Armstronglivs wrote:
Rekrunner has argued that if doping played a part in women's marathon running then Paula's performance would not be an outlier. It is an outlier. He is therefore arguing from the fact of her performance that doping does not play a significant role in marathon running (or at least women's marathon running).
This is not an inference that can be safely drawn. We see outlier performances in other running events - from the sprints to the middle distances - where we know doping is significantly present in those events (if not necessarily in the outlier performances). The existence of outlier performances does not allow for a conclusion that doping plays an insignificant role in those events. Athlete surveys have shown that at least 1 in 3 championship athletes have doped. It is unlikely that marathon runners will be an exception.
More precisely, I argued that if it is a partial explanation of performance, it must be a small part, as opposed to a big part.
I did not argue "significant role in marathon running", to the extent it only implies "use", but significant role in explaining the fastest marathon performances, which implies "benefit".
We can safely infer that doping did not bring anyone below 2:18 (within 2.2%) in the 14 years after Paula ran 2:15:25, and did not bring many, below 2:20, nor even many below Ingrid Kristiensen's 2:21:06 (4.2%) in 1985 (before 2018, only 42 women ran faster over 33 years -- hardly an impressive demonstration of high prevalence plus high effect).
The outliers you have shown for the other events were not as great in magnitude as for the marathon.
The women's sprints were just over half as large (1.3% and 1.4% versus 2.2%).
For women's sprinting, we would also have to consider other factors unique to women's sprinting, in terms of doping and anti-doping. This still wouldn't rule out, by the same reasoning, that FloJo's superiority is explained in part by other factors than doping, despite a known large effect in events benefiting from superior strength from women taking steroids.
I'm not sure we know that doping is significantly present in the men's 1500m.
Suggesting prevalence as high as 1 in 3, combined with an assumption of a "big part of the explanation" seems to work against the likelihood of a single outlier, by almost twice as large as your next best example.
You still seem to be conflating use of doping with benefit. Suggesting "marathon runners will (unlikely) be an exception" only speaks to "use", and only clouds a discussion about alleged superior performance benefit for the fastest performances.