Well, actually it’s possible… It really depends on the average of that 20min section. If you were bouncing between 225 and 250 for 20 mins combine that with the average of the first and last part of the effort which looks to be well over 250… as long as that 20 min section averaged to 241 watts overall, then that is a legit 229 FTP.
But that also means you for sure can hold 229 for 10 mins (given you’re bouncing above that consistently for 20 mins correct?)
my guess is Yes. But it is just a guess. If you open the ride in Strava you can highlight that 20 minute section in the analysis tab and you can see the average watts.
You can do a similar analysis in Garmin Connect using your power data plot for the entire workout and ask it to select your max avg 20’ power. Multiply that value by 0.95 and, hey presto, you’ve got your FTP.
Here’s how TrainerRoad suggest you ride the 20’ test portion of the workout,
" How To Pace a 20 Minute FTP Test
From the start, build your power quickly but controlled, then hold it high and steady. Avoid large surges near the finish, which can artificially lift your power number, skewing your results. The simplest advice is don’t go out too hard . At the same time, you don’t want to go out too easily.
We recommend riders reassess their pace every 5 minutes during the test, and that any adjustments be just a few percent at a time. Many athletes ride these long efforts too conservatively and surge their power during the last 2 minutes; this skews the results and should be avoided. You should arrive at the last 5 minutes just able to hang on and maintain steady power.
When pacing a 20-minute FTP test, pay close attention to how you’re feeling during the last 5 minutes of your test. If you paced accordingly, you should be just hanging on. If you’re able to elevate your effort substantially at the end, it’s a sure sign you didn’t pace well".
The test is correct (the long IS the 20-min test, it’s just that the short one reduces the warm-up)
The problem is the statement that you couldn’t hold threshold (FTP) power for 10mins… when you did more than that for 20mins.
Either you’ve changed something since the test (calibration, trainer, etc) or you’re more fatigued than you were when you did the test.
Did an experiment. I hope w some important conclusions to share. I owed you fine people this, after all the info you’ve shared, and all that I’ve learned here. Again, a deep ‘Thank you’ !
Felt great, solid dinner night before, slept well, solid lunch. Tons of water day before and day of, as always. Good hard ride the eve before, but nothing nuts, legs feeling good day of. So a very normal / avg day. [I don’t consider days where you’ve rested for 3 - 5 days ‘avg’, and have no interest in that benchmark; not rly relevant, bc never happens.]
Kept W tight as poss, flat course to keep variables from gearing low as poss. [V. Flats.] Did a stepped warm up, bumping by 20 W each up to 220. Held 220 step long enough that HR stabilized. Think this is important.
At 18:30 spun it up to 229 = 230, pushed as long as poss. [Some bouncing near 5 mins due to hills & gearing, sorry!] Wasn’t ‘Pushing for 10 mins’, I was playing ‘TWO MORE MINS! COME ON! JUST TWO MORE!’, and each two mins deciding if it was Go or NO. HR was high enough I had no clue how long I had gone, srsly. Ended up just making it over the 10 min line; 11:36.
Observations:
@ 220 HR stabilized. It’s a very solid effort, felt sorta OK, and HR was doing OK at least for that short time. I don’t think I could hold it long though.
Bump to 230, HR spikes right away, towards 160. [For me: 150: All day. 155: Pushing, but cool. 158 - 160: Pushing, not so cool anymore. 162 - 163: Ok, srsly, pls, where is the finish line? I got like a few mins max here. 172: Personal record, held for a few secs.] Managed to focus up and smooth out the pedals and got it back down to 155 ish for a bit, but not long. Legs running out, HR ups to compensate, quickly over 160, held on as long as I could, pulled at 30:00.
Could I have done another 2 mins? I rly don’t think, without my head in a bucket.
So I guess you guys were right, and I was wrong. : ) I can, technically, hold 229 for 10 mins… 11:36.
An hr? No chance in that hot place.
Some have commented that my & others’ concept of what FTP means is mistaken, and maybe even a bit silly. : ) And that FTP is not your “1 hr sustainable power”.
But consider: 1 - That’s the original definition, in the study where it was invented. [Thanks again @Y.iannis !] 2 - A quick Google shows this is the understanding across much of the cycling world.
And when you think about it, isn’t your ‘Max theoretical sustainable power for an hr’ a valuable metric? I think so. Going for a ride, I’d like to know what line I can push to try and hold, to both pace & challenge myself to get stronger.
Conclusions I put to the group:
1 - FTP really does mean “1 hr power”. So if any platform wants to talk about another metric, like 20 min power, 10 min max, whatever, cool beans, but it should prob use another term.
2 - The ‘FTP’ shot out by even the full 73 min ‘FTP test’ is way high. I doubt I could hold even 215 for an hr. Maybe. Probably more like 210. So 230 is about 8.7% high. That’s a lot!! All seem to agree the ramp & 20 min are even less accurate. Wowza.
