Zwift's New Ramp Test (FTP)

  • “The test will let you warm up for 5 minutes at whatever level you’d like. It then starts you on an ascending series of 1-minute intervals beginning at 100 watts and increasing by 20 watts each minute.”

Jordan also said one thing he really likes about the ramp test is that this is a great workout just to do. Looking for a short, max effort? Want to test your limits? Hop on and crush a ramp test!

I always thought one of the strengths of the Ramp Test is that it isn’t a good workout. As in, it doesn’t disrupt training like other, more stressful protocols. In fact, if I’m not mistaken, many TR riders follow up a Ramp Test with a short endurance or sweet spot session.


Just going on the screenshots, if I reached 360 watts for 30+ seconds, I’d hope the actual FTP it gave me was more than 225…

1 Like

Looks like the rider reached 28 seconds of the 360 interval for a total ramp time of 13:28. So it looks like they use about 65% of your last 1 minute power I’m guessing :man_shrugging:. Or they have a completely other formula.

1 Like

Can I get a higher FTP with a power up :thinking:?


If Zwift implemented a 4DP, then that’s going to cause issues I’m sure. As for implementing a years old non ©/IP/trademark step test and saying they’re copying TrainerRoad… nope.

The questions I have to take things forward - Are either TrainerRoad or Zwift taking into account male/female protocols, or estimate calculations? Are they adjusting FTP estimate calculations based on a less ‘elite / trained’ participant levels a lot of these protocol papers are based on? There’s a lot more to take into account than slotting all humans into one bell curve.

The platform that can take the ramp/step test and move it forward, clearly, and with quality data, will lead.


Sure, step tests have existed for many years. But I’m not aware that most of those prior ones were used to determine FTP in the same way that TR has done. I could be wrong, as I haven’t researched it deeply, but I think most step tests that I have seen are looking at VO2 Max and other markers, not FTP.

As to 4DP, that’s a separate discussion. But there were significant and deep discussions about it having strong parallels to much of the direct work from Dr. Coggan himself. The detailed (and oddly amusing) thread on Slowtwitch showed something further back than the Sufferfest test alone.

  • Anything in that direction would be news to me, as I don’t know of any FTP testing protocol (or any other I’ve seen in cycling) that take gender into account.
  • Based on some the comments in recent podcasts, it seems those general considerations may be working their way into the TR universe. But I don’t think we know how far they may go.
  • TrainerRoad in particular used the many thousands of users and beta Ramp tests (over 7000 at the time of their Ramp Test debut, about a year ago) to refine the protocol and their calculations to obtain FTP. Presumably it includes a wide range of amateur riders up to pro levels in some cases.
  • I don’t think we have much if any info on the particular foundations of the Zwift test. The release is fresh and I may have missed a detailed article.

Obviously the Ramp test has pros and cons, like any other test. Based on feedback here, it is not a solution for everyone in a single form, but it seems to work well for many. Personally, it is the most repeatable and reliable for me compared to the 1x20m or 2x8m FTP tests.

It will be interesting to see the reception in the Z world and if it parallels or diverges from what we see here.


Isn’t this what the TR ramp test does as well? Not that there’s anything wrong with that, but it basically backs into FTP mathematically rather than actually testing threshold power.

The only ramp test that would actually measure threshold would be a blood lactate test as far as I know.

The 20 minute and 8 minute tests don’t test threshold power either.

Everything is an estimate. Some tests are just more repeatable than others.


I don’t think the Krebs cycle works differently based on gender. At least I hope it doesn’t :rofl:


Well, strictly speaking, anything less than an hour test “backs into FTP”, via looking at a prescribed amount and/or point in time of a test, and reduces a value by a set factor to estimate FTP.

My point about the TR Ramp test is that it was the first step test I know of that focused on getting an FTP estimate, not some other value, directly from the power data and not another data point (lactate test, for instance).

I don’t doubt that others have likely played with it for that very purpose and I never claimed anything along the lines of TR inventing it. I do think the adoption of a Ramp test in Z is not surely coincidental, and it’s presence was driven at least in part from the apparent success of the TR version.

But it seems to me that TR put extraordinary effort into analysis for the purpose of estimating FTP, and linked to power specifically.

Then they made a large push to share and encourage people to use it for that purpose. That is what I see as unique about the TR approach.


On a more serious note, I’m sure that the TR folks are aware of the individualization issues with various testing protocols. Building something into a product that is powerful but simple to use takes time and I think the TR team will deliver something on this front eventually.


Agreed. The hints in the forum, about the Ramp test and potential for even moving past them at some point, are a signal that this is far from the end of progress on that front at TR.


