Here's a video of testing different tyre pressures
https://www.youtube.com/watch?v=7r8f3w89XeM
Brilliant find, Macspud. When I first came into cycling on the net, I was outrageously abused by clowns who claimed to be scientific cyclists for saying what the fellow in video now preaches as received wisdom. Back then only Sheldon and Jobst stood up for real science, which I'd brought with me from racing car development, as had Jobst Brandt.
The most interesting thing here, for me, after the rigorous test method, is the large input to the results that the presenter accords the rider, irregardless, as we used to say down under. An instance, which is a key origin point in this long-running thread, is that I ride 60mm tyres, at the time I started doing so considered by many cyclists as grotesquely wide, at pressures between 1.5 to 2.4 bar, which many considered grotesquely low ("Double the pressure and you shrink the tyre's contact patch, and get a more efficient bike!"), but which on my rough lanes curling down steep hills, gave me the roadholding and stability that made me so much faster on my touring bike than the local peloton on their road bikes riding on the highly pressured narrow tyres of received cycling wisdom. See, I knew something else, besides the scientific purity of pressure and tyre width in iterated, controlled tests: a comfortable driver/rider is less distracted, even less fearful, and therefore translates greater confidence into riding faster. The fellow in the video admits as much where he talks about professional racers on cobbled stages after he hobbles away from the cobble test. Engineers don't like this sort of talk, and pure mathematicians visibly squirmed when I explained the art of statistical market research*, but in many fields of research there is a necessary element of art beside the purely mechanistic motions of exposing the inner working of whatever, useful throughout the process but (usually)most obviously seen in the design stage of the experiment (so that, to name only one possible error, you don't test the wrong hypothesis) and then again in the interpretation of the data (so that, inter alia, you don't mistake correlation for causation).
BTW, we should commend this presenter for describing the losses resulting from wrong tyre inflation in Watts, large numbers much easier to visualize comparatively than the numerically small efficiency loss which is always normalized to a percentage, and which therefore seems tiny. However, notice something? The efficiency losses between correctly and incorrectly inflated tyres are
larger than those some cyclists obsess about between derailleur and Rohloff drive systems!
* They're more inclined to be open to the meaning when it is called "psychology" rather than "art", but it is actually art in the sense that all lateral thinking is art, even when it is about structuring Popperian research correctly.
