Why I failed at TrainerRoad (I think)

Dear all,

I hereby wish to share my experiences using TrainerRoad. My goal is to contribute something to this community. I would love to acquire more knowledge about my physical condition and training in general.

First, we’ll start off by taking you back to March 2019:
I went to a sports medical institute to perform a cycling lactate threshold test. The test protocol consisted of a ramp increasing by 50 watts every 3 minutes. The test ended at 350W, 18m30s after starting and ending with a blood lactate level of 16.3 mmol. FTP: 240W (derived from the best 1-minute power)
The accompanying counselor summarized the results and catergorized me as being a classic ‘weekend warrior’. I. e. cyclists who hardly train during weekdays and go too hard on weekend group rides. Otherwise stated: cyclists that hardly ever spend time in the endurance zone and mostly train their bodies in the ‘higher’ zones. During these lactate threshold tests the lactate blood levels of the so-called ‘weekend warriors’ rise relatively quickly as their aerobic base fitness isn’t that good.
Reflecting back to myself we were only to conclude my base fitness wasn’t well trained and I relied heavily on my anaerobic capacities during this test as my max blood lactate level was exceptionally high.

Fast forward to September 2019:
Me purchasing my first indoor trainer (Wahoo Kickr). FTP: 232W after performing the Zwift ramp test and thereafter commenced with a a 6-week zwift training plan.

November 2019: Crash and burn!
I managed to complete the 6-week Zwift training plan but without any recovery week in between this really broke me, physically and mentally! The following weeks I was fatigued and lacked motivation to train any further. Instead I tried to just have fun by doing outside rides and on Zwift by joining group rides and some unstructured workouts.

March 2020: Off to a new start, TR-time!
My first TrainerRoad ramp test: FTP 224W. I embark on a low volume plan.

April 2020: Gains!
TrainerRoad ramp test result: FTP 239W (6% increase)
At that time the workouts already felt pretty though. A 6% FTP increase is nice but the zones increase by 6% as well. Each workout my heart rate would eventually go into the anaerobic zone while only doing sweet spot work and an occasional VO2max workout.

May 2020: More gains!
TrainerRoad ramp test result: FTP 254W (6% increase)
After this ramp test I commenced with a build phase which consisted of threshold work versus sweet spot work in the previous base phase. This combined with the 6% FTP increase meant that I suddenly had to ride 10-minute intervals at 16% above the sweet spot work of the previous phase.
By that time I could barely hang on during workouts and I would fail to finish every single workout.

June 2020: Crash and burn (again)!
TrainerRoad ramp test result: FTP 230W (10% decrease)
Mentally and physically drained I cancelled my TrainerRoad subscription and during the next months I stayed away from any form of structured training.

September 2020: back to Zwift
Zwift ramp test result: FTP 256W (all-time high) (Zwift ramps up quicker with 20W steps instead of 15W in my TrainerRoad ramp test)
At the time I was very suprised by the result. I had stayed away from any structured training the previous 3 months and I still managed to break all previous FTP results.
The next period of time, I did some sweet spot workouts combined with outdoor rides but no real structure.

December 2020: To ramp test or not to ramp test, that is the question.
Looking back at the past year of indoor training I must admit that I am a bit disappointed. All the hard workouts barely paid off as I overreached myself at least 2 times.
So I start to think: “What is FTP? FTP is the measure of power you can hold for one hour.”
Then it struck me, never in the past year I would have been able to maintain my FTP wattage for an hour. I know this for a fact because I even struggled to complete sweet spot workouts with intervals at 90% of my so-called FTP.
So very recently I executed the Zwift 20-minute ramp test. You start off with some sprints and a 5 minute effort above threshold to drain the anaerobic tank and thereafter you go all out for 20 minutes.
The results: 231W over 20 minutes. If you calculate FTP you multiply this value by 0.95: new FTP is 221W.
Could it be that over the past year my ramp test results where a gross overestimation of my FTP?
Could that be true because I am an ‘anaerobically stronger’ rider and during the 1 minute in which the ramp test determines your result I happen to excel because of my capability to cope with high lactate blood levels?
FTP should be the result of a 1-hour ‘aerobic’ effort but me doing ramp test probably gives me different results? In fact 13 % off if you would compare my 20-minute test with the September ramp test of 256W.
My last sweet spot workouts went really well and my heart rate usually stays in my tempo heart rate zones.

