A new approach to decision making in basketball - BBFBR program

Abstract

The developed solution is named BBFBR, which stands for Basketball Board for Basketball Referees. The current implementation of the solution is based on drawing the ball's movement on the court, which is the input vector of the neural network, whilst the output vector of the neural network consists of the movement coordinates of the referees. The key segment of this paper describes, in detail, the structure of the input and output vectors of the neural network used in the BBFBR project, as well as the methods, and the advantages and flaws that where noted during the training of the neural network. There are two methods used while training the neural network: the method of sequential repetition. The current solution enables calculating optimal ways of the referee's movement in the case that the movement of the ball in an action consists of not more than 15 key points, e.g. guiding the ball from one point to another, passing the ball to another player, shooting etc. The application value of the current solution bears only educational purpose because it is possible to apply it to training young basketball referees in the terms of their movement, to enable them to be aware of an action and to be able to analyze it. This paper also describes the methods and the developed software which, based on the action and the movement of the ball, using the neural network, determines the movement of a basketball referee on the court, in order to gain the best view of the action. The solution is developed in Microsoft Visual Studio 2010, written in the programming language C# referring to AForge.NET Framework for the support in the aspects of configuring, training and usage of neural networks. AForge.NET Framework is published with LGPL v3 licence.