Is the FTP warm up too hard?

My non-Zwift FTP estimate put me at around 256w. That consisted of a lap around Watopia to warm up and then 20-minutes with Golden Cheetah for an average power measurement. I assume that is accurate because Zwift recently corrected my FTP to 252 based on my riding (I guessed 200w when I signed up for the beta). 

Zwift’s short FTP warm up had me do 275w for 3-mins, 300w for 2-mins and 330w for 30-seconds. That works out to 290w average for (what could have been) 28% of my FTP time.

My limited research said FTP warm-ups should be light, but this burned me out.

I posted this question in the Zwift forum on Strava too. There seemed to be a mixed bag as to the theory with such a hard warm up. 

A month ago the first ftp (shorter) test I did on zwift put my result at 303w. Well today (after a month of training) I tried to redo the test and got a result of 296w but the warmup was a lot harder - it had me doing 350-370w for 5 mins right before the 20 min test, which I found hard and I’m sure that had an effect on my result. Maybe I should have lowered the ftp slider a bit? :/ 

I just did the FTP shorter test again and took the “warm up” intervals somewhat easier than the last time (which felt flat out!) and managed to average 14W higher in the actual 20 min section for 1 bpm less average pulse. So I did think those warm up intervals were too hard for me the first time and was struggling before even started the actual test. It seems to me that how you do in these prelims is ignored anyway for the FTP calculation and it is all about the 20 mins. However even though you take 95% of the watts for a 20 min effort to estimate an hour one, are you actually meant to be somewhat tired before you start the test? Sounds illogical however to me as isn’t that the point of the 95%. You could of course just ride for an hour and take your average power as your FTP but a bit extreme!

Exactly. I did an FTP test a couple weeks ago and improved quite a lot. I had a rest day before, did it with ERG mode enabled, and the warm up intervals were spot on. This time, although I had two full rest days, the intervals were so hard I didn’t even get to the 20 minute section! Are these intervals supposed to be like that?? Is it cheating (is the result inaccurate) if I just turn off ERG mode and ignore the wattage targets before getting to the test section?

I haven’t done the structured test since my report above but did have an FTP improvement following one of the KISS 10 mile time trial races which I think is an ideal way of testing yourself as I found my focus was more on the “race” and not just staring at a number. With no drafting on TT bikes it is a “fair” effort too.

I would not do an FTP test in ERG mode as it kind of predicts the result before you do it and anyway I found some “strange” things happening in ERG mode on my trainer (Tacx Genius) when I tried an interval session. Wattbikes don’t do ERG mode anyway.

Ultimately you are riding hard for 20 mins to predict something that is actually meant to be a metric for approximately an hour. I then look at the computed FTP and think to myself could I really keep that going for an hour. Might need to enter some Zwift races around the 40K mark to check this out!

Another factor is repeatibility. When I did those first FTP tests, they were actually on a Wattbike at the gym. I am very tall and even bought an extended seat pin so can get to the right height but then I find myself so far back that I just don’t feel am in an A1 optimum position. It is handy to use them at the gym however as are right next to the treadmills when want to try Zwift running and I don’t have my own treadmill. When I use my own bike on the trainer at home it just feels “better” and indeed the watts I can put out on the trainer on Zwift are more similar to what I see on my TT bike with a PM on it. I think I might “lose” 20-30W on the Wattbike for what seems a comparable effort. Those tests on the Wattbike were trying to do the “warm up” intervals at a FTP level from my own bike and perhaps not surprising I struggled.

Given the fact that Zwift only seems to “care” about the 20 min effort then I think that is the focus but ultimately you can become good about doing 20 min efforts. It is a bit like saying I want to know how well I can do in a 40K/25M TT but I only want to ever do 10M TTs.

Of course for any hard effort you need to do a decent warm up. I would not want to do 3 mins way over FTP as part of that however. I did discuss this in the context of the FTP test with a few people however and the consensus was that the hard intervals are meant to introduce some level of fatigue to make the 20 min simulation that much more realistic.

I saw on GCN they did an FTP test on Trainer Road the other week. I only recall the 20 min bit however. Trainer Road says “During an assessment, riders first follow a consistent, structured warm-up before riding at a challenging effort level for an extended period of time, e.g. 20 minutes or 8 minutes.” Anybody got an idea whether their “structured warm-up” includes a 3 min “very hard” effort?