Originally Posted by Harold74 (Post 22252969)

When I look at the graphs, I see a rider deliberately holding cadence to roughly constant, preferred values which forces power to vary linearly in relation to crank torque.

It suggests that being in the right cadence zone has a larger impact on power generation than does crank torque.

Originally Posted by tomato coupe (Post 22253097)

Your second sentence totally contradicts the first.

Originally Posted by Harold74 (Post 22253113)

I disagree. The second point speaks to the utility of correlation in the context being discussed. The third point speaks to things that may or may not be inferred by the trend of the graphs. The trend exhibited by a data set and and that data set's coefficient of correlation or separate things with different potential uses.

Originally Posted by tomato coupe (Post 22253121)

Your discussion of correlation is off target and meaningless.

Originally Posted by tomato coupe (Post 22253121)

If power is equal to torque multiplied by angular velocity (i.e. RPM), and the angular velocity (RPM) is held constant, then all variation in the power are due to variation of the torque. This is as simple as it gets.

Originally Posted by Harold74 (Post 22252969)

When I look at the graphs, I see a rider deliberately holding cadence to roughly constant, preferred values which forces power to vary linearly in relation to crank torque.

Originally Posted by Harold74 (Post 22253146)

Yes, so simple in fact that it's precisely what I said myself in my first point.

Originally Posted by tomato coupe (Post 22253167)

Sorry. How about irrelevant to the point, instead of off target and meaningless?

Originally Posted by tomato coupe (Post 22253167)

Yes you did get it right in your first point. But, you totally contradicted it later, which was my point.