Computer data analysis as an instrument in sports analytics

The level of information technologies development and means of photo and video allow to accumulate huge amounts of statistical information in various sports. Multivariate analysis methods implemented in the statistics, allow to reveal hidden patterns which can be used in making certain correct managerial decisions in preparing for sports competitions. For example, the analysis of statistical data of the Russian championship football season of 2013-2014 the possibilities of computer data analysis methods are shown to identify statistical regularities representing a certain interest in football analytics. With the help of statistical methods implemented in STATISTICA package - correlation analysis, cluster analysis, multidimensional scaling comparative analysis of team games parameters were conducted (3 parameters - Wikipedia data, 53 -company Opta data) in the championship individually for each of the 56 indicators, and the most important set of 24 indicators. In assessing the performance of teams on 56 indicators such statistics as a selective average (arithmetical mean), a sample standard deviation were used. The uniformity of teams were highlighted. Leaders' cluster: CSKA, Zenit, Lokomotiv Rostov. Mid-level cluster: Krasnodar, Spartak, Terek, Kuban, Rubin, Ural, Dynamo. Outsiders' cluster: Volga, Krylya Sovetov, Tom, Amkar, Andzhi. The analysis of the degree of similarity (difference) between the teams were conducted by estimating the distance between them as the points of the multidimensional space. A factor to assess technical and tactical training level of the teams of Premier League on the results of the games in Russian Championship of 2013-2014 was offered.