The customer is a large company–the biggest lottery operator in one of Eastern European countries. Its main and only type of business activity is operating lottery games.
In total, the company has introduced and ran over 30 various types of lotteries, including combined, draw-based, and instant games.
A bit of statistics on just one of the largest lotteries operated by the company:
- 510 drawings (held on a weekly basis)
- More than 50 million dollars played, including more than 80 apartments, over 8 kg of gold, 261 diamonds, etc.
Due to an unstable prize fund distribution algorithm used by the lottery operator, the remainder of the prize fund for each game would vary in size dramatically.
It was necessary to make sure that the size of the prize fund remainder stayed within certain limits. Thus, the main need was to study and redesign the lottery drawing algorithms.
The algorithm was studied under various sets of parameters. The research showed that it is impossible to achieve a good result by selecting parameters.
We developed new algorithms for lottery ticket generation and prize distribution, which met the customer’s requirements. While working on the new algorithms, we used combinatorial methods and the classical probability theory to generate lottery tickets as well as statistically analyze and forecast the results of lottery drawing.
After the updated algorithms had been implemented, we were able to increase prize amounts by reducing the size of prize fund remainder. As the result, the lottery became more attractive and the turnover increased considerably.
Contact us to request your free 1-hour consultation with one of Altoros’ machine learning experts.
Artificial Intelligence practice head