Originally Posted by
asgelle
Go to the wattage group on Google and read about regression analysis or "Chung Method" then you can do the right experiment correctly. As it is, you have controlled for nothing and reached a conclusion that might have seemed obvious but is totally unsupported.
If I'm biased and will subconsciously try to ride harder on P2SL (which is kinda likely now) statistics will not help. For test to be proper now I'd have to be blindfolded and given bike at random
. The only realistic thing I could do now is to get access to special fitting bike and try to find a geometry that let's me to produce most power and then make sure that the bike I ride comes as close as possible to that geometry.