If you are still reading by now I would very curious to read your thoughts and advice on my story. Any misassumptions I made or experiences of your own. I would be glad to read and learn about them.

Thanks in advance!
Kind regards,
A fellow cyclist fanatic,
male, age: 30, weight: 69kgs


Look at post on sweet spot progression to help build your base and kolie moore testing procedure for alternative testing processes…of can only do 230ish W for 20 mins then ftp is certainly no higher than that and likely lower. For some people ramp tests work very well to estimate ftp…for others they do not.


Thank you! I will definitely look into those!


@Gurbe_Willebrords I’m going to just re-hash what the TR ramp test is (which you know, but just for the sake of completeness) mention a couple of papers supporting the notion the among weekend warriors and highly trained cyclists there is significant population of athletes that will find the TR Ramp model unusable…and then copy & paste some text re: TR ramp test that I’ve posted before (pay close attention to the example rider in that text)

TR ramp test is a Maximum Aerobic Power (MAP) Test. MAP tests are an age old way to estimate VO2max. TR FTP is an estimate of FTP based on the estimate of your VO2max (MAP). So an estimate of an estimate. FTP as a percentage of MAP can vary substantially from rider to rider. Way, way back in the day Gollnick did a study on enzyme effects of cycling training…subjects in the study did a ramp assessment & then cycled at 75% of their ramp assessed VO2max 4 times a week for 5 months. Here is what Gollnick observed at the beginning of that process:

“Initially subjects could not tolerate this load for the full hour and it was reduced to about 65% of VO2max during a portion of the exercise bout…At the end of the training program most of the subjects were working for 1hr at 85% to 90% of their VO2max”

If you go read that paper pay special attention to Table 1, 2nd subject. PD Gollnick’s paper…subject PDG. Hmmm. Not just an egghead observer.

Also way back in the day, Coyle/Coggan looked at 14 cyclists with mid-60’s VO2max numbers & a lot of cycling experience. They did a ramp test to establish VO2max and took lactate curves to establish Lactate Threshold. Among those well-trained cyclists LT expressed as a percent of VO2max ranged from 59% to 85%. Mean LT/VO2max was in fact 74%. But standard deviation was over 9%!! Even among well-trained cyclists, out of 14 cyclists, only one of them would have been well-served by a 75% estimate. The rest would have been either overtraining or undertraining. If you read this study pay special attention to Table 1.

PD Gollnick Effect of training on enzyme activity and fiber composition of human skeletal muscle.
Coyle/Coggan Determinants of endurance in well-trained cyclists

Here is what I’ve said about the TR Ramp Test in the past:

"I suspect correlation between TR MAP-based FTP estimate and actual FTP or actual hour performance is not that great. For a few reasons:

1.) TR is pretty honest with their user base. They never talk about this. We talk about it a bunch. What does that tell you?
2.) A progressive ramp test is designed to estimate Maximum Aerobic Power, or MAP. Pegging FTP at 75% of MAP is using an estimate to make an estimate. Usually such things don’t have good R^2. :wink:
3.) There is a lot of data from other sources suggesting that individual variation from the 75% rule can be substantial. Anecdotally, I think dialogue on this forum only serve to support this notion.

MAP-based FTP is a tool. It designed primarily to improve testing compliance & I think it’s good for that. Directionally, it’s a good measure of what’s going on with FTP. On an absolute basis, it can certainly OVER report FTP and UNDER report FTP. Individual users should be aware of where they sit on that continuum and either adjust plans accordingly or take steps to correct deficiencies (see my hour of power threads for more thoughts on those steps)."

" Here is what I think is going on: TR uses (primarily) a MAP ramp test to determine FTP. Most would perceive all TR plans to be FTP based but they are not. They are Maximum Aerobic Power based & the first thing you do with your Maximum Aerobic Power is multiply it by 0.75 to get your ‘FTP’. So, really, all training plans are based on 0.75*MAP…not FTP.

