The goal of this project was to design and develop an algorithm that can predict the outcomes of lottery games and to backtest the algorithm. The client was satisfied with the performance of the algorithm.

 

 Technically, I have defined probability-like functions that take inputs from a space of 10 by 6 matrices of integers and from a space of 7 by 1 matrices of integers. Some of the tools used are Python, Pandas, and Excel.

 

The lottery data was organized by US state. For each state, there was a spreadsheet with historical winning numbers and another spreadsheet with data organized as 10 by 6 matrices.

 

For each of the 10 by 6 matrix, there were a number of patterns to search for and permutations of numbers from these patterns were to be compared with the winning numbers from the other spreadsheet.

 

The algorithm I have created computed certain probability-like functions using the data from both spreadsheets and make predictions about possible winning numbers for the future. I have used Pandas to load the data as dataframes and implemented the algorithm in Python from scratch.

 

Then I ran several backtests for games from different states and calculated the success rate along with distribution of these results. The client was pleased with the results.

 

The final product was a desktop Python application that the client can run on any Windows computer.