Dynamics of variation of sports performance in light of Time Series based on Artificial Neural Networks in swimming.

Background: General availability of sports results, such as current world records, world rankings, and results of the Olympic Games, offers opportunities for the analysis variation in many different competitions and sports. Current trends in the progress in sports performance have been analysed based on e.g. freestyle swimming. The focus of this study is on the analysis of variability of sports results in swimming achieved by women and men in 11 Olympic Games in 1972-2012. Material and methods: The analysis was based on the results of top eight finalists in all four events. Four 100m sprinting events (men's and women's) included in the program of the Olympic Games (freestyle, backstroke, breaststroke and butterfly swimming) were analysed. Results: The analyses showed that a statistically significant difference in women's real and model results was found for the most recent Olympics in Rio de Janeiro in 2016 for the 100m breaststroke swimming. Conclusions: The predicted results suggest that during the next Olympic Games in Tokyo, dynamics of progress in women's results is likely to be faster compared to men in three discussed events: 100m breaststroke, 100m butterfly and 100m backstroke. The above trend may not be observed in these events. Therefore, future research studies should be aimed to verify this tendency and the dynamics of progress in the results in breaststroke, backstroke and butterfly stroke. [ABSTRACT FROM AUTHOR]