What is FTP and Critical Power
In terms of cycling training Functional Threshold Power(FTP) is considered the gold standard. Andrew Coggan, the author of Training and Racing with a Power Meter, first popularized the idea of measuring cycling fitness by a Functional Threshold Power. FTP was first defined as the amount of power a cyclist would average in a maximal 40km time trial. What a concrete definition… Surely this number has no underlying confounding factors that could affect TT time without changing power output.
In all seriousness Coggan’s model of fitness has been incredibly crucial to the advancement of cycling training. FTP has later been more defined as the power a cyclist could roughly average over an hour of maximal effort. The accepted best way to test for this is via a 20 minute test. The correct protocol for such test is a 5 minute maximal effort, to expel the glycogen stores in the muscles, to ensure that the 20 minute maximal effort is recruiting aerobic energy systems. The average from the 20 min test is then multiplied by 0.95 to determine the FTP.
This FTP number is then used to determine one’s training zones, via Coggan’s 7 zone training model(shown above). This helps determine a starting point for new riders to find their different zones of power to and rough durations to ride in to hopefully find the desired affect from their training. It is generally accepted that percentage of FTP is the standard for making workouts that transfer between riders.
Critical Power differs noticeably from FTP. Critical power(Never abbreviate Critical Power) as coined by Monod and Scherrer in their article Critical power as a measure of physical work capacity and anaerobic threshold, is the maximum rate a muscle can produce a power for a very long time without fatigue. Hrmmm, even more vague and perhaps more useless than Coggan’s definition of FTP.
In practice, Critical Power tends to be a value higher than FTP, translating to a quasi threshold where a high power can be sustained without expedient fatigue or time to exhaustion. Many athletes seem to be able to sustain Critical Power for around 30 minutes before failure.
One more metric that coincides with Critical Power is W’. W’ is the maximal anerobic work achievable before exhaustion. This metric is measured in work, Kilojoules, the close cousin to power, measured in watts. These are of course related to each other by time. Work is the total energy used to complete a task, while power is the rate at which the task in completed, or work/time. Many in tune cyclist will be familiar with kJ as a way to measure the total work done in a cycling session. W’ is the theoretical amount of work possible above an athletes Critical Power. I like to think of this as the anerobic battery, one that is drained by power done above the Critical Power and is restored by power down below the Critical Power.
How is Critical Power determined?
The typical determination of critical power is found via multiple tests. With 2 or 3 maximal efforts across different(1-20min) time periods, an equation is used to calculate both the Critical Power as well as W’. The particulars of this calculation are both uninteresting to learn, as well as being of little practical use, however if interested Tom Bell of High North has a breakdown. I’m here to vouch for another Critical Power test, the 3 minute all out test.
The 3 minute all out power test is an unpaced, continuously maximal effort test. Meaning you start the test by sprinting, then keep sustaining as high of power possible until the ending of the three minutes. This is in some terms a mentally easy test, as there is no mental energy used to pace, but otherwise a very difficult, easy to forfeit on. Every pedal revolution truly needs to be maximum. The Critical Power is then found from the average the last 30 seconds of the test. W’ can be calculated, as the total amount of work done above the ending Critical Power.
Some dots may be connecting and lightbulbs being ignited as you read this. For myself this was such a satisfying point to learn about this test, as it perfectly makes logical sense with both the Critical Power and W’ values. This test relies upon the depletion of the anerobic battery(W’) to find the maximal possible power to be sustained without rapidly fatiguing. What an elegant way to test and model fitness.
Above is a screen shot from my Intervals.icu file of which I attempted this test. The first screen shot being the real time power data over the three minutes, while the lower is a 30 second average of my power. In both, my cursor has been placed at the value my ending Critical Power to show a horizontal green line representing my final Critical Power value. It can be seen in both, particularly the 30 second average, that the maximally obtained power asymptotically slopes down from the first and highest power numbers, to the relatively sustainable Critical Power Value.
How well this test actually predicts Critical Power and W’ may be shocking. To me a seemingly simple and short test that predicts two measures of fitness is almost too good to be true. It is likely true however. Mark Burnley explains on his Youtube channel, All-Out Physiology, a great resource by the way, that in the initial study testing this 3 minute test done by his college Anni Vanhatalo, the correlation factor between the conventionally tested Critical Power and this newer method of testing was 0.99(1.0 being the perfect score). This is certainly good enough for me. In fact I repeated this test twice, on two different days, with different amounts of riding beforehand, I got Critical Powers of 337 and 340, with W’ of 14.6 and 15.9. Good enough for me!
One final note on the three minute test, and how it is tested. The test itself is quite difficult to mentally complete. It is very hard to keep yourself accountable, Mark recommends strong encouragement from the test administrator, I would practically recommend a training partner accompany you for the test. The mental tax to finish the test after the first minute cannot be understated. In practical sense, one should not see either time duration, nor power for the test, as both lead to pacing. The key to this test is fully depleting the energy stores of the body, this cannot be done properly if any restraint or pacing is used. Pacing is easy to spot in the power file as any uptick in power through the test. If any signs of this occur, the test is invalidated and must be repeated. Prepare to give a full effort before embarking on this test!
