All FTP tests are wrong
(but some tests may be more right than others)
Your functional threshold power (FTP) is something athletes and coaches, especially those that are number geeks, love to talk about. By definition, your FTP is the maximum power you can hold for an hour. We treat it as an absolute number and use it to structure workouts, calculate training load, set race targets, and measure progress. The reality is, the FTP we assign is an estimate. And in addition to being an estimate, it is influenced by how we measure it.
You may be saying, “Hold on. If your FTP is the maximum power you can hold for an hour, then if we ride for an hour at max effort, that will be the ‘true’ FTP.” Okay, let’s chat through a few things that may influence the test.
Are you well rested (if not fully tapered)? If not, you’ll likely be able to push more watts.
Can we drop the temperature and humidity? Let’s add a few watts.
Moving the test from the trainer to outdoors? Some athletes may see a bump up due to the mental challenges of being on the trainer.
Now that we’re outside, if we flatten the course and avoid tight corners, we should see a slight increase.
Move out of the aerobars/drops to an upright position? More watts
How well did you pace the effort (i.e. did you fade or have more left in the tank) and how smooth was the effort (i.e. was your variability index low)? Improve these – more watts.
There are also a few device considerations that could influence the watts up or down:
Did you properly zero-offset your power meter?
Are you consistent in which device you’re measuring from? Many athletes have multiple power meters for different bike as well as a smart trainer that provides power output. Devices can differ by 10+ watts.
Seeing a trend? None of the above influences the athlete’s fitness at that point in time, but they can all influence the number of watts the athlete is able to put out on the day and therefore what your “FTP” is.
Another huge issue with the one-hour test is that the athlete needs to be well rested (if not fully tapered) to give an accurate measurement. Also, due to the amount of stress the test itself imposes, you will likely be fried for a couple days afterwards. We are basically then losing two weeks of training to run the test. As a result, we often use a shorter, less stressful, test to extrapolate to your FTP. The issue here is that the more we shorten the test, the further we must extrapolate the result out, and the more we can be off.
So if all tests are off, why bother to measure it at all especially now that we have AI predictions of your FTP?
Three reasons I believe testing still adds value:
The tests can provide insight into your strengths and weaknesses as a cyclist.
Testing can provide a consistent way of measuring if we are achieving the adaptions we are trying to achieve.
With regards to AI, the AI is built on a model, and the way you perform may not fit well into the model.
Let’s say we have you do both a ramp test and a 20-minute test. If the result from a ramp test is coming out lower than the result from a 20-minute test, this could mean we need to spend more time doing work at your VO2 max. Now we spend two months working to increase the amount of watts you can put out at your VO2 max. Remeasuring with another ramp test tells us if over time we were able to increase your max power.
I have argued so far that all measurements of your FTP are going to off in some way. However, having an estimate of your FTP is still useful for the reasons mentioned above (structuring workouts, calculating training load, etc).
At Revel Racing, in order to set our athletes’ FTP, we take a holistic approach. The FTP we set is using a combination of performance during workouts, feedback from the athlete, race results, information from AI, and, yes, testing. Once we have a number, we can use those results to dictate training and monitor progress.
While not without its shortfalls, my personal preference is to use a ramp test for most athletes. During a ramp test, the intensity increases every minute until the athlete cannot progress any higher. The FTP is estimated as 75% of the best one minute. I like this test because pacing is less of an issue, it can be conducted with some fatigue, and it does not impose a lot of fatigue. This test is especially good for newer athletes. The downside is that because the test uses a relatively large offset, the prediction of the athlete’s FTP is prone to more error. Because of the short duration of the test, it will likely overestimate the FTP for a cyclist that is better sprinter and underestimate the FTP of athletes that are better at long distances.
Most of our athletes are focused on long distance triathlon (half and full distance), so we spend late fall through spring focused on building FTP through high intensity workouts. In the spring through the end of race season we tend to focus on building endurance and race prep.
I tend to use the ramp test for monitoring through the majority of the winter as we are building up the FTP. Sometime in the spring, I may use a 20-minute test before I set the FTP heading into race season. Where the longer test requires a smaller offset, using 95% of the effort as the FTP, it is less prone to errors due to extrapolating the result. It does require the athlete to be a bit more rested for the result to be accurate. Pacing also becomes more important here. This is partly why I save this test for later in the season. Once we have a sense of what the athlete’s FTP is, it’s easier to prescribe a goal for the 20-minute effort.
Once we head into race season, when the focus is on building endurance, I will set the FTP for the remainder of the summer. I will tend to leave it there unless I see something in the athlete’s performance during workouts or in a race that suggests it should be modified. Once the offseason hits, we will start the cycle over again. With the focus on endurance, it is not uncommon for the FTP to decay, but the athlete is now going into the next season with a better base to further increase their FTP.
Monitoring your functional threshold power is a powerful way to assess changes in your fitness over time and to ensure you are training effectively. It is important to recognize this value will fluctuate as a function of many different factors and to always be cognizant of the long-term trends. Testing is a critical component to establishing your FTP as it is a way to provide routine monitoring, but it should not be used in isolation. Other factors such as workout performance, feedback from the athlete, AI predications and race results should all be used as pieces of the equation.