Too funny, considering the facts including the extremely generous threshold, and what you yourself write below about the cat-and-mouse game.
Obviously the 12% is an underestimation (how did the authors call it, minimal estimate?), just like the 44% from the survey...
No. Correct is:
Ashenden told us that athletes played a cat-and-mouse game with blood values. This means that the low values post-2007 likely do not reflect true blood doping prevalence.
That article was about, I cite, "professional endurance athletes", titled "Current markers of the Athlete Blood Passport do not flag microdose EPO doping".
Spoiler alert 1: the thresholds are too generous.
Spoiler alert 2: their "microdose" went from 10 IU/kg rhEPO twice weekly to 40 IU/kg...
Spoiler alert 3: this led, on average, to an Hb increase of 5.8%.
Comment, I cite:
"This dosage/frequency combination spanned a range shown previously to enhance maximal oxygen uptake: 20–40 IU/kg (Berglund and Ekblom 1991); 20–50 IU/kg (Russell et al. 2002); 60 IU/kg (Thomsen et al. 2007)"
As for, does EPO work and cat and mouse:
"we surmised that in an environment where rhEPO is unreservedly acknowledged to be performance enhancing and thus highly coveted, an athlete selecting a doping regimen would place greater emphasis on finding the highest dosage which evaded detection, rather than the lowest dosage which yielded a performance enhancement."
No. Any such suggestion presumes that these trends - while the values are only the tip of the iceberg - are indicative of the real extent of the problem. Anecdotes of runners playing cat and mouse were just discussed here in this thread. You can add Kiprop and Jeptoo to that list (though they got caught after a while).