Good post, thanks. I agree that the study is well done. And I am not an expert in these matters either, just an interested layperson.
I did not know that the probability requirement of flagging an abnormality had changed from 99,99% to 99%. Within the experimental design of the study, I think would assuming a 30% caught/flag rate is pushing it, but 20% would appear quite reasonable. One subject had an abnormal off-score before the rhEPO treatment began and these things happen. I'd prefer the trigger point being lower still so as to be a guideline for aggressive targeted testing, despite the potential for some collateral damage in the form of more flags. From memory this was Ashenden's criticism early on as well, though he put it more eloquently to be sure.
However, I don't think that this conclusion of a 20%-30% catch rate can be applied into the real world as such. I don't mean to claim that you suggested so, but I want to emphasise this nonetheless. Such an argument simply assumes that athletes behave as though the 99,99% trigger probability is in place in a world where the trigger rate is 99%. This would be comparing apples to oranges, surely there are no "pre 2014 athletes" in the "post 2014 world". Ie, of course the athletes will adjust and act accordingly.
Nonetheless, I will concede this is a potential improvement - that is, making it a bit more risky to pursue high yields - assuming the governance of the sport and AD actually works. This is a big assumption, though.
As for maneuvering within the confines of the passport, the authors point out, "three parameters can be adjusted with regard to how rhEPO treatment is administered—the route (intravenous or subcutaneous), the frequency of injections and the dosage used." In the study a particular combination was used: intravenous, twice a week, incrementally rising dosage.
When I used the phrase room for play in the previous posts, I meant it as a shorthand for manipulating these parameters. In practice, the trade off seems to be between a regimen of high frequency low dose (where glowtime is primary the limiting factor) and a relatively high dosage lower frequency (where blood value fluctuations are the main limiting factor).
In the study, there were four steps in ramping up the dosage. In absolute terms the steps consisted of 750IU, 1500IU, 2250IU and 3000IU, respectively, injected intravenously twice a week.
As ToreBear's citation made clear, the study states that mimicking "doping in the field" as a possible limitation. This point is made in the sense that athletes might not be able to reproduce a careful dosage regimen in the field, which is a genuine possibility. But I think this point cuts two ways. In other words, in the field regimens may well be more conservative than the one utilised in the study, but still yield benefits. In comparison to what Hamilton has described the elected dosage regimen is not necessarily very conservative.
Per Hamilton's testimony in the book and the link posted by Red Flanders, for instance, a lower dosage per injection plus a higher frequency regimen can be used instead of the one that was utilised in the study. In the link posted by Red Flanders, he indicates that a 400-500IU injected every night was closer to the actual practice. The difference in dosage and frequency is quite large. Now, 400IU four times a week would be close to the least aggressive step in the study with a 1600IU load per week; 500IU seven times a week would amount only to a bit more than half the IUs of the most aggressive step totalling at 3500IU per week.
Mind also that Hamilton described practices from the pre-passport era, which means that the athletes were careful even before the passport. This at least raises the question about the contemporary evasive measures in the altered environment. On the other hand, if we assume that improved testing has increased glowtimes, there is bound to be a limit to trading dosage for frequency too. But is it twice a week a ceiling? I would tend to doubt this.
In the end the take home message is this: given the lower trigger point post 2014 (99,99% vs 99%), athletes can be assumed to adjust and switch to a combination of the 3 parameters that is a bit more conservative than the one used in the study. Perhaps it yields a boost in the range of 5% +/-2 or so, depending on the individual. This dirty-rough-and-ready figure is but an illustration on my part. If it anywhere near the mark, the boost is still substantial. For some it can be higher.
Moreover, as Yaco (perhaps unwittingly?) pointed out, simply going to train at altitude will change the dynamics, ie. make the blood profile more messy. While the physiological benefits of altitude training are there, one could also argue that there is a bonus, which is tricking the passport and/or getting a free hallpass ("I was at altitude"). In this case, a higher dosage less frequently might be the technique of choice, as at least some of the resulting blood abnormalities can be explained away by altitude, and the lower frequency minimises glowtime.
So, in summary, I think that the dopers are, and are bound to be, ahead of the AD both in the technical sense (beating the passport & tests) and political-institutional sense (administrators appear reluctant to actually catch cheats).