I think I’ve come across the worst piece of scientific (if you can call it that) research ever published.
The conclusion of the study (done at Oxford nonetheless) is that women will suprass men in the 100 yard dash at the Olympics by 2158. How do they come to this conclusion? Well, they took a linear extrapolation of olympic winning times over the last 100 years. That’s it.
Now, there are a few slight problems with this extrapolation.
1) If you accept that things will continue in a linear fashion (and accept that they have been changing in a linear fashion), eventually humans will run the speed of light. At some point in the future, the Olympics will be won by the Flash. Its has to. Just take the graph and keep extending the lines. Eventually you reach a time of zero.
2) Look at the data points. Especially look at the dots representing the last 3-5 olympics. One could just as easly conclude that the winning times have been bottoming out. If that is true, then you can’t and shouldn’t make a linear extrapolation.
3) Winning times at the olympics isn’t a very good measure of performance. Each olympics is different. Different altitude, different political circumstances, etc. Its a snapshot of performace on an arbitrary week every four years. A better measure would be the progress of world record times. They aren’t constrained to a few days every four years. If someone is injured for the olympics, they still have other chances to set a world record. How does the progression of world records look compared to the olympic winning times?
Well……I did some analysis of my own 🙂
I plotted the progression of world record times in the 100m by men and women. Before I did it, I expected to see a more dramatic change with women than I did with men. Why? The men have been doing it longer and records go back further. Also, its only been in the last few decades that as many resources have been put behind women’s track and field. Nonetheless,in the long run, the over all trend should be the same. You should see a gradual bottoming out over time. Records should be broken by smaller margins at longer intervals between records.
If you look at the data you can see several interesting things. 1) Many of the world records held by women happened during olympics. This might have been due to a lack of other competitions available to women. The olympics was it. This would skew the data for women becuase the Oxford studies only looked at olympic winning times, which correspont more to world record times for women. 2) The current women’s 100m record is currently 16 years old, held by Florencew Griffith Joyner. The men’s is 2 years old. That means the end of the trendline on the women’s chart is essentially flat. For the Oxford study to be true, the olympic winning times for women (and men) will eventually have to break the world record time.
Also, the trendlines for the world record times for men and women tell a very different story than the Oxford study. It clearly shows a bottoming out of record times, which is exactly what you would expect. (and this isn’t just true in track, its true in any human physical activity)
The women’s record has cleary dropped farther, but that has to do with the fact that it had a higher starting point. Since records have been kept, the men’s record hasn’t gone down a full second. The womens record has fallen by more than 2.
The reason why this “study” pisses me off so much is that is so totally obviously flawed that even someone in middle school should be able to see the problem with it. Human athletic progress can’t go forward in a linear fashion. Somewhere there is a limit. We can approach it, but we can never cross it. Even with drugs and genetics we will never see a 5 second 100m. Never. Not with anything that we today would call a human being.