue.

Online gambling on websites regulated by the Autorité française de regulation du jeu en ligne(ARJEL) offers the advantage of being continuously under observations. The activity data are collected and stored by the ARJEL for each individual player. Therefore, it seems appropriate to study the data capacity in predicting problematic behaviour of online gamblers. The development of a prediction model for problem gambling based on the analysis of data registered daily offers relevant possible answers on two aspects:

(1)   An individualised action, oriented towards a specialized support, for detected as well as
       pathological gamblers;

(2)   Measures of prevention or reduction of risks for detected and problematic gamblers.

The Observatoire des Jeux and ARJEL recently launched a study to design such a model. It will use statistical analysis of the “behavioral markers” that can be obtained from regulation data, in order to determine their ability to predict the diagnostic made by the detection tool choose for this model, the CPGI index, and to analyse and maximise the specifics and sensitivity of the theoretical model.  The CPGI test will be given to a sample of at least 10,000 online gamblers for whom we have the online activity data.

We will discuss methodological, operational, and political issues surrounding the realisation of this study, as well as its challenges and limitations.