Further, and more importantly, the ‘FTP’ Zwift gives you is likely not even sustainable for 20 mins. It’s very possible your max push may land between 7 - 15 mins. So it’s simply not the case that “By ‘FTP’ Zwift doesn’t mean true FTP, they mean your ‘20 min FTP’ .”
3 - Given 2, trying to hold Zwift’s ‘FTP’ for an hour isn’t realistic. If you want to push yourself for an hour, it’s going to be much lower. I guess we have to figure it out on our own!?! LOL
And if 1 & 2 are true… shouldn’t Zwift be making some adjustments to the ‘FTP test’ ?! : )
Also interesting: It does appear Zwift is using their calculated ‘FTP’ to scale the workouts, and that scaling is much more accurate than tested ‘FTP’ being your true FTP. I’ve (finally) got up to 0% FTP bias, was using 6 - 8% for a while. I feel right on the line as I cross the finish line of the last push in various interval workouts. Not a gram left in the tank, and also haven’t vomed (yet…), so it seems right where it should be.
I hope this info is at least a tad interesting to some.
I will spend some time in the forums trying to answer newer riders’ questions, to pay forward the help you’ve all given me. Again, thank you.
This is the part I still don’t quite understand:
The FTP Zwift gives you in the FTP test is 95% of what you actually sustained for 20 mins as part of the Zwift FTP test. So you can sustain your FTP for 20 mins since you had to average above that for 20 mins in the FTP test to get that value.
Perhaps I’m not understanding a subtlety in the semantics?
Nope, you get it just fine, I’m dumb. : )
I totally missed that. Yes, obvi, during the FTP test I held 241 avg for 20 mins = ‘FTP’ 229.
But… how? I was really genuinely maxed in the experiment ride. IDK, my only guess is that because in the experiment, I did a stepped ramp over 18 mins, so the tank was partly empty, but in the test, you jump right into it off of a 5 ish min break around 125 W ?
Still doesn’t make sense though, how I could hold 241 for 20 mins. Nutty.
Frick. Now I have no clue. Is it possible that daily variance is just that huge, like can be up / down by 10 - 20% or more? That doesn’t seem to make much sense either…
if your hr is dropping that quickly after you think you’ve emptied the tank then you probably haven’t emptied the tank. generally when you are actually physically exhausted at TTE (time to exhaustion) in a given zone it will stay elevated for quite a few minutes as your power falls away.
in your graph, unless i’m looking at it wrong, your HR seems to fall quite linearly with your power. i’m not a coach or a sports scientist but personally when i see that in my own efforts it means i’ve cut the effort short on purpose or it’s submaximal
Except that it was never defined as such by Andy Coggan, not by Hunter Allen.
The question still remains, however: how can you hold ~240W for 20mins (and after a 5-min race effort, I might add!) but then can only hold 229 for 11mins? Either your FTP is wrong or something’s changed in your set up
I expect I’m a massive anomaly compared to the average Zwifter, in that my best 60mins in the last 60 days is only ~78.6% of my best 20mins.
However, I’ve only done ~20 races or TTs over 25mins this year, with perhaps half being “race pace” efforts of 40-58mins… While as an extreme example, yesterday I did three sub 15min races and a 28min race (which I’m suffering for today). I’d guess I usually do 5-10 short sub 15min races a week these days, if I can’t ride outside.
Sub 20min race intervals transfer pretty well for me to my outdoor rides, where I like to chase my previous bests up my local 250-700 foot climbs. I’d love for climbs like the ~6.9 mile ~1209 foot Road To Hell in Denbigh, which I climbed three weeks ago while visiting family, to be on my doorstep!
Edit: Devil’s home is banned word!
My long winded point is, 95% of 20mins power can be a long way off 60mins, especially if you don’t practice pacing or train for longer efforts.
I think it all depend on the type of rider you are. My max 60min power is 332w at a FTP of 351w, so about 94%.
And NO I am not interested in doing a 60min test. 
Hey Chris, I think I’ve read through the whole thread and didn’t see mention of this. Not sure what your power device was for the test, but had you calibrated it properly before the FTP test (and before all your rides)?
@William_Harrison1 Yes, good thought. Saris M2. Not wheel-off, [not $ possible for me] but quite a few reliable sources say the accuracy is significantly better in testing than the claimed +/- 5%. I inflate & calibrate before every single ride, for that exact reason.
So shouldn’t be anywhere near producing 7 - 10% or even higher variances… right?
No, but your legs might be.
A couple of Summer ago, when I was at my highest recent level of fitness, Zwift had my FTP as 317, calculated through a combination of ramp tests and event riding. At that point, I managed to get up the Alpe in 58 minutes something (my only sub-hour effort to date) holding 311 average for the whole climb.
Just a week later, did another Alpe climb and, despite my best efforts, was several minutes slower, averaging only 291 for the hour. Didn’t feel any different, prep was the same (including setting off at the same time of day) but just couldn’t manage the same power numbers.
Could be because I am well into my 50s, but it doesn’t take a lot to create variances in performance. I would imagine being able to minimise that and produce consistent power output and performance is part of what marks out a pro from the rest of us.
