Just like yours. There was a curve at the beginning - first few tests - that I was getting ridiculous improvements. I put it entirely to pacing - initially, I couldn't and then I got better at it. I've been using the Sufferfest FTP test video for a few years, and I find even now that the first time I test in the fall after doing mostly road tests through the summer, my pacing is off.

Cheers!

---

Munq

Darn, so I haven't finally managed to make huge off-season fitness improvements?
That's very disappointing! :(

A bit from column A... a bit from column B... But look on the bright side. The more-accurate FTP means more effective (and hurtier) workouts going forward.

---

Munq

Yeah, the first thing that occurred to me when I saw the increase was "Oh oh, that most of 20W added to my sweetspot/threshold interval targets. That's gonna hurt!"

I found small variations depended do the day, I could do FTP assessment and variations could be up to +-5% on any given day. If you rest 2 days before the FTP have good sleep and proper food, test will have different outcome when done after heavy week, less sleep and not the best choices of food. If you keel all variables close, that threshold will be smaller, but still exist.
There is no set/fixed FTP number it is threshold, that number means nothing, it allows you to set mathematically driven zones that again mean nothing, but allow you to have sense of consistent training. You are measuring performance and that performance depends on your physiology which is affected by so many factors and is not constant on day to day basis.

Here is best example why those numbers mean noting:

Cycling gym: everyone just completed their FTP/lactate test, they all have their zones defined and set, training starts everyone pedals based on their FTP number so in theory they are all exposed to the same load, they should be all the same tired.

I look around and what I see: One guy is puffing and huffing, other is sweating buckets, third has casual conversation when you feel your heart racing already. Who is really having best workout based on those zones???

I would say both fitness gains and pacing.

I actually test better on the 1st one inside after summer testing outside. My training shifts to more Vo2 work which helps with a 20min test and also I have no clue what to aim for power wise but go off feel. The next tests after this I usually aim too high and burn out after 10 mins in. I'm going to take a different approach on my next test. Replicate my last test until the 15min mark and then see what I have in the tank.

I had taken a break after an Ironman and was just back training properly about 6 weeks when I did the first test so I'm sure there was scope for improvement anyway. I felt like I paced the first one reasonably well. No matter how I start a long interval I always seem to sag a little just before half way and fight back after about 2/3 distance. Both my FTP tests had a similar profile but I was probably just slightly more consistent the second time. Both times I was averaging extremely close to my eventual 20min average for the first couple of minutes and then sagged, losing about 10W from the average over the next 10 minutes before stabilising and then clawing back all of that in the final few 2 or 3 minutes. So my power over the closing minutes were significantly higher power (maybe 10-15%) compared to the preceeding period. However, despite repeated attempts I don't seem to be able to hold a steady, and equally fast, pace over a long interval.

Everything you said could make sense, but can I ask where you live? Northern or Southern Hemisphere. Maybe your room also got cooler moving into the cooler temp resulting in better overall cooling. Also did you change the fan you used?

You are testing FTP.
This is not anaerobic threshold. FTP is higher as you can only hold it for an hour.
Furthermore a 20min test only approximates your FTP from the 20min test by assuming you are like everybody else and can hold a certain percentage above FTP for 20min.
You are not everybody.

Some people have very little extra over threshold.
Some people have a vast amount.
Depending on you, your FTP estimate may or may not be accurate.
The only thing you know for sure is your 20min power.

If you were measuring actual threshold, then results are usually much more consistent.
But you are not, so what you are really measuring is your state of rest.

Plenty of rest, and your 20min power goes up, not enough rest and it goes down.
Heat effects will show more at FTP type powers as well.

This is the danger of using 20min tests to estimate FTP.
If you had of run a full HR vs power vs lactate, it may very well have shown that your actual threshold had not changed at all, but that your power above threshold was better in the second test due to whatever factors.

6-8% is a pretty decent gain, congrats. The next few weeks will not be fun. And then you'll start feeling strong again and crushing workouts. And then you'll retest...

I think cooling makes a huge difference on an indoor ride... especially as the 'pain cave season' starts and I sit in a basement with no real airflow. The first few rides seem like pure torture.Once I figure out the cooling, and my mind buys into the reality of it, I can do much better. If you improved your cooling system, I'm not surprised to see a bump.