Application of Critical Power
Critical Power can be used as a metric to assess one’s medium-duration fitness on the bike, similar to that of FTP. For fitness and progression tracking, Critical Power is substituted directly for FTP. In addition to this, Critical Power also gives a W’ metric, letting an athlete track their punch, or anerobic strength as well.
Critical Power and W’ also have a completely unique application. With the knowledge of the total amount of work able to be done above critical power, if you have either a time duration or a power target, you can reverse engineer a to find the value you don’t have. For example, if I have a Critical Power 340 W and a W’ of 16 kJ, and want to have a power estimate for a 5 minute effort, I would divide 16000 J by 300 seconds, then add that amount of watts to my Critical Power. This gives me a power of 393, which seems to be relatively close to what I would expect, given my relative fitness compared to my previous peak at 415. This can also be done in reverse if you have a power target, the estimated duration you can hold said power for can be calculated.
Above is a figure showing W’ as a blue block of area, that keeping area fixed, can find where power and time will intersect along the Critical Power curve. The general equation to estimate duration given a power target is as follows: Duration(s) = W'(j)/(Absolute power(W) – Critical Power(W)).
The general equation to find power with a known time is Absolute Power(W) = W'(j)/Duration(s) + Critical Power(W).
Past this, it is difficult to find more application of Critical Power. This is because the Critical Power and W’ model only claim to model power usage above critical power and below neuromuscular strength realms, roughly efforts from 1 minute to 30 minutes. Trying to glean any useful data outside this window, like trying to find an endurance pace, would be extrapolations beyond the limits of the model. Despite this, some useful data can be estimated.
Tom Bell of High North Performance says FTP tends to lay around 96% of Critical Power for most athletes. This estimation can the be used for the standardized training metric of percentage of FTP. While this is extrapolation, I believe it can be used as a starting point for athletes to begin training off of. Side note: for what it’s worth, FTP and Critical Power are close enough to each other that Intervals.icu treat them as the same metric to calculate specific values. Again, another close enough moment for me….
Why I think you should switch to Critical Power
Here comes my view of the Critical Power vs FTP situation. I have begun to much prefer Critical Power to FTP for nearly the entire reason of the testing procedure. I recently have preferred this testing procedure as it is relatively easy and mentally reassuring to complete. In my experience the standard 20 minute FTP test while being the gold standard has both an enormous mental energy to commit to, but also carries a large amount of fatigue after the test. The amount of fatigue is self explanatory, and in my experience the 3 minute test has much less carry-over fatigue. Logically in my eyes both tests are very depleting, however the short duration of the 3 minutes limit the amount of fatigue when compared to a longer test.
In terms of mental energy, a 20 minute test is much more daunting. Not only am I committing to 20 minutes of suffering, I am contemplating pacing and expectations prior to the effort, as well as finding the ideal road and conditions to complete the test. In my eyes, a 3 minute all out test is a bit more out of my control. In a similar way to how a ramp test is less daunting, I begin the test and simply give all I have. There is no strategy to contemplate, as long as I go as hard as I can, I did it right. I would also argue that it is a lot easier to find a section of road for a 3 minute effort as opposed to a 20 minute effort, given I would prefer a steady gentle gradient with no stops for both tests.
For these reasons it is easier for me to do a 3 minute test, meaning I will do it more, which is essential to track fitness closely. As well as giving more frequent data points, 3 minute tests give 2 data points. To me the advantage is clear
Now I will attempt to invalidate all the advantages of the FTP training model. Of which I only see one, being the training benefit of using a the more popular FTP model. More specifically, if you use FTP you not only have the concrete Coggan zone model, but you have easy cross-compatibility to resources that share workouts that use the % FTP model, which is essentially the only way to share workouts. The easy way to disregard this is to recommend the 96% rule that Tom Bell noticed. I would like to recommend more nuance to this however.
I would say that for newer riders, the 96% rule works well enough. Particularly when considering that most newer riders will have lower FTP and Critical Power values, meaning that the difference between them is mathematically reduced. For example 96% of 350 is 336, 14 watts different, whilst a lower Critical Power of 250 would equate to 240 FTP, 10 watt difference. This mathematical reduction lends less error to the less fit rider. When a rider is experienced enough to have a higher power output, and there is a greater chance of the 96% rule failing, I would argue those riders should need not be structuring their training off the numbers that define zones.
Overall I believe that Rate of Perceived Exertion to be king over training to power. An experienced rider should know the feeling a particular workout should elicit, e.g. they can tell when they are riding harder than their endurance pace, not because their power meter and zones say so, but because they can feel it. I generally believe that effective training is reassured by feeling, not by power numbers. Power number are useful to collect, as they offer insight to progression or lack thereof, but power numbers from a ride should only be investigated after the ride. I view power as a way to look back on a session, not as a guide for a session. This is the topic for an article another day, but the relevant part of my point stands for this article.
In sum, I think that the 96% rule is useful for riders whom aren’t yet familiar with the “feeling” of these nebulous zones, while experienced riders gain more from riding to exertion to satisfy the intent of a workout. In my personal views, Critical Power and its 3 minute test give me an easier way to track fitness that I will personally use to more effectiveness than the 20 minute FTP test.