Here is the problem with that MAP-derived FTP number: if you compare the actual FTP to the actual MAP of a population of riders with similar MAP you get a BROAD distribution. Remember that paper from way back in the day by Coyle and Coggan where they did exactly that? Determined FTP as a percentage of VO2Max for a bunch of trained cyclists? Some of them were 60% & some of them were 85%. A lot of them were less than 75%.

TR has no way to deal with those riders who have sub ~70% FTP. And in fact, at least for the sustained power build plan, TR puts those riders into a destructive positive-feedback loop. Imagine our trained cyclist who has an FTP that is 60% of their VO2Max. TR uses a ramp test to estimate VO2max, then assigns an estimated FTP of 75%*VO2Max.

Yikes! Our 60% rider’s next workout is Avalanche Spire! The poor fellow is doing over/unders at (nominally) ~118% of his true FTP. That’s a VO2Max workout . So he slogs it out. Does what he can. It’s a super tough workout…maybe he can’t complete it all. People on the forum tell him it should be ‘hard but doable’. Especially those riders who are lucky enough to be in the >80% FTP-to-VO2Max club.

Our 60% rider is one tough cookie. He hammers through workouts as best he can. Then, when it comes time to do the next ramp test, guess what? He hasn’t been doing sustained power for the past few weeks AT ALL. He’s been doing a ton of VO2max work. Guess what that does to your Maximum Aerobic Power? It makes it better…so now his next ramp test…surprise, surprise…is a little bit better.

But his FTP probably hasn’t improved that much.

That’s what I think was going on with me. My physiologic profile favored Maximum Aerobic Power. As a result my MAP test results caused 75% of MAP to overestimate my FTP. So when I executed a TR workout at 95% of TR-ramp-test-derived FTP…I as really doing intervals at >110% of true FTP."




No, I’m afraid not. This is a popular misconception, and there are many threads around the internet on this. There is an apparent threshold above which fatigue comes increasingly rapidly, and below which decreasingly rapidly. It’s a quasi-threshold* plateau in the graph of anyone’s stats, and lasts roughly somewhere between 40mins to over an hour.

Rather than exhaust ourselves performing at this exhaustion level, we perform tests that estimate this threshold of output; RAMP, 20 min, 2x 8min, 40km TT, and so on.

Any test tells you how you are at that moment, they don’t tell you how good a cyclist you are or how good a person you are. TR training scales according to how fit you are on the bike, so regular testing is useful and beneficial even if you have concerns about it’s accuracy or relevance…it does the job of scaling your next training session to you.

Real world results should be your actual benchmark for how good a cyclist you are, imo. Race results are race results. A new PB up a local hill. Finding your weekly group rides getting easier.

I decided to go back to 20 min test earlier this year and found I tested higher than on on the RAMP. I also raced a couple of virtual 40k TTs and I’m about to do another one this weekend…all of which inform my abilities to race but I won’t be changing my FTP until I do a TR test.

  • I may misremember the exact terminology

I feel like the ramp test has been serving me well, so far. Maybe I’m one of the lucky dogs that fall in that happy medium.
However, I’m debating going back to the 20mins test, to actually SEE how i fare on one of those.

1 Like

As I said on one of the other “Ramp Tests are bad, okay” threads, I trust the TrainerRoad guys have the data to back that it does work for the majority in setting the training zones for TrainerRoad. They did beta testing, and now have months (years?) of data from ramp tests and completion of workouts data.

I’m in the minority, where it really doesn’t work except for setting VO2 max zone. I can complete Fang Mountain +1 at 100% at an FTP significantly above what the ramp test gives me. I just think us outliers are noisy about it! Just use a longer form format if you’re sure it doesn’t work for you.

Is it (or is any “test”) a true reflection of FTP? I think that’s a different question. But also different to what the TR team have claimed - I would say from the podcasts they’ve been pretty consistent that it’s just a method of setting zones for training.


