Team Blog

Can the ultimate 100 m time be predicted?

Historical data doesn’t help much with predicting future race times. wikimedia commons

So this is quite an interesting if slightly pointless question, but nevertheless it pops up every now and again…so let’s give it a stab and see what the research says.

The 100 m has progressed from a stately 12 s in 1896 to a phenomenal 9.58 s in 2009. It is the progression of the 100 m over the last 100 plus years which have been used to predict future performance.

Researchers have added various curves to extrapolate these data points to come up with a time which will be the limit to sprinting performance. I’m not a mathmatican, therefore I’ll leave explaining the different curve fitting methods - unless any of the other team bloggers wish to comment on the most apropriate method?!

Linear extrapolation (fitting a straight line) of these trends by Tatam et al in 2004 suggested that at the 2008 Olympiad the women’s 100 m race could be won in a time of 10.57 ± 0.2 s and the men’s event in 9.73 ± 0.14 s. For male sprinters Kruper and Sterken agreed with Tatem et al and predicted that in 2008 the world record would be 9.71 s and that the ultimate limit would be 9.55 s. The predictions by Kruper and Sterken for women in 2008 were a bit wide of the mark – 10.17 s. The actual times at Beijing were 9.69 s and 10.78 s for men and women respectively. Tatam et al stated that if these trends continue, the projections would intersect at the 2156 Olympics. The time for 100 m would be 8.079 s for women and 8.098 s for men.

OK, so you’re probably way ahead of me in seeing the problem with this. You can’t fit a straight line to these data. Why? Eventually you will have a 100 m time approaching 0 s and if you continue to extrapolate further you will start to time travel with a negative time.

The other issue with comparing men to women is that the women’s world record has not been broken since Flo-Jo in 1988. Suggesting that performance has stalled/platued based on world record progression.

In 2008 Deny described Bolt’s “records as only small improvements on the existing records for the 100 m and 200 m races, 0.3% and 0.1%, respectively”. Which at the time this was indeed the case. So, based on the historical data up until that point the predicted limit for 100 m was 9.48 s (based on an average speed of 10.55 m/s).

Again, you are all probably way ahead of me here…these papers were all pre/post Beijing but more importantly all were pre-Berlin where Bolt’s improvements on the 100 and 200 m world records were no longer small but in sprinting terms – massive.

So what have we learnt from this? Predicting 100 m times from historical data are only as reliable as the day it was analysed because someone comes along and re-writes the record and rule book…and that someone was Bolt in 2009.

Next time I’ll throw in some biomechanics and explain a small aspect of how Bolt runs so fast and some mechanical limitations to speed.