Four in ten Ghanaians are prone to gambling addiction, fueled mainly by the hope to alleviate poverty. In particular, sports betting has been recorded as the predominant gambling activity. Due to the dramatic increase in the availability of smartphones in the Ghanaian communities (over 50% of Ghanaians have a smartphone), we proposed the development of a mobile app, Chancer. This will be a platform that incorporates a machine learning (ML) model to increase the odds of winning (above the chance 33%) and allow gamblers to manage their bets to prevent their financial betting from reaching an untenable state.

Chancer uses enhanced football analytics and the ML algorithm to increase the success rate of the gamblers’ bets. The model predicts the outcome of a game, either a home win, an away win, or a draw. We compared several MLs and identified Naïve Bayes as the best model with an overall accuracy of 51.6%. Despite not having the highest accuracy, it showed excellent performance in draw predictions, approximately 30%, a significant increase over the 15% precision in most models. The model ultimately gives the bettor a better chance of winning but still allows the betting agencies to profit, thus not alienating them or banning the application.

Furthermore, research has demonstrated that mobile apps can address many health behaviors. For example, advertisements that incorporate behavioral change techniques of persuasion to end addiction found success and were associated with a lower probability of drug abuse. Thus, including behavioral change techniques such as visual persuasion and self-monitoring of bets spending can curb gambling addiction. We are conducting qualitative research to determine the most effective methods in the Ghanaian context. 

In conclusion, Chancer will support the gamblers in making better-informed betting decisions preventing financial ruin while encouraging behavior changes and, eventually, helping them stop gambling.