The predictions of the results in a race walking with the use of predictive models

Abstract

Aim. The issue of forecasting records in measurable sports
disciplines has a very long tradition. For this purpose, methods
known in mathematical statistics and econometrics are used. To
date, approximation theory has not been used to predict the future
course of the sports careers among athletes reaching phases of
relative stabilization in their results. In this study, we presented our
own proposal for adapting predictive models in solving the signalled
problem. Basic procedures. Time series of the best results obtained
during the 21-year sporting career of the three-time Olympic
participant were analyzed. Using the method of least squares for
approximation of the results obtained up until the end of the
observation, based on the developed curve (parabola) and the
nonlinear 2nd grade model (y = ax2+bx+c), we estimated further
prospects for the development of sports championship in racewalking
for 20 and 50 km distances. Research results.
Characteristics of the sports biography provided valuable results to
understand the development trends of the contemporary model of a
champion in professional sports and to develop training and
recruitment concepts for future sports champions. Conclusions.
Predictive models should be used both for forecasting the
development of sports disciplines and planning the development of
careers of players reaching a phase of relative stability in sports
performance.