So all your explanations make sense, and I agree a 3rd test should prove which FTP number is closer to reality. If you go up another 20 watts in a week or two, send your files on to Sky's team ;-)

---

" I take my gear out of my car and put my bike together. Tourists and locals are watching from sidewalk cafes. Non-racers. The emptiness of of their lives shocks me. "
(opening lines from Tim Krabbe's The Rider , 1978

I am in the same spot as you. I started Trainer Road and am new to power training. First FTP test was 234 calculated based off a 20 min test. Did it again 4 weeks later and it bumped to 256. Did it again 4 weeks later (last week) and got it to 271 based on the 20 min test. That is a big improvement based off of 8 weeks of training. I do believe I have made some good gains in general, but a large part of those gains in the numbers were a function of testing experience. The first test, I had no idea how hard to go, and suspect I just did not go hard enough, even though it was a challenge. The second test, I had a little better idea, and the third test I had a solid target in mind. I was consistently really close to that target for the entire 20 min and ended with an avg and NP 1 watt over the target, so I consider that spot on with my goal. I was really struggling the last three minutes but holding power and when the test ended, I really felt like I gave it my all which made me think my target was spot on.

We'll see how high I can push this over the next 8 week block, but I certainly expect the improvements to diminish quickly here. I would be curious to hear what your next test results in.

When you are just starting out, big gains can be made just from the body getting used to the pain it takes to hold threshold for an hour. As your pain tolerance gets better, you are able to endure more pain, and put out better wattage results. This only last so long until you bump up against your genetic abilities.

I started using Zwift on a Kickr at pretty much the same time as I started doing Triathlon, seemed like the best way to squeeze in some decent bike training post work.

These are all my FTP tests (ever), all done on Zwift, between them I've been doing the 6wk FTP builder & some long rides on the weekend (30 to 60 miles, prep for a 70.3)


Date cp20 FTP cp20/kg FTP/kg Hrmax HRcp20
03/07/16 210 199.5 2.69 2.56 172 159
04/19/16 231 219.4 2.92 2.78 177 160
05/30/16 253 240.3 3.18 3.02 177 162
07/22/16 268 254.6 3.35 3.18 179 160
09/21/16 297 282.1 3.74 3.55 178 166

Had some time off after that, started the course again 4 weeks ago, I'd like to crack 300w, but the recent 280w intervals have me somewhat doubting I'll get there next test.

Anyway, I'm pretty sure that I've both improved physically & got better at taking the test, I can sit at my March CP20 (210W) for hours now & my HR will get no where near ~160, however I'd bet that I could've got more out of that March test if I did one a week or two before it.

 
Lower fatigue level the day of the test?

Structured test vs 20 min is a huge difference. The FTP tests on zwift or training peaks provides you the chance to warm up fully.

I have found that as you continue to attempt them they become easier.

I just did one myself using my power meter. Gained 21 watts in 6 weeks on my current program. All training I do is Sweet Spot now from 60 - 120 minutes 2 x a week plus cross train (x5 times).

For my last 3 tests I've tried to arrange them with a mini-taper, getting the cycling TSB to ~0, and overall TSB to >-10.

I think this is now at least partly psychological 'It should be easier than training as I'm more rested', but also gives me a rest before the next block & more consistant testing condtions.

Yep, I'm in the northern hemisphere (Ireland) and as I mentioned in the original post, I have better cooling now. When I did the first one the temperature in the room was probably about 16 deg C and I was relying on airflow from windows open in front and behind me on a breezy day. So the room was cool but the airflow was modest. I moved house 5 weeks ago and have now made my new garage into a training studio. It's not heated and I have a fan. I did the second test with a room temperature around 12-14 deg C and with an 18" fan providing reasonable cooling during the 20min test interval. I used to sometimes feel uncomfortably warm during trainer use but I don't remember this being noticeable during the first FTP test. Of course that doesn't mean it had no effect! I expect this will have made a difference to performance, but it's hard to quantify how much - even order of magnitude.

Thanks for your feedback.
Yep, it's true to say a 20min test to estimate a variable that's specified in terms of a 1hr period is a bit odd. As far as I'm aware, however, it's somewhat the norm. In fact, unless I missed something, Zwift doesn't provide a 1hr FTP workout (not that you couldn't do it yourself but you wouldn't have the same automated average power and updated in-game FTP figure).
It would make more sense to base training intensities on 20min power instead of applying a 95% correction first and calling it FTP, but the result is essentially the same.
I'm not about to do regular all out 1hr efforts in order to get more accurate FTP figures. The test hurts too much and would interfere too much with other training the week I did it. I can hurt myself just fine, as I see fit, in races. The motivation would be hard to conjure up for consistent all out 1hr FTP tests and this would possibly make the results less consistent. Not something I plan to start doing any time soon, and probably never.


Incidentally, my average HR was very similar on both tests. I don't have the figures to hand but I think the first was about 167bpm or 168bpm and the second was 169bpm. I would also think I was similarly rested for both having done moderately hard bike sessions 2 days earlier and no training the day before in both cases. Both were done at a similar time of day but I can't say whether my amount or quality of sleep in the previous days was similar. That's a level of detail that isn't likely to be sustainably kept equivalent by amateur athletes with somewhat "normal" lives. If that makes the tests pointless, so be it.

