Test and Regression Analysis of the Ramp Test...what's the r?

Hopefully I am not rehashing a topic here, but has TrainerRoad released the data they used to statistically validate the ramp test? Meaning, I’d like to see their charts, the sample size, the regression analysis, and the correlation coefficient of the analysis between all the variables. I ideally, I’d like to see 4 variables compared: ramp, 20min test, 2x8 min test, and 60 min test.

1 Like

What’s your hypothesis?

How would they “validate”? Workout success/failure rates?

My best bet is that they won’t be sharing core data with anyone outside the TR org.


oh man…tough one. It’s only a guess based on my experience with ramp test, 2x8 and CP20.
I would hope for at least r=90 between the ramp test and CP20. Between ramp and 2x8 perhaps r=95. That may be wishful thinking.


Dream world? :wink:

Please clarify. I assume there is a test design that would yield reliable data. I don’t want to re-invent the wheel here and attempt to lay it all out, but let’s say you had a sample size of 50 cyclists, intermediate through pro (we would want to draw correlations from these sub-divisions as well). A simple test would have each cyclist do both the ramp and the CP20. There would need to be some recover time between tests (2 days? I don’t know), and some conditions laid out on when, how, etc. But the tests would be performed just as they would ordinarily be performed. We are comparing test protocols. If they follow the protocol, the test is a success. I am certain some sports physiologists have done this, but maybe not with the ramp test. I will keep searching,

Total dream world. .9 is near perfect agreement.
Ramp and 20 min… maybe .6- just moderate correlation especially when looking at TT versus sprint power.

1 Like

Y axis is power? What’s your x axis?

I would be very suspicious if that were the case. In the world of peer reviewed science, that would disqualify any study on the spot. Legit scientific studies publish there results to be reviewed by the community.

X are the samples (cyclists), Y are the results of each test. But I pity the cyclist who would do this study! 4 FTP tests in how many days! Not feasible. Those aren’t real results btw. I was just showing an examples. That’s only n=12

Here is a study correlating FTP CP20 with maximal lactate steady state (MLSS). R=.91.

Now just need to find one comparing different FTP protocols.

Is this relevant?

This study compares the 8 min test with several other protocol (e…g, lactic threshold, VO2max). this one is promising as well, so both CP20 and 2x8m aren’t entirely ruled out.

I would need to track the source study down and study it. That’s an interesting one.

Isn’t that what Coogan has been claiming all along? I.e., FTP = power @ MLSS?


wow, a CP3! Check this thesis out. I have heard of a 3 minute FTP test before. Insane.

Suspicious of what? If you don’t like the ramp test, don’t do it.


TR’s not doing peer-reviewed science; they’re consuming peer-reviewed science and combining it with their own dataset and analysis to support a business model. Not only do they not have an obligation to release their data, but releasing it could easily breach their privacy policy.