First Aspect wrote:And Neeb, as someone who claims to work in biology (which is an ology after all) you have actually managed to take 2W at the outside (less than or equal to 1% in power, lets say, with error bounds considerably more and which is extremely generous as it is), but quoted data at more like 5%,which is exactly the same value in willymetres.
FA, if you look back at what I actually posted you will see that the only "data" I "quoted" was an entirely hypothetical example designed to illustrate how one might notice a consistent trend in the relationship between power and speed over multiple rides on the same route. At no point did I say that 23mm vs. 25mm tyres caused me to ride 0.5mph faster (I assume you were referring to the 19.7-20.7mph vs. 19.2-20.2mph example and are equating a +/- 2.5% average speed difference with a +/- 5% power difference at around 20mph, who knows.. I'll give you the benefit of assuming that you didn't think my example equated to a 5% speed difference!).
I've no idea what the exact aerodynamic advantage of 23mm vs. 25mm tyres is (on optimal rims or otherwise), I never claimed to. So I didn't "take" 2W from anywhere. I was describing my observations and explicitly stated
from the outset that they would not be statistically signifcant:"Yes - I have multiple average speeds and average power figures for a couple of 30-40 mile routes - not controlled conditions obviously and doubtless not enough repetitions to be statistically robust given the other significant variables (basically wind and 2 or 3 traffic lights), but strongly indicative."
To which you replied:strongly indicative eh? On a good day in the summer I'm about 5 to 7% faster on the same bike over the same route than on a bad day. How much faster are you on 23c tyres than 25's exactly, because I'm sure as hell buying a set of your 23c tyres if you can really tell the difference over that much scatter in your data
In the process not only being unnecessarily confrontational but completely inappropriately equating the variation in your own absolute speeds with what I was talking about, which was variation in the relationship between power and speed.
Obviously the extent to which such variation due to a single cause would be statistically significant would depend on its magnitude, the sample size and variation due to other factors. But even if 23mm tyres only equated to a 2W difference, if this was completely consistent it would be detectable given a large enough sample size, even if other factors caused speed from one ride to the next to vary by considerably more than a figure equating to 2W power.
But this isn't the point of course - I wasn't claiming to be demonstrating anything, I was stating what my impressions were, which I am quite entitled to do at less than 95% confidence intervals, thank-you-very-much!!
First Aspect wrote:Woodster - we know that GCN does science is fun. It is like a busman's holiday for me. What suprises ME about this forum is how intractable people are and how little traction reasoned argument makes. Perhaps that's why I descend into poking people with a stick so quickly. This too is more fun than repeatedly banging my head on the keyboard.
And why do you think reasoned argument makes so little traction on forums? Has it occured to you that it's because there is always someone like you around who is determined to polarise the debate by being confrontational, thus forcing people into defensive positions that they can't easily back down from - quite often positions that they didn't even advocate in the first place?