Sorry just a little confused, probably missing something but you keep saying that TR was the first to implement the Ramp test to calculate FTP, and sure Freil and Coggan look to more “traditional” ways of calculating FTP, but I was under the impression that the Ramp Test (MAP) was age old, and a quick google has articles < 2015, infact the article on autobus has steep differences for male,female and none elite athletes. Although the Autobus article doesn’t give a direct FTP it gives the 7 training zones which is basically what the purpose of the FTP test is (rather than a brag figure)

In fact reading about seems to indicate that a lot of coaches have preferred the MAP\RAMP\Step test of none elite athletes for a while, due to it being more “likeable” the exact reason that TR start using it in June 2018

Again sorry if I’ve missed \ miss understood

< 2015



From the zwifthacks article (2016)


When you press ‘Calculate’ the following will be returned:

  • MAP (final completed stage + percentage of uncompleted stage)
  • Estimated FTP (calculated as 75% of MAP)
  • 20 minute target value for an FTP test (FTP/0.95)

So what they seem to want is for you to do the MAP test, calculate the FTP at 75% and MAP, and then reafirn the FTP with 20mins at /0,95 … now that would be a good workout


Would have shown some integrity and independence for Zwiftinsider to at least reference TrainerRoad’s ramp test. Bit too cosy that site with Zwift HQ, but that’s another story. Shame there isn’t a website that covers all indoor training only in a fun newsy way.

Anyway, really loving that Zwift have got round to a ramp test. Wonder what the calculations are, will have to give it a shot at some point. I test too high on TR’s ramp test every time. Not sure if it’s too high a test or just a lack of confidence to set my sweet spot intervals at the test scores, either way the numbers are a bit scary for me and I have great success by knocking 5 to 10% off. Wonder if Zwift will be the same.

Why would ZwiftInsider reference the TrainerRoad test? This isn’t anything to do with TrainerRoad. See above, ramp/step tests to estimate FTP have been around a long while. As much as I’m a one-eyed Zwift fan, I’ll give some unsolicited advice here for the benefit of this community - Do your homework before pointing fingers. What I’ve read here today comes across as petty whining, uninformed petty whining at best that reflects poorly on this place. Just the other day I mentioned even I’m sick of reading about Zwift on here.

Rise above it. Make this post about how much BETTER the TrainerRoad methods of FTP estimation are. What happens next week when FulGaz implement an FTP test? RoadGrandTours? Maybe CVRCade will get the jump and pump out some kind of AI data analytic based FTP estimation test. While interesting, and worth discussing the merits of… this has descended into you being unhappy with some Zwift fan-site covering Zwift’s FTP estimation test. That’s one hell of a rabbit hole.

Someone once tried to spin up a webshow that covers indoor training and bike tech in a fun newsy way. Peloton didn’t like the idea. They told them to f’k off and kept doing it.

According to what I’ve seen posted the Zwift estimations are 75% of the completed 20W/60sec step.


I appreciate the links and need time to review them. Again, not saying TR was the first ever or only on Ramp leading to FTP.

But they refined their own protocol (after reviewing many thousands of tests and follow up workouts by testers) and made it the default option because of their results. That style of trst is something I have not seen widely applied (by far reaching apps) until them.

Ask most people about what FTP test they know and use, and I suspect you’d get lots of 20m, some 2x8m and likely a tiny amount of 60m or step type.

I just don’t think step testing for this purpose was widely known or applied like TR made for its users. Again, might just be my flawed perspective.

I fully admit I may be wrong in the grand scheme, and may be based.on deeper research (as already stated).

If anyone know me at all, they will know I rarely speak in absolutes. This situation is no different. I made no such claims about TR being the “one and only” for this test.

But I still stand by the claim that I believe Z adopting this test on their side, is a direct response/result to the success of the TR version. Just an observation more than anything.


I’m curios to try the test qnd compare to a TR version if I can fit it into my schedule. I want to see how it feels with the direct wattage steps in Z vs the percentage scaled ones in TR.

That and the dedicated start wattage are interesting and will lead to test length variation for users in a different way than it does for TR.

1 Like

Wow, my impression is the opposite. Appears you bring some negative biases to the discussion. To me, TR users are poking a bit of fun and amused at Zwift’s attempts to “borrow” ideas from proven successes of TR methods.

If you’re sick of reading about Zwift issues/concerns/discussions on the TR forum, it’s as simple as not clicking on that topic to read :man_shrugging:. You don’t have to involve yourself in the discussion to defend Zwift, but somehow you feel obligated to do so. I’m not sick of reading about road racing or triathlete related topics on the TR forum because I simply scroll right past those threads.

This Forum is all things cycling related and many TR users are also Zwift users. So unfortunately for some I guess, Zwift discussions will never end :confused:.