One other variable that I thought of last night that is almost impossible to account for in tests like this is human bias. Anyone who is an athlete or involved somehow with the business is going to of course have sort of bias. Let's say you were to bring in complete outsiders who know how to conduct aero tests (eg, aerospace), give them the protocol, and completely remove yourself from the equation...you're still not going to completely account for this variable. Even when we look at things objectively, we always have SOME sort of bias in our heads. Maybe we like the color of shape of a certain frame batter or the font used in a logo.
There's a reason why we do double-blinded, if not triple-blinded, trials in the pharmaceutical industry. Even then, it's sometimes difficult to remove everything. Maybe a patient or investigator likes the color of one pill over the other, even if they don't know what is in it.
I know Dan has his own standards, but one thing that would have helped in this situation is to publish your protocol ahead of time, outline all of the bike configurations, and make NO changes to the test once they've been made public (despite all of the comments you might get from others). Every test is going to have it's flaws, but making last minute changes like you are is already introducing more variables and weakening your test results.
Overall, best of luck with the test and look forward to seeing the results, wherever they are published*.
And if you really wanted to be kick-ass, you would have gone about this in a real scientific fashion and had it published in a peer-reviewed publication...;) (and I won't get into bias there)
"Most of my heroes don't appear on no stamps"
Blog =
http://extrememomentum.com|Photos =
http://wheelgoodphotos.com