Why I failed at TrainerRoad (I think)

It does, but also offers other options. The plans are based on FTP, not 75% of MAP; you can see this readily with threshold and over-under work. They are based on “FTP = 75% of MAP” for VO2Max, but with a ton of notices that you may need to adjust the level to compensate.



yes, just looking at the means and std you can tell that the difference is significant…however, the sample size is problematically small. Here’s another way to go about it

1 Like

yeah. Your hypothesis could be correct or it might not be. Definitely if you just give the chart an eyeball test it looks like there might be something there. Not enough data to conclude.

Let’s all train hard for another 4 years and see how it turns out!


Some more data…items in red & yellow are my own annotations.



Effect size of nearly 4?? Yowzer!!

1 Like

Completely agree here Brennus. Same thing here as my theoretical FTP is a lower % of my MAP and over-unders were causing me to be in VO2 max barely hanging on for minutes on end. I utilize a 4m version of ramp test with my own lactate meter to track progress and set training zones vs taking an estimate of an estimate of an estimate. Basically if I’m working on V02 max they are either all out efforts for the “ON” period or the highest 1m of my MAP test which is a good number to use for V02 max power. Then for my base rides, etc. I’m using the information gleaned from the ramp test and my lactate levels.

This strategy is far more accurate than the FTP model even if you actually have done a 1 hour test in my opinion. In that example my 1 hour FTP might be 240 watts, but if I apply 120% to that to complete my v02 intervals I end up with 288 watts which is 20-30 watts below my actual v02 power.

I think, but am not totally sure, that people lower on on the “FTP” ratio to MAP would be better served doing longer steady state efforts actually in the correct zone. Unfortunately the way it is done with just a 75% from the ramp test you are riding too hard all the time and never get the aerobic adaptations needed to push that number closer to 75% or higher.


That’s a given, no?

(Rhetorical question. This has been accepted since at least the 1970s.)

Pretty much everything you said above has gone through my head. I love TR, and I have done over 700 workouts (well over 1,000 hours). 75% just doesn’t work for me, but I only do the ramp test because the others are just too stressful. I foolishly fall into the trap after each assessment thinking that this is the time I will be able to handle the ramp-test-determined number. I know that for some people, 75% works just fine, but I think that TR should lower it to somewhere between 70% and 75%. You might be saying, well, it’s working just fine for me… just wait until your FTP starts to plateau.

Just look at the failure rates of higher intensity factor workouts like Spencer +2, Carpathian Peak +2, Raymond +whatever, the list goes on. Few riders are “nailing” these workouts, and I have researched the ones that are finishing. Many many of them are relatively new to structured training and are on the rapid ascent part of their FTP development (as I once was), and they’ve already far surpassed the fitness of their last test. So if TR is even looking at the success rate data, the success rate numbers are skewed high by athletes who’s FTPs are rising rapidly, which probably comprises a large part of their user base.

I love the podcasters, but it makes my blood boil when I hear them talk about the fact that athletes just need to become accustomed to the discomfort. I’ve done so many indoor and outdoor efforts that brought me to my limits. On the contrary, many of us need to realize that we don’t need to be digging so deep every day. To the podcasters credit, they acknowledge this.

I actually skipped my last ramp test last time because based on past experience, my test would very likely have come in higher. I am currently experimenting with modifying (reducing) my assessed FTP to a level that appropriately sets the intensity level. Right now, I am at about a 3% to 4% reduction from my tested level (or 72%-73% of the max 1 min from the ramp test). That small adjustment makes a HUGE difference. Still, I am struggling to finish some workouts. Of course, this all depends on my V02max/FTP ratio not fluctuating too much going forward, or I may end up setting it completely by feel in the future.


I can’t tell why you had a less than optimal experience with TR, but here is how I look at the ramp test. It’s only a gauge to estimate how “hard” to make the workouts. While I can’t lie that I don’t pay attention to the FTP number provided, the way I use it is in dialing up or down the intensity of the workout.

This is my first go with TR, and the FTP set with the SSB plans was perfect for me. Not easy, but not super hard. I finished all of them. Coming to the Build phase, it’s becoming much more difficult, and I’m having to decrease the intensity in the app to 95%. With this I can finish and still be (very) challenged. So depending on the type of workout, even for the same athlete, it may be perfect for some but not for other workouts. I am seeing dramatic progress almost weekly as I ride outdoors only during weekends. It’s always better than the prior week.

TLDR the FTP number doesn’t matter as much. As long as the workouts are not too easy or not too hard, you will be making progress.

1 Like

But that’s the whole crux of the issue. All of the workouts are based on your FTP. So if you FTP is too high then all of the workouts are going to be too hard. And if they aren’t too hard individually then they are when done consecutively in a plan. And then the opposite problem if your FTP is too low. So if you are doing workouts and training plans that are based on %FTP then getting that FTP number correct is crucial.

