Mathematical models that predict 2012 London Olympic Games swimming and athletic performances

The prediction of future athletic performance is a recurring theme in sports science. Predictive mathematical models can provide new standards which athletes can aspire, assist with talent identification, develop realistic qualifying standards for international competitions and serve to motivate athletes to achieve new objectively set goals. The aim of this research was to develop predictive models based on past Olympic performances for swimming and athletics based on regression curve estimation. Data was averaging the top six performances in the finals of each swimming and athletic event. Nonlinear equations based on data up to 1996 accurately predicted 2000, 2004 and 2008 Olympic performances. The best fit mathematical functions for men and women freestyle swim were 50m (inverse), 100m (men-compound, women-cubic), 200m (sigmoidal), 400m (cubic), 800m (women-sigmoidal) and 1500m (men-cubic) and R² from .543-.997 (p<.01). The best fits for men and women athletic events were I00m (men-inverse, women-cubic), 400m (sigmoidal), long jump (men-cubic, women-inverse) and high jump (compound and logistic) and R² from .659-.947 (p<.05). Predicted values for these preceding events at the 2012 Olympics for most swimming events are logical and attainable, whereas performances in the athletic events of women's 400m and men's high jump appear to be overestimated.