I think I am driving myself crazy here but I cannot seem to find a clean or clear protocol for performing a ramp test.
Friel says in his book 20 watt jumps every minute then take 85% of the last minute for FTP. TR and Zwift use 20 or 25w every min but then use 75% for FTP of the last minute. I have also found 82.5% of last minute referencing Rick Sterns work. Why are these percentages all over the place?
And then the jump per minute changes as well. I’ve seen 15 for women, 20 for elite male and 25 for non elite male per minute (that one really confuses me, cannot find the logic there). I also found 25 per 2.5 minutes. I can understand that using different ramp rates would be looking for different things but what are they used for? What am I missing?
Does anyone have a clean protocol for establishing FTP and then a separate one for performing a lactate test? I cannot seem to find good consistent info on performing either of these tests.
A ramp test is a good way to overestimate your FTP. As training is based on a % of FTP, an overestimated FTP can result in your easy rides being too hard. Ego wants the highest possible value.
FTP, by definition, is the power you can hold, in a quassi-steady state, for about 30-70 minutes without fatiguing. It’s important to understand that a ramp test is not in a steady state, as you’re increasing power until failure. The MAP, Lactate and DFA-alpha1 tests are step tests, not ramp.
There are a few versions of this but these two (inside red blocks) are among the best; you can find these in my shared “testing protocol” folder, on my profile page:
Easiest way to establish FTP is CP testing. Once you know CP, then you can determine the appropriate ramp rate or step length (ramp test and step test are both graded tests) for your lactate measurements.
The reason you do CP testing is because it gives you a really good idea where mlss is, so you can focus on certain areas to get good lactate samples.
I’ve had good experience with this method in both running and cycling.
Ultimately I’ve come to the conclusion that lab testing is fun, but the only real measure of performance is the performance itself.
Gerald. Thank you for sharing such detailed info. I will definitely take a look at the FTP tests you shared. I am curious as to your take on why there are mainly two different numbers used to calculate FTP off of the ramp test. Friel 85% and Zwift/TR at 75%. Those are big differences and I’m trying to understand why. I can also see to your point if you use the wrong number it will definitely overestimate the number. I get that these are just triangulations based on population level data but they are helpful when used with other confirmation points to get close. Obviously the only way to truly get accurate number is lactate testing which I want to try as well. Thank you again for the quick and thorough reply.
I don’t know the answer to why one would be 85% and another 75%. It would be foolish of me to pretend I know.
I guess the same can be said of using 92-95% of the 20-min effort. Not everyone is the same, and some people can be closer to the 92% of the 20-min power.
I have a lactate test protocol too, which I don’t think I moved to the shared library folder. There is also the test protocol for Inscyd, as well as Wahoo’s 4DP (full and half Monty). That said, the 4DP is a power profile test over one effort, similar to the power profile test I shared from the Training and Racing with a power meter.
Consistency is important, and if you test the same way each time.
Personal preference… I prefer to do the traditional 20-min test, or the Kolie Moore version, then build progression in the training to a point where I’m capable of doing at least 45-min at the same power (extending TTE). Then build intensity on top of that, and the fitness level (not CTL fitness, but ability to hold the power) seems to remain high for a while. To get even better, do some high intensity (5m max effort, then go do some endurance/tempo work and try the 5m max effort again (durability).
I typically do 2-3 time trials ranging from 3 to 40 minutes when I’m trying to identify CP. I’d recommend 3 and 12 minute trials for the first iteration, and then try to do 30-40+ minutes near the calculated CP from the short duration tests. The long test will help bring the CP estimate to a more representative value. Here’s an example of how I analyze things in Excel. It’s quick and dirty, but it works.
Once you have a good estimate of CP, you can then try 10 minute steps for the graded test. Try to place 3-4 of your step intensities below CP and 2 above CP. Then plot the data.
Basically you just need to do some all-out efforts to estimate CP. Lactate will just tell you what you already know once you get the hang of regular testing.
This is posted on the FastTalkLabs forum, and given it was sourced online, I’m sharing it as posted.
Found this online: Calculating Critical Power
Multiply the average power you held over 12-mins by 12.
Multiply the average power you held over 3-mins by 3.
Subtract the number calculated in step (2) from the number calculated in step (1)
Divide the number calculated in step (3) by 9. This is your Critical Power.
There are research papers recommending that the longest effort be no longer than 15 minutes. You can find more here: https://kar.kent.ac.uk/104738/1/77Final_Thesis.pdf
from pg 40 of the above link:
Despite multiple research studies supporting the use of CP as the gold standard for
determining the MMSS (De Lucas et al., 2013; Jones et al., 2019; Nixon et al., 2021), some
concerns persist regarding the accuracy of CP estimation (Muniz-Pumares et al., 2019). The
conventional CP determination protocol requires 3 to 5 different points (power outputs for
given durations) on the PD relationship in the severe exercise intensity domain. All
determination trials must be between 2 and 15 minutes (Munix-Pumares et al., 2019). Any
intensity that lasts less than 2 minutes or more than 15 minutes would risk the intensity being
outside the severe domain or unable to attain V ̇ O2max, which would affect the CP estimation
(Hill et al., 2002; Vanhatalo et al., 2016; Burnley, 2022).
To estimate CP, you can plot 1/Time (in seconds) vs power (in watts) and use Excel to give you a linear regression. The slope is W’ and the intercept is CP. Or you can use a back of the envelope calc like Gerald provided. I actually use an additional 4 models to estimate CP/FTP from power data - 2parameter, 3 parameter, ExtendedCP, and FTP with WKO5. They are all very close.
