Indeed there are significant differences to athletics.Note repeat offenders like Armstrong and Ulrich are only counted once in the statistics, so don't abnormally skew percentages like 38% of athletes and 65% of athletes by double counting them.This is unlike Sunday Times statistics that counted percentages of medals won or London Marathon winners, which would inflate the percentages by multiple-booking the same athlete.In distance running events, unlike cycling, athletics has nothing close to 38% "confirmed dopers".The marathon (subject of this thread) have, for men, 0 in the top 10 (actually top 40) and 2 in the top 50 for men, and for women, 2 in the top 10, and 3 in the top 50.Women's shorter distance events like 1500m do have comparable percentage of "suspected dopers", but then the relevant question for this thread is, did the women run fast due to EPO (subject of this thread), or male hormones (confounder), or both?The Sunday Times analysis of "Seppelt's" data misleadingly lumped "highly likely" together with "at the very least abnormal" values. (Ashenden and Parisotto marked the values red "likely to be doping" and yellow "suspicious requiring further investigation" respectively). I'm not sure how well "at the very least abnormal" compares to "clear link" or "strongly suspected". But we have seen that altitude training (very common among top distance runners, especially East Africans), and improper collection and analysis of blood (before 2009) can make "abnormal" false positives.In the TDF in the Armstrong years, only 9 out of 70 (13%) "unblemished" riders could break into the top 10. Compare this to Sunday Times statistic that 89% of world championship or Olympic marathon medals were won by athletes who were never "at the very least abnormal".Of course, there are a lot of apples to oranges to bananas comparisons here, so take these differences for what they are worth:- The TDF statistics looks only at men, only at endurance -- specifically a grand tour lasting 3 weeks, and includes all doping- The "leaked blood" data looks at men and women, looked only at blood levels, which may or may not be doping- The Tuebingen (alt-129 doesn't work on my Macbook) questionnaires asked men and women, from all events presumably including sprints, fields, and race-walking, about ADRVs, which can include doping, as well as banned methods, failure to get a TUE with a genuine medical need, whereabouts failures, association with banned personnel, and recreational drug use like marijuana.