Have a look through the Inigo San Millan Training Model Thread for an alternative way to train. Iñigo San Millán training model

1 Like

Exactly this! :+1:t2:

1 Like

@Gurbe_Willebrords it doesn’t completely obviate the TR Ramp Test, btw. I’m just saying individual riders need to be more aware of where their threshold falls as a percent of MAP. TR is not going to overtly help you achieve this awareness…you gotta do it on your own.


Yes…remember for some (quite a lot of people…just those it doesn’t work for are more likely to say) but not for all. I regularly do a 40 min effort to gauge/estimate my ftp and often do a TR ramp test around a similar tine to compare and results are usually within a few watts for FTP estimate…however I prefer the longer efforts as good workout by itself but gives me confidence to KNOW I can hold X watts for 40+ mins rather than assuming I can from ramp test.

But the longer efforts work better for more experienced riders who know/can predict their ftp with reasonable accuracy in order to pace well. The best bit for the ramp test (and a main selling point from TR) is no need to pace, just go until you die! For newer riders this is a big plus point.


Agree with everything you said and appreciate the detailed post.

Not wanting to derail the thread but this:

stood out to me as TrainerRoad developed plans well before introducing the ramp test, when the 8-min and 20-min protocols were used to estimate FTP. All of the zones in workouts are based on Coggan classic levels. You’ve made this point before, but without giving a reason.

And then there is this:

and later in same FAQ:

and this:

All of which support my conclusion that TrainerRoad believes a) the ramp test provides a fairly accurate estimate of FTP, and b) that the workouts are in fact based on FTP and Coggan classic levels. For an example of MAP based zones, this Ric Stern article from nearly 20 years ago has some details on his ramp test protocol and his MAP based training zones.

While there have been minor tweaks to individual workouts, to increase compliance, there have been no changes to the plans to accommodate the switch from 8/20 min protocols to the ramp protocol.


To play devil’s (TR) advocate for a minute…It’s in TR’s best interest to keep people engaged, training and renewing their plans. If the OP’s experience was representative of most users, they had to be terrible business people not to look at their data and do something about it.

So by this logic, they must be seeing that for most ppl the Ramp Test is ok and better than the alternatives in terms of renewing subscriptions. FWIW, in my case, Kolie, Ramp and 20min tests give similar estimates.



This is very interesting…do you have a link to the studies raw data?..perhaps the median would have been more appropriate. Were there any outliers?..I’ts really hard to work with such a small size.

1 Like

Believe it was this study:

and FWIW, I’ve seen the ramp give inconsistent estimates. When it gives an inaccurate estimate, my ramp estimated FTP is 20-40W BELOW (under-estimates) what I can do on sweet spot and threshold workouts. And FWIW I’ve had more good ramp tests than inaccurate test results. The inaccurate ones are when WKO shows a high fractional utilization (FTP is 88-90% of estimated vo2max), however I’ve had good results at similar fractional utilizations.


There are a lot of descriptions like this on the Polarized Training thread.

There is also a considerable and under-appreciated “learning how to do a Ramp Test” going on IMHO. When looking at the Ramp Tests I have done, my max HR has increased by 18 BPM. Does this mean that I have somehow broken through to a new level of fitness? Maybe, but I think it is more likely that I have just gotten better at suffering and have “learned” how to push to exhaustion better.

I have had good luck with Ramp Test results estimating VO2 max power appropriately, however it appears to over-estimate aerobic capabilities (20 min and above efforts) for me. This has led me in the past to have too few days of low-intensity, and too many days of high-intensity, which then leads to the exhaustion you described above.

As always, YMMV.


ok, the story is a bit different…there were 2 groups with 2 different cycling training experience

But this thing is from 1988, is the status of endurace research that poor, that we don’t have a better reference?

1 Like

Based on the study you cited, they way to deal with this is to use cycling training experience as a triage.

1 Like

Your idea is that as years of training increases LT as a percent of MAP will increase? I think that’s an interesting hypothesis. Definitely the Gollnick study supports the notion that if you just do an Hour of Power 4x/wk even recreational athletes can achieve hour thresholds that are a material portion of MAP.

That was basically the conclusion of my ‘it’s been a great year’ thread. Lots of steady state sweet spot work closes the gap.

1 Like