That's a hell of an achievement. I can TT a hilly windy triangular 10 mile course in 29 minutes 20, so just over 20 mph. I certainly couldn't keep that pace up for 20 miles let alone 50. I wonder how much difference a TT bike and kit would make over my bog standard road bike.
I have some data from 15+ years ago for 10 miles on the same course, over a period a of few years (insufficient for 25 miles, and no road bike data for 50 miles)
Sometimes I'd have clip-ons for the road bike, I can't split the data for that as I don't have records.
There's a significant difference! (2-sample t-test) and and the mean TT bike times are about ~1.6 mph quicker and have lower variance.
The road bike data and the TT bike data are not very interspersed, bulk of the road bike data is for the first part of the period and TT bike data from when that bike was purchased (any training effect is not easily estimable- although I could perhaps use 'athletes' 10' times to do this crudely)
Descriptive Statistics: Average
| Bike | n | Mean (mph) | StDev |
| Road Bike | 22 | 23.01 | 0.719 |
| TT Bike | 27 | 24.57 | 0.467 |
Difference | 95% CI for Difference |
-1.561 | (-1.922, -1.199) |
Individual value plot (minutes) for time