JasoninHalifax wrote:
The problem with testing outdoors is that outdoors is a far more variable environment than indoors. You cannot control wind velocity, wind direction, temperature, humidity, ground slope, etc, etc. You are stuck at testing at whatever yaw angle is present as riding. You have variation in rider postition (however slight), road surface.
Each of those things introduces error. So if you want a true picture, then you really need to do the testing in a controlled environment, which means a wind tunnel, indoors.
My beef is really that there are no actual industry standards for conducting such testing. Everyone comes up with their own.
Actually, as per the links Andrew Coggan put up above, despite those difference you claim, you can still get very precise, and reproducible results, fairly simply.
And truth is, if racing outdoors does in fact involve conditions so variable that it masks all the wind tunnel effects, then you've proven my point completely.
Even an imperfect test would be highly, highly useful: take 5 riders who can each reliably output 200 watts for 10 minutes or so, wait for a near-ideal day with no wind, and have them ride outdoors on many bikes on a flat road . This is exactly what Dr. Coggan did in his links above (he just used himself). He included wind measurements in his studies, but tried to minimize them.
Then plot out the results, including the standard deviation to show variability. If you can't tease out a reliable advantage of one bike frame in such conditions, I'd definitely assert that you will be hard pressed to show the same advantage come race day with even more variable conditions on a variable course and without ideal wind conditions. It doesn't matter even if your wind tunnel says you should be +5 minutes per hour theoretically - if it doesn't pan out in racelike conditions, it doesn't pan out.