A similar analysis for the Men’s 200m shows that
- The asymptote is 18.8 seconds, with an exponential time-scale of almost 37 years, meaning in 37 years it will fall to 37% (1-e^(-1), just a coincidence here) of the current gap 0.38s = 19.19s–18.81s.
- Pietro Mennea’s 1979 record of 19.72 s was “on schedule”, though it eventually stood for 17 years. In fact it was Tommie Smith’s 1968 WR of 19.83 s which was about 8 years ahead of its time (Possibly due to altitude?).
- When Mennea’s WR was broken, neither Michael Johnson’s then astonishing 19.32 s in 1996 nor Usain Bolt’s current WR were too far from the predicted values, and so were NOT “ahead of their times”.
- For the Women’s 200m (ignoring times simply equaling the WR by Heike Drechsler), a 3-parameter exponential fit seems overkill, a linear fit serves just as well (which seemingly doesn’t make sense, but it does imply that women’s WRs are still in the linear “early era … as training, technique, equipment and overall physiques improve steadily”).
- Griffth-Joyner’s WR (21.34s), though long-lasting, is as predicted by the linear fit. The current linearly predicted value, almost 3 decades after she set it, is close to 19.7. What explains this longevity and the large gap from expectations? Most likely that Women’s WRs in the 200m have already transitioned to the “incremental” phase, and that all predictions are nonsense until we get more data.
- Women’s 100m (@Benjamin T. Awesome): Again, for the data we have, a linear fit is quite enough (the parameters of the exponential fit are a bit nonsensical).
- FloJo’s 10.49 s from 1988 was about a decade ahead of its time and should have been surpassed by about 2000. The linear model would suggest a predicted current WR of 10.1 s. Again, we seem to have entered the “incremental phase”.