The model predicted well 11 out of 15 outcomes for the 2000 Olympic Games when the number of poles was equal to 5. Sensitivity analysis for higher values up to 10 (meaning a significant effect of the last 40 years to the predicted future value) gave similar results. Figures 1 to 15 show graphically the results for various track and field sports. For example Figure 1 shows the results for the men 100m since 1896. The final data point in graph A is the prediction using our model for year 2000. The results indicated that we should not expect better performance in the year 2000 compared to year 1996, and that was true. For the year 2004 it predicts worse performance than year 2000. The prediction of results in future games is always a risky attempt. The athlete`s performance is based on many important parameters, which could be unavoidable (e.g. high wind or rain during the event, health problems of the athletes, injuries, etc). This study was based on one factor only which was the previous winning results. Our main purpose was to provide the tendency of the sport in a binary mode (better - worse than previous event). The use of linear prediction may be applicable in this area which extends a lot further than the coaches and the athletes. It could be used as a source of information to the community who is interested to see the Olympic Games. We shall wait until September 2004 to confirm/reject the above findings.
© Copyright 2004 Alle Rechte vorbehalten.