Interestingly, eFTP / AI FTP estimates are starting to converge and I think my next ramp test will be close to these values as well.
Following the 244 AI FTP Detection and my 260 FTP ramp test immediately after, my workouts were really well tuned, it definitely didn’t feel easy at any point, I wasn’t expecting AI FTP to give me the 269 FTP today given that I had marked workouts mostly hard and two very hard. Although I had some good zwift climbs which may have helped.
I’ve got a rest week next and then my first time doing a build phase, excited to start and see the difference.
just go ride alpe du zwift and see how long you can go at 270. not that this graphing is a lot of work, but if you just did a longer effort like AJS states above you wouldn’t have to go through this whole rigmarole of comparing different imperfect estimates
I think a lot of people wonder how accurate various estimates are, AI FTP, eFTP, zwift’s zFTP etc.
I’m providing a few data points for comparison. I have the spreadsheet anyways for tracking, you’re right that adding a graph isn’t a lot of work.
Next step will be a more fully featured graph that adds my weight (loss) over time and my w/kg.
That will give some other data points for questions that are frequently asked by people new to cycling wanting to know what kind of progress they can expect, presuming they are in a similar age/height/weight range.
Precise perhaps but the ramp test is probably more consistent for repeatable results, even if they are slightly less accurate than doing a full ftp test well.
It’s really not. IMO the biggest issue with ramp tests is that the offset from ftp changes throughout the season depending on a riders ability to work above ftp. During base it could be close, during a “build” phase it could be way off.
You are always better off doing an actual ftp test.
Do you mean that since it starts at 46% of ftp (and ramps 6% per minute), that depending on the progress you’ve made with your ftp since the previous test, you’ll spend less time above ftp in the test?
Let’s say I do a ramp test during base where my ability to work above ftp is low, I am going to fatigue quicker in the latter portion of the test and get a lower ftp result.
Then if I do the same test after improving my ability to work above ftp I can get a much higher ftp from that test but my actual ftp may be the same.
Everyone’s ability to work above ftp is different and it changes throughout the season based on the type of workouts they are doing. The ramp test uses that to determine an ftp and is therefore less accurate than a long form test and more importantly that accuracy changes through a riders season.
A lot depends on what you are using your FTP number for and what definition you choose to use.
If just using FTP to set training zones in TR, it really doesn’t matter if it’s accurate because it’s continuously self-correcting based on workout feeback and adjusted progression levels. Honestly, TR could have completely removed the FTP concept from adaptive training and just tracked progress in zones independently, but FTP helps establish starting points and it’s a concept that people have grown attached to. So FTP lives on, but being in the ballpark is good enough for the purpose of determining workouts.
The only place FTP accuracy is really critical is when using it in a race for pacing. At that point, what you really care about is what wattage for a given duration. If it’s an ~hour effort, then go test at/near that duration to get your answer. If it’s a ~30 minute effort, test for that duration. Etc., etc…
If it’s for bragging rights and self-worth (all good, I’m not judging), find a protocol that gives you the biggest number.
Uh… NO. The relationship between power and TTE is highly nonlinear above FTP. If your FTP is over estimated by 10%, then what’s supposed to be a very achievable 1x30 threshold (100%) workout becomes an unachievable VO2max (110%) workout.
Adaptive training isn’t going to be able to compensate for that, because AT isn’t adjusting your FTP. It still uses the typical %FTP vs TTE ranges to set the upper and lower ends of PLs, which determine prescribed interval durations.
Yeah, maybe at 10%+ is getting too big to compensate for, but that should be a pretty obvious deviation. People operate at significantly different percentages of FTP (particularly as you get into the higher zones). When I’m highly trained, my FTP is at the very high “normal” end of my lab-tested vo2max “MAP” power. So, inversely, if I try to do vo2max training based on a “typical” percentage of an accurate FTP number, I’m actually beyond my vo2max range and need to scale it back to an appropriate level. I can either adjust that down manually and pick vo2max workouts at lower % of FTP or let AT do it based on workout feedback. There is nothing magical about using percentage of FTP to determine zones. It’s far from an exact science and the zone percentages are based on averages and always going to wrong to some degree at the individual level. AT can account for those individual differences (to a point) via progression levels. Having a “wrong” FTP is no different than being an individual who operates at percentages that’s aren’t in line with the average. Even when training in Z4 (where you could argue FTP is most critical to be accurate), you can work at lower PL’s to account for an overstated FTP and you can be prescribed higher PL’s if you are knocking things out of the park with an underestimated FTP.
i’d argue that it does matter because, in my opinion, I find lower progression level workouts (from some I’ve seen) run contrary to the intended reason for doing a workout in that domain (such as sweet spot, threshold). You’ve got coaches who will not give anything below 10 mins as an interval length for threshold or 15mins for sweet spot, because interval lengths below that won’t really help athletes meet the intended goal of those types of intervals. I find it awfully hand-wavy when people say TR FTP is fine to “just set training zones” but if the workouts people get kind of stink and don’t help an athlete accomplish what the workouts should do (build TTE, for example), then an incorrect FTP setting does matter.
Using a ramp test and having your FTP be off by 10% is actually something seen on this forum a lot for cyclists new to training with power. It happens when someone is untrained/detrained or really fresh. When that’s the case, I have very high anaerobic capacity relative to my FTP, allowing me to over perform on the ramp and get a very inflated FTP. If I’m not fresh or I have trained my FTP so it’s closer to my historic level, then the ramp is more accurate. I have the experience to know that, many people don’t.
And as far as zones above FTP, it’s well known that those vary among individuals. That doesn’t change the fact that by definition your TTE is much different above and below FTP, so trying to train SST, Threshold and VO2max is going to be pretty suboptimal with an inaccurate FTP.
What’s the point of training zones if you don’t bother to dial them in? It’s better to ride by RPE then and forget the power meter, because almost all metrics derived from power factor in FTP.
“A lot” I dont think it happens “A lot” and the forum userbase is not necessarily an accurate representation of the TR userbase.
Is it though? Sub-optimal by how much? 1% 10%, does it really matter? I think what really matters is consistency. How much do we really think is being left on the table at the moment?
Exactly. My point is that the zones are “squishy” ranges, not exact buckets. And they vary by individual and also where someone falls on their training progression. That’s precisely why we need to dial in our zones/workouts at the individual level. An FTP (whether it’s dead on or off by 10 watts) is still only serving as a starting point to start the dialing in process.
We use power as an objective measuring stick as part of the “dialing in” process based on RPE. Even with that “perfect” FTP established, your first vo2max workout of the year might be ridiculously easy (or hard) and you will adjust based on RPE to make workouts harder or easier (ie - dialing in the power/duration range you need to illicit vo2max adaptations). So yeah, power is very useful for many reasons, but RPE is critical to figure out what power is appropriate at the individual level.
