The influence of the Olympic Games on athletic performance

There are many examples within scientific literature whereby historical trends in human athletic performance have been modelled. Many models consist of exponential decay functions which are generally considered the best way to model performance trends. However none of the current performance models consider the influence of regular global sporting events such as the Olympic Games. To account for this, a new model was proposed which utilized an existing exponential decay function with additional sine functions to mimic the effect of the Olympics. Human athletic performance form 1948 up until 2010 was characterised by collecting yearly top 25 performance statistics for standard distance running events. The new performance model was applied to performance data using a bespoke graphical user interface (GUI) created in Matlab. The GUI performed regression analysis using optimisation functions, giving the following: best fit model parameters, r2, r2, adjusted and sum of the square of the errors for the different sporting disciplines, It was found that the r2 adjusted values increased when applying the new Olympic sinusoidal functions in addition to the original exponential model, leading to the conclusion that the addition of sine functions was a valid way of modelling global sporting events within human performance trends.