Pareto distributions in online casino gambling: Sensitivity to timeframe and associations with self exclusion.

The Pareto effect (also known as the 80/20 rule) describes a skewed distribution of engagement that is observed for many products. In this study, we investigated Pareto estimates for online casino gambling, and tested their association with voluntary self-exclusion (VSE) as a marker of gambling harm, and examined their sensitivity to varying time windows. We used one year of betting data from the eCasino section of a provincially-run gambling website in British Columbia, Canada. The data contained 30,920 account holders who placed at least one bet on the platform from October 2014 to September 2015. The top 20% most engaged gamblers accounted for 92% (based on total number of bets) and 90% (based on net losses) of eCasino gambling activity over the year. The top 20% of online gamblers displayed higher levels of VSE enrolment than the remaining 80% (total bets: 13% vs 7%; net losses: 16% vs 6%, respectively). Pareto estimates increased with longer time windows, from one month to one year (total bets: 80% to 92%; net losses: 81% to 90%, respectively). This accumulation was driven by the relative loyalty of the most engaged gamblers, coupled with the influx of new and more transient gamblers on a month-to-month basis. These data strengthen links between concentrations of engagement with online products and measures of harm, but also highlight the dynamic nature of these estimates. One year estimates are preferable for estimating the degree of concentration.