Why is there so much difference between the TR bell curve and the Coggan chart? I’ve now reached the grand heights of just over average, kind of.
But the TR power ranking, and the TR bell curves are very different to the Coggan chart, or the Cycling Analytics comparison. The TR power ranking is definitely, erm, more positive, especially for the shorter duration power.
They’re talking about different things.
Coggans chart is talking about the average stats of competitive racers.
The TR charts and the other thing you’ve linked to are average stats of riders.
They are different populations of riders.
Yeah, our Power Ranking page shows your abilities compared to other TR users. 
The Coggan chart was also just invented and is not based on data. Well, the top levels of the male chart were based on actual data (or were, at the time of its invention), and possibly the bottom, but the rest of it is just guessing/interpolation. And the values for women are just 88.9% of the male values.
It’s a rough rule of thumb, shouldn’t be taken as gospel, and should probably be re-made with actual data.
On the Coggan Chart, does anyone else find the 5s and 1min power levels to be incredibly high compared to the 5 min and FTP?
I’ve never really done proper all out efforts in most of those distances, but both my 5 min & FTP according to that chart are about spot on. The other two I’m not remotely close. Maybe I need to try some 5s and 1 min proper efforts.
(I understand it may not be a very science based chart, according to some of the comments)
That could be including track cyclists and full on sprinters. Jasper Philipson can probably top that chart for the 5 seconds.
But that is genetics, specialism and hard work.
Jonathan Milan can supposedly hit over 1,000 for 45 seconds and has a peak power of ~2,000 watts. 
Key Point: While the TR data isn’t representative of the population as a whole, it is real data (power meter vagaries not withstanding) based upon a large sample size. The Coggan data is neither representative of the population as a whole, nor is it based upon large sample size - basically it is a guess / estimate.
Between the two, the TR data is much better representation of the relationship between power at different time lengths, with the big caveat being it probably tops out somewhere between 5 - 6 w/kg. Tops out meaning it doesn’t have enough people in this range (+ higher w/kg) to have good statistics.
Ok interesting, thanks all!
