In Reply To:
In Reply To:
In Reply To:
Hmmm. Well, first, let's all agree that whether you could or couldn't do such a thing is not dispositive about this particular case. Second, as I've said, I didn't show all of the things I looked at or analyses I did. Just 'cuz you can't think of a way to detect alterations or manipulations doesn't mean no one can.
Here is my thought. Someone send me a PT file of a climb (it could be done on anything but if it is on the flat it is harder to do do and make it undetectable if there is substantial variation and we are dealing with a climb in this particular instance). I will then make 9 variations of the file showing 9 different average powers varying as much as 20% and see if the experts can detect which one is the original out of the 10. If randomized properly I say they will get it right about 10% of the time. If they get it right substantially more than that then they know a lot more of the nuances of this stuff than I do (they probably do anyhow, but I don't see how what I would do to "manipulate the file" could be detected)
Well, this is an interesting challenge and from a purely egotistical point of view I'm very tempted -- but, really, what would we learn? You could, after all, be quite clever and deceitful, or you could just be bad at fakery. In either event, it tells us nothing about the current case. More importantly, I might have to reveal some of the other analytical approaches I've used and I'm not too keen on doing that at the moment.
Just an update here folks, I thought it would be possible to modify a PT file in a fashion that it would be "undetectable". It may still be possible but I couldn't do it. Ken Willet sent me a PT file and I modified it 9 different ways then sent 10 files to Dr. Chung to see if he could find the "unaltered" one. First he came back and told me they were all altered. Well, they were as I didn't want the file times to give it away so one was "modified" with zero change and ten new files were produced identically within a few minutes. Despite that he was able to choose the correct one, rather easily I presume also. While there is a 10% chance he got it right by chance I suspect this is not the case. If Dr. Chung says a file is altered you can safely assume it is. If he says it is unaltered, I would even give that a greater chance of being right.
I am totally impressed. Sort of like some of the "tricks" doctors have to pick up malingering I suppose.
--------------
Frank,
An original Ironman and the Inventor of PowerCranks