| Article ID: | iaor20113649 |
| Volume: | 22 |
| Issue: | 2 |
| Start Page Number: | 157 |
| End Page Number: | 169 |
| Publication Date: | Jun 2011 |
| Journal: | IMA Journal of Management Mathematics |
| Authors: | Volf Petr |
A statistical model for development of track and field records is presented. The progression of the best results of each year is fitted by a non‐linear regression model with non‐constant variance. On this basis, the probability of a new record occurrence as well as the distribution of a new record improvement are derived. We also discuss the problems of record prediction and the existence of ultimate records. Our application deals with the men's 100m sprint and the men's long‐jump data.