Want to break a world record? Don't throw a javelin

NEWS RELEASE

TILBURG UNIVERSITY

************************* Econometrists at Tilburg University calculate the ultimate athletics records

The 100 metres can be run in 9.29

TILBURG, Netherlands - In an as-yet-unpublished scientific article, two econometrists from Tilburg University (the Netherlands), Professors John Einmahl and Jan Magnus, have worked out what the ultimate records are in 14 athletics disciplines, for both men and women.

The article is entitled 'Records in athletics through extreme-value theory.'

The data they have gathered relate to the marathon, the events in the men's decathlon, the women's heptathlon, and the 10-kilometre run. In order to be able to make a pronouncement about the chances of the current best time in the marathon being improved upon, the personal best times of 1,546 male and 1,024 female athletes were analysed.

An athlete may only feature once in the list.

According to the method of approach used by Professors Magnus and Einmahl, male marathon runners can hope to better the present record by no more than 49 seconds.

Their female counterparts have more to aim at: in theory their record could be beaten by around nine minutes. For the 100 metres, the personal bests of 970 male athletes were ranked.

The fastest time is 9.77 (the world record), and the slowest, 10.30 seconds.

The researchers believe that in theory the best potential time for the men's 100 metres could be almost half a second less - 9.29 seconds.

That is a relatively much greater reduction than that of the marathon.

In the case of 578 female 100-metre runners, the personal best times varied from 10.49 (the world record) to 11.38 seconds.

Here, too, the fastest time has scope for improvement: on paper, 10.11 seconds should be possible. There are great opportunities for discus throwers as well.

The world record for men stands at 74.08 metres, but that could be stretched by almost three metres, while the maximum theoretical improvement for women is 8.20 metres. The extreme-value theory is an area of statistics that attempts to answer questions about extreme events (which by definition do not occur very often) using information on less extreme events.

The theory is normally applied in the world of finance and insurance as a means of estimating the risks associated with extreme levels of damage resulting from air disasters, burst dykes, oil tankers that have run aground, and so on, in order to calculate premiums.

This is the first time that this theory has been applied to the performances of athletes. The researchers also calculated what the best-ever athletic achievement to date has been, in the sense that the current record is more or less unbeatable on paper.

A bit like comparing long-jumping apples with high-jumping oranges, perhaps - how is the 'best' record to be defined?

It seems that javelin throwers can claim the distinction: in 2005, Osleidys Menendez threw 71.70 metres and in 1996 Jan Zelezny achieved a distance of 98.48 metres.

These records would appear to be unbeatable.

At least on paper. John Einmahl (1957) is Professor of Econometrics (particularly statistics) at Tilburg University. Specialisms: extreme-value theory, statistics, and probability. Jan Magnus (1948) is Professor in Econometrics at Tilburg University. He worked at the LSE in London and since 1996 has been associated with the Tilburg CentER for Economic Research. Specialisms: econometric theory, sensitivity analysis and sport statistics (especially tennis). 'Records in athletics through extreme-value theory' can be downloaded as a PDF document by clicking here.

*************************