Well as I said. This was a 20min test rather than the hour. I'm used to pushing hard for 2x20min intervals which I do most weeks and did all last winter too. An all out 20mins is certainly harder but I had a good idea what I might be able to sustain based on the 2x20 sessions and I think my pacing was fairly good on both attempts as a result. Though it was slightly better on the second where I kept a pretty even effort until the last 80 seconds at which point I managed to raise it a bit and just made it to the end.

Pretty consistent with my experience. My boost wasn't quite as much - 11W improvement from 1st test to 2nd test, 6 weeks apart, but I had some interruptions during that 6 weeks due to illness and a couple of weeks travelling with work, so to be honest I'd put pretty much 100% of the improvement down to pacing as I don't believe I made any fitness gains in that period.

20 min testing was developed for one reason only.
Convenience.
As were all the other whacked out tests developed before it.
A full test is not easy for a home based or basic coach to do.
You need to be experienced to control variables and be skilled in interpreting what you see.
Also not many people had access to lactate meters or even HR monitors when all this started.
In a well trained, well rested athlete that is in the middle of the bell curve, the 20min test works well.
It is also easy for untrained people to administer.
The results however do not fall well for edge of the bell curve athletes or anything but well rested and well trained.
In some respects it is just a slightly better version of things than the garbage HR estimating formulas based on witchcraft and magic.

Doing a full HR step with or without lactate vs power is the only proper way to know what is really going on.
When you look at a full test, you will see in a well rested and reasonably trained athlete, that the graph is basically a linear progression up to the Aerobic threashold where you will see a small flatening as you start to see the presence of increased lactate from basal, and then the graph continues in a linear fashion once more up until the anaerobic threashold where it flattens once more.
These two areas of linear increase will be parallel in a well rested test. (testing using power or running speed, as these progress linearly)
So you will see a line, a small flattening, then another line to a flattening with the two lines parallel to each other..

Several things can happen that will change this general graph.

A preponderance of long distance aerobic training without any higher intensity stuff will flatten the first line compared to the second one.
So will not being rested, a little sick or overtrained by compressing the second section.
But if you have a good solid graph to base further testing on, you can see if it is an extension of the low output aerobic zone that is causing the flattening or a steepening of the higher output sections that create the disturbance from parallel.

A concentration of high output training will do the opposite, compress the first line and flatten the second line.
So will a simple lack of base training.
Being overtrained will also flatten the second line as it is harder to push into those harder heart rates and it may at first look like you are making good gains in the upper end with the flatter HR, but in fact, you have gone too far and need to back off.
Again comparing to a known good previous graph will help interpret the graph.

There is heaps more, but basically there is a flattening of either the first line or the second.
And you need to understand why. Understanding the phase of training the athlete is in as well as general health and state of rest is needed.
But the graph will tell you everything you need to know.

Not something that everybody has the knowhow to do or interpret but less work than trying to slam out an hour TT and it still gives you a shitload of data that all the other estimates don't even hint at.

So really, a 20min test tells you how good you can do a 20 min test.
That's pretty much it apart from some amateur sleuthing in regards state of health and rest.
So don't read too much into each individual test.

What the test is good for is if it is showing a trend over time.
Are you getting higher power each time in a nice progression?

There is too much variability in the result for individual tests to be taken too importantly as the test cannot tease out the reasons for the differing results.
You can over time, figure out what's going on with the help of other info, but really a single test tells you not much.
This is one of the reasons HR training gets a bit of a bad wrap, it's not being used properly.
You must have proper full testing for it to be relevant and a home based amateur and most entry level coaches just can't or couldn't be bothered to do it.

For anyone reading this thread:

1) FTP has never been defined as the power you can maintain for exactly 1 h. You should question the understanding of anyone who has told you that it is.

2) On average, 95% of maximal 20 min power provides an accurate estimate of FTP, but the exact relationship between 20 min power and FTP can and does vary not only between individuals, but even within a given individual depending on how they have been training.

Your statement is incorrect.

As anyone who has been hanging around endurance sports the last couple of decades can tell you, heart rate monitors were ubiquitous before powermeters were available, and many coaches had invested in lactate analyzers (and even metabolic carts) before powermeters became readily affordable.

People have moved away from such physiological measurements simply because "the best predictor of performance is performance itself" and "that which gets measured gets improved."

If that is what you have actually observed, then peoples' FTP values are set incorrectly.

At any duration beyond a few minutes, expressing the exercise intensity relative to FTP (a measure of muscular metabolic fitness) largely eliminates any differences in physiological/metabolic/performance differences between individuals.

The above is in laboratory studies we have normalized exercise intensity relative to LT (another surrogate measure of muscular metabolic fitness) instead of VO2max, and is illustrated in the data shown in Figure 2 here:

http://home.trainingpeaks.com/...the-new-wko4-ilevels