In recent years, there have been several technological advances that have brought numerous innovations that have caused several transformations in the most diverse aspects of humanity. Currently, we are experiencing a new phase of technological revolution caused by the gain in the ability to store data as well as to transmit and process them, which makes it possible to draw valuable conclusions that helped in the decision-making process in the most diverse contexts and aspects. The world of sports could not be left out of this revolution that is currently taking place, the importance of both collecting and processing data that can generate insights that allow decision-making that can be essential in the most diverse sports aspects. Assistance in decision-making at a crucial moment in a game, assistance in the preparation of training that maximizes athletes s performance as much as possible and even injuries prevetion, among others. Among the many possibilities, one is the use of artificial intelligence algorithms to determine the best team in an NBB (Brasilian League of Basketball) Basketball match. The main idea is to be able to evaluate the performance of the teams that are facing each other in a match through a time window evaluating the composition of the players on the court, analyze variables related to the match and to each player individually and at the same time extract individual characteristics of each player. Thus, it is possible to know what would be the ideal composition of players among those available so that the team as a whole does better than the opponent in a match as a whole. Such an analysis was made more precisely for the Flamengo basketball team, so the objective was to create a system that could predict which would be the composition of available Flamengo players that would be better against an opposing team. In this manner, a data analysis procedure was carried out to select variables, to cluster players, create variables and metrics that could be useful in the analysis, and finally the development and training of two neural networks that will work together to determine what is the best composition of players. Throughout the process, several lines of analysis were tested and the procedures will be explained in more detail below.