I ride 165s on my tri bike and 170s on my road bike. To be honest I can't really tell the difference--that being said (a) I know I would feel the difference if I threw the 170s on my tri bike and (b) i've also dialed in my fit around each. I also wanted a little more leverage on the road bike for sprinting.

Mtb 175
CX 170
Road 165
TT 155

Crank length is a tool for a purpose

I ride 160 on my tri bike and 170 on my road bike have varied lengths on both but they are the best lengths for me. Totally different positions tri with a closed off hip and road more upright so best length on each bike is different and pedalling dynamics are different. I put 165 on a CX bike I had for a while and hated it going back to 170.

Road 165
Tri 145

Road 135
Tri 110

Road 172.5
TT 165

I don't feel the difference at all.

Road: 170
Tri: 167.5

---

-------------------------------------------------------------
Tough Times Don't Last, Tough People Do.

This does not make sense. for sprinting you want to crank up RPM quickly at high crank torque. This is generally easier on shorter cranks and shorter cranks are technically a higher gear than long cranks. Longer cranks are like adding more teeth to you rear cluster or reduing teeth on your front derailleur.

Most track sprinters prefer shorter.

tri 165
road 175
never noticed anything, i went 165 on my shiv to maximise the drop i could get.

The road bike and tri bike are different machines optimised for differnt jobs. Your position is different (or should be!) and there is no reasonable point in trying to match crank lengths either. The same crank length in a different position is still a different setup.
I use 172.5mm cranks on my roadbike and 160mm on my tri bike. I have absolutely no qualms about using different lengths. The two bikes feel different for many reasons. Crank length is one of the more insignificant differences.

I see what you're saying. I always thought the longer crank length allows you to apply a larger torque/power (with the same Force) since the moment arm is larger.

You are correct in theory, but that means keeping RPM constant at the same pedala force since power = Force x distance x RPM

Let's use the extreme case of 2 foot long cranks vs say 150mm cranks (half a foot). Since distance in the above equation went from 0.15m to 0.60m, you would think that power would go up 4x. But it would be very hard to keep your RPM the same and your pedal force the same (imagine using an arm curl machine in the weight room and making the lever four times the length so you're outsid the natural range of the human limbs).

So that's an extreme example, but inside the range of human limbs you still have some this effect going on. Increase the length of the lever and you can't keep up RPM, or if you keep up RPM, pedal force may drop because how quickly you can or cannot contract muscle fibers.

Another way of looking at this from human evolution side is iwth running. If you want to run faster (meaning higher power), you can take these massive strides and bound at which point your "RPM" is down because your Force and your range of motion went up, or you can go with lower ground force and lower range of motion and higher RPM. There is a reason top sprinters all end up in the range of 4.7 strides per seconds while they do sub 10 second 100m races. After that its just ground force and range of motion. In biking we lock in your range of motion (frank length) so your only variables are RPM and pedal force. Running you can play with three variables.