chas58

Yeah, I was concerned about the overly large time gap.

I’m thinking that on the 3 hour ride, I just got exhausted carrying the extra 10 lbs of mountain bike (and heavy wheels), and ran out of steam. The CX bike is a 20lb ultegra equipped race bike, while the mountain bike is your basic big box store Diamonback bike (but with very nice tires).

To filter out any variables, I matched up 9.4 miles of exact overlap from the beginning of my ride. This compares to what is basically a $2000+ aluminum CX bike with a basic big box store $500 mtb bike running the fastest tires possible (yes, the tires totally transform that bike).

MTB compared to cross
+ 2 minutes (5%)
- 0.7 mph (5%)

So yes, for gravel grinding, a CX/gravel bike is going to be pretty similar to MTB. I expect that a comparable high end MTB would minimize much of the gap I found.

Biggest differences were weight, acceleration, climbing, high speed (aero @<20mph). But for longer distance touring and a moderately steady pace - the differences are minimized (other than my endurance was better on the CX bike).

When it dries out, I want to try this with my ‘90s carbon XC race MTB (with bar ends and aerobars). That era bike is capable of doing just about anything.

Spoonrobot 

Make sure to run a regression analysis when you've done a few dozen trial runs in variable temperatures and road surface smoothness. We can discuss P-value choice once you have the data.

I'm with you though, slower/heavier tires are most obvious on hills and can add a significant amount of time to a given course.