Determining FTP from the Ramp Test

So I have done a few ramp tests from various sources including Zwift, TheSUF (Half Monty), TrainingRoad and had my FTP auto detected by various Garmin devices. Of interest is that the results are always very different.
Today I did a Training Road Ramp test and it determined a result of 196w. The Garmin 1030 which captured the same test via my Vector 3’s (which records power 11w less than the Kickr Core which the TrainerRoad data was based on) determined a value of 205w.
Using the same data recorded by TrainerRoad and Garmin and using the method outlined in the Hawley JA1, Noakes TD. Paper: I get of MAP of 261w (Kickr) and 254w (Vectors) which result in FTP estimates of 215w and 210w respectively.
These seem to be significantly different results with subsequently large impacts on training efforts that would be based on them.
So I chose a value of 200w and entered that into TrainerRoad and complete Abbey. This was Sweetspot intervals and were not too difficult to complete (my HR only crept into Zone 4 due to HR drift toward the end of the workout.
I’m not superfit, I am overwieght and 60, and just started exercising again (to get fit and lose weight) 2 mths ago. Should I experiment increasing my FTP value?

1 Like

TR takes 75% of your best one-minute power from the ramp test for your ftp. Other ramp test might define it slightly differently.

If you’re going to be doing a TR plan or TR workouts, starting with the number you got from their test is probably your best bet. (Keep in mind that sweet spot work isn’t supposed to be too difficult; the whole point is to target roughly the same adaptations you’d get from many more hours of lower intensity work, in less time, without being gradually limited by fatigue.)

TR calculates your FTP as 75% of your best 1-minute power during a ramp test (261 * 0.75 = 196). They’re pulling that calculation from Ric Stern’s work, which estimated FTP at 72-77% of MAP, and from a lot of internal testing and validation based on people who used the TR ramp test and then did TR workouts. TR has written about it a bunch.

There are plenty of other ways to estimate your threshold power based on a known power curve, but they all use roughly the same formula (X percent of your max power for Y minutes). Your Garmin is probably filling in different values for X and Y.

(All that said, 4 watts is well within the margin of error for all your power sources, so using 200 instead of 196 probably won’t make a different in practice!)


I’ve been with TR for a couple of years now. Prior to using the platform I trained using Joe Friel’s training bible book and did 20 min and 30 min tests.

I found that the ramp test would invariably give a number which I would think is too low. I’d even prove it to myself by riding for 40-60 minutes above the ramp test figure.

The problem is, I don’t think I ever completed a TR phase without bailing on some of the final workouts, or if I did, I felt pretty wiped out and would tend to do my own thing for a bit before returning to a plan.

6 weeks ago I started mid volume base 1 and, although the number TR gave me on my ramp test was around 30 watts lower than I’d been using a few months before, I went with it.

I’m in the rest week now, having nailed every workout, except for one 30m Weds session which I missed. I feel good and ready to move on to base 2.

Some might say that if the ftp estimate is low then you aren’t training at your true potential. I’d say that it’s better that I complete workouts and train consistently. That’s why I’ll accept the ramp test value next week (as long as it’s the same or higher!:wink::grin:).

I guess what I’m trying to say is take the TR ftp and reassess after you’ve at least completed a few weeks.


Consequences of FTP set too high are a way worse than little bit too low. Especially that we are talking about training zones not fixed numbers. I would go with lower number and simply check how it goes.


No. Train with the lower FTP. There’s almost no downside to training with an underestimated FTP (within reason). There are massive downsides to training with an overestimated FTP.


Thanks all. Will keep it where it is.



1 Like

If your FTP is set correctly, you should just be able to complete longer threshold workouts - e.g. Palisade and Mary Austin in SSB LV.

TR hasn’t planned Mary Austin for me until 9 Jan :grin:

It does make me wonder, if TP just use the last completed Step, what the point of trying to partially complete a step is. Hawley and Noake use 82.5% of MAP. Interesting how many variations there are based on the same testing regime.

They don’t - your FTP is evaluated as 75% of your last minute power. It doesn’t matter whether that last minute encompasses a complete ‘step’ in the test.


OK so the last 1 minute average power.

1 Like

Yes - more here:

We take 75% of your best one-minute power. If you are above target for the final minute, the app will reduce your result by a small amount


1 Like

Depends of the rider and their experience. If you are more inclined as a “diesel” types these workouts are not particularly hard. Definitely not on “just” side of things. But this is very specific case.


Do they use the same step structure? The TR ramp increases each step by 6% of whatever the FTP value was before the test (e.g. if your pre-test estimate was 150 FTP, each step would be a 9-watt increase over the last), and keeps you at each step for one minute. Zwift I think does either 10 or 20 watts flat, and I know other protocols do more like three minutes per step. I don’t know if there’s been an academic survey of ramp test variations but I’d totally read it.

Best, not last. If you fall below the target near the end, you’re not hurting your result.

Do you have another paper to share? Because I just read the one you posted, and it does not say anything about estimating FTP from MAP.

Here are the conclusions, where Wpeak = MAP

It states that MAP accurately predicts VO2max when using 2.5 minute steps in a specific ramp test (not TR test).

It then states MAP is a good predictor of finish times on a flat 20km time trial.

Nowhere does it equate MAP with FTP.

I don’t see that anywhere in the paper?

In fact, how could they have proposed using anything to estimate FTP back in 1992?

Didn’t read the paper myself - just working off a reference provider by Shane Miller and his Sports Physiologist Dr Stephen Lane from HPTek. I just relied on the math that they supplied.

1 Like

My trainer road ramp test results are so much lower than my outdoor result that I wouldn’t even barely break a sweat if I used it. Indoor ramp test = 205. Outdoor = 270. Most likely, my kickr was messed up. Sold it anyhow for rollers. Since I train 90% outdoors, I use 270 and can complete all the workouts. Now, I only do the easy days and some of the sweet spot workouts indoors on rollers, still using 270. The sweet spot is much harder than doing it outdoors where I can stand when needed, but I can still complete them if I’m not too fatigued. I say use what feels right. VO2 max repeats should leave you barely being able to complete each interval, but still able to do them all. Sweet spot should feel uncomfortable but doable. Over unders should feel almost like you want to puke and you can’t wait for it to be over. That’s how I know when my ftp is set correctly !


But they used 3 minute step I believe. Don’t use that math with TR’s 1 minute step.


2.5 minute steps:

Again you can’t use that math with TR 1-min steps.