And yes, after much research, thought, and practice…I understand FTP as a proxy for MLSS (or MMSS) and CP as a mathematical failure envelope that happens to come out very, very close MLSS, at least for me and another athlete I monitor.
Gerald, I appreciate the candor. I thought there might be a rift in the community about the differences in testing protocol. It just seems to be so aberrant that I figured I was missing something. It was really more of a matter of interest I guess, that I figured would lead to better understanding.
I have always used the 20-minute test as my gold standard but again just looking for the new information as I’m getting back to training after years away. I really appreciate you sharing the protocols with me and will dive into that. I should really like to look at the lactate test protocol if you’re willing to share that too. Like GarageLab I’m interested in doing some testing on my own, the sports science really interests me on a personal and professional level.
And thank you for the workout ideas I really like the idea of intervals bookending endurance. That makes a lot of sense as opposed to just doing the intervals at the end. Durability is a major factor that I believe gets overlooked today especially given all of the ultra distance events.
Is there somewhere to access the Inscyd or 4D profile or is that all behind a paywall? I do want to be mindful of people’s intellectual property
I in no way qualify as a strong or experienced rider and will most likely put myself to great shame with this reply/question/input but I know a bit about power measurement on race engines and I apply that “ideology” to my own training.
I’m close to drag racing on a fair level having been part of teams that have collected over 30 European Championships over time. 2 and 4 wheel vehicles running on gasoline, alcohol or nitromethane.
Noone in the genre or on that level care about the absolute power output number in relation to anyone else except maybe for a high estimate to put in pressreleases or on hero cards.
What is important is consistency and repeatability in testing. Simplifying it greatly but you start out the day with a baseline test using a set protocol and equipment. You then change a parameter and note the delta in power in the next test. Act on it. Repeat.
Not that it happens IRL, but let’s say that your engine shows 100 on your dyno and you somehow know that your competitor’s engine shows 110 on their dyno. Noone cares one iota about that as it means nothing to either team. The only thing that matters is the delta between two tests with your own engine and your own test setup.
And of course, ultimately how the vehicle performs on the race track.
Isn’t it pretty much the same in training? My bicycle power test numbers are mine and have always been derived the same way. When that number changes using the same test procedure and equipment, I should try to analyze why the delta went one way or another and use that knowledge to my benefit.
Please set me right here but isn’t that what testing is all about for bicycling as well as say motorsports? Repeatability and consistency in the testing procedure of your choice and staying with it no matter what test procedures others use?
Yes, I can see benefits for a rider in choosing a specific FTP procedure over CP or vice versa but IMO it still boils down to sticking with a testing procedure and analysis method that is based on your choice relative to you, not others.
That’s what is called ‘performance testing’ and yes, it is very important to track which actions increase/decrease performance.
A big advantage of this type of testing is that you can do it at any intensity. As long as the protocol is the same, you can check if your power at HR=x has improved or not. Or your decoupling over x hours is more or less. Etc…
In cycling the FTP value is strongly overrated, but it was developed to get an idea of how one cyclist compares to others. FTP in W/Kg and VO2max will give you an idea of the level of performance and the type of cyclist. Still, cycling is more then putting down the most watts. Motivation and clever racing to minimize energy expenditure is just as important.
For most of us as amateurs, FTP only has some meaning to calculate Load. An FTP within the ballpark is good enough to give you a consistent load measurement, allowing you to track Fitness/Fatigue/Form.
Performing regular FTP tests is for most not an enjoyable way of tracking performance. Whichever protocol you use, it is really hard work in an environment where you’re not really stimulated to set a good performance.
Let’s be honest with ourselves, lots of amateurs use FTP as a bragging number and tend to skew the results of an FTP test on the high side. Which is not a good thing for your training because rides that are supposed to be easy, become to hard.
Per Trainer Road. “ It estimates the highest average power you can sustain for one hour, measured in watts.” “ FTP marks **the power or effort that you can hold where your lactate levels are being used and buffered effectively and not flooding your system”
Every test under an hour is simply a method attempting to derive the results of an hour test without spending an hour. The 20 minute test became useful because of its close correlation to the 60 minute test. But, that’s wasn’t short enough. People complained and once the ramp test was introduced, that took off.
Blame it on marketing or competition, but imo fear of losing customers compelled all the trainers to have a quick and easy ramp test. It isn’t accurate and each organization has a formula to attempt to make it so. I imagine somebody is sitting down at their table as we speak and having a laugh rationalizing in as complex and convoluted a manner as possible their ramp test model just because lazy cyclists don’t want to spend the time needed for a reliable ramp test.
So many cyclists are focused on metrics down to the watt or heartbeat that this was bound to happen. With a lot of emphasis on being exactly in the zone and measuring improvements in minuscule amounts, it shouldn’t come as a surprise that a quick and easy ramp test was developed to cater to those cyclists.
Trainerroad is wrong about the time limit. Ask Coggan himself, I’ve interacted with him at TR’s forum.
The problem is most people conflate the estimate with the thing they’re trying to estimate. FTP is a actual physiological phase change and can be observed, and much more importantly - FELT, even with power-only data. That’s why CP testing is one of the 7 deadly sins. You can closely estimate long power with short power tests, and at the very least you can iterate onto your CP with further workouts.
We focus on metrics because we want an accurate representation of the work we’re doing. Accuracy is important. It becomes important if you’re tracking training load with any degree of precision and using your power data for performance modeling.
Most people have no real experience with experiment design, data collection, and interpretation and that’s why cycling perform assessments across the platforms are a mess. People don’t understand what they’re looking for.