1 Like

Honestly you are not giving it enough time and there needs to be more of a long term goal to keep on track. People that get fast have years of “base” fitness.

Keep working and don’t get discouraged.


Not sure. That number sometimes is spot on, but if it isn’t (too high or too low) you’ll need to adjust. You’ll also be able to follow the trend of multiple ramp tests. So for training it will always be pushing you to progress.

1 Like

Isn’t this was the % adjustment is for?

I use the ramp test and for the last few years it’s been ok. However I reached an FTP high during a recent test and started sustained build MV. Over-unders were ok (just :grimacing:) but for the V02max I had to turn down 2-3%. (the first time I’ve had to do this)

The goal of each session is to get training stimulus and I got that depsite having to knock a few % of the intentisy. Hopefully I’ll be able to redcue the reduction as I go through the plan.

Don’t become to fixed on the absolute number and having to achieve it at 100%, it’s the training stimulus your doing the workout for and you get that at 95%.


After a while you learn if ramp test works for you or bit over/under estimates and learn to adjust ftp. So if you know ramp usually overestimates for you then manually adjust ftp down a bit.

Same issue with other tests though as well, for example the normal 20 min ftp test procedure takes 95% x 20min power. I know for me that 92-93% of 20 min power is better. For others it might be 97%.

I’ll chime in here FWIW. Tests are tests, and estimations of what the outcome could be. Even looking at education tests etc. they are not a fool proof method of doing something. If you test well you may be good at retaining specific information or if you test badly maybe that is your weakness and not necessarily knowledge. That being said, a tool like TR has to try and provide some way of making things worthwhile from both their perspective and from the user perspective. If you wanted precise data and “perfect” methods you would be better off going to a sports physiologist and having scientific tests done.

Additional to that is the use of different metrics, while not perfect, PWR, HR and PRE are metrics at your disposal, and I would caution anyone that when those do not line up with the goal of the workout then there may have been a flaw in the test/estimation. If a single steady state workout is tough then fair, maybe it was a bad day etc. if they are continuously tough and you cannot maintain the effort and you creep into threshold zones with the other metrics then maybe you need to consider that something isn’t right and adjust and understand why that may be. OP actually did that in a way but the human condition is that we all need to be “normal” when we do not conform to the journal data on something it isn’t accepted and we rather push forward doing more harm than good. Physiology has incredible ranges. If things work out at 75% great if they don’t try look a little deeper, you can be a good scientist to yourself.

Additionally I would add that no method is perfect, and have tried, ramp tests, varied ones, 20min tests, 1h of max effort up a climb, as well as Sufferfests testing, all have pros and cons. If the workouts are too easy or too hard then try to understand these factors. The tool is only as good as you can apply it to your circumstances.

Maybe I’m ranting because my planned race is postponed :frowning:


I still come back to TrainerRoad have the data!

During beta testing of the ramp test they continued to tweek the %. If they were seeing widespread failure of workouts, as has been suggested, they’d adjust the ramp test figure. More often they’ve suggested on the pods that the failures are the people who haven’t followed the ramp test “FTP” or have been inconsistent in the lead up, if I recall correctly?

They’ve also adjusted some of the workouts in plans with high failure rates (Mary Austin comes to mind), but I think they’ve always done that so not necessarily ramp test related. The workouts mentioned above are bloody hard! I’d expect them to have higher “failure” rate than easier ones.

There’s plenty of threads on the data that TrainerRoad has, what they could do with it, what they want to do with it, but we’re to believe they’re sticking with the ramp test despite it not working? Sorry, I really don’t buy that!

Again, speaking as someone for whom the ramp test doesn’t seem to work.


It would be interesting to see the results of a poll or from data mining the percentage of users who are perfectly fine with ramp test results, what percentage have high anaerobic contribution and over estimate from ramp tests, what percentage have high fractional utilization and under estimate from ramp tests.

Probably the more important would be the percentage of users who over-estimate (strong anaerobic relative to aerobic contribution) as that is more likely to cause these kinds of problems. I’m guessing that percentage would be relatively low or at least low enough that they think it is worth the risk vs advocating 8m or 20m protocols as they are sticking with the ramp.

Under estimating just means you won’t be training hard enough but should be able to complete sessions, over estimating seems more likely to lead to blow ups and non-compliance.

1 Like

If you feel the ramp test doesn’t work for you there are choices.

  • Turn down any workout by a few % that’s too hard
  • use the 2 x 8min test
  • use the 20min test.

TR were quite open that in their opinion the ramp test is better for most people as it is less demanding and doesn’t rely on accurate pacing.


A perfect post to sum up what’s in multiple threads on ramp test vs FTP. Nicely done! :+1:t3::+1:t3: