Is usually end up with something like this…
It now finds the median cadence of the Z2 HR data points and keeps those within 0.92 and 1.15 of that. I will deploy that tomorrow morning to recalculate everyones stuff. Its a reasonable match for eFTP for me for the last 2 years and nice because you don’t even have to do max efforts for this, just ride easy Z2 and you can figure out how fit you are!
5 Likes
I also added CSV download of the HR lag adjusted power vs HR minutes for an activity:
start,secs,movingSecs,watts,bpm,ratio,cadence
0,60,56,147,115,1.28,43
60,60,36,168,126,1.33,63
183,60,60,118,112,1.05,61
243,60,60,140,111,1.26,60
303,60,60,158,116,1.36,83
...Out of curiosity, would it change the graph at all if you switched the zone filter to power? Thus plotting Pw/HR for time spent in power Z2.
Probably not much unless your power zones have changed i.e. FTP changed. The big advantage of using HR zones as a base is that they generally don’t change during a season.
They do as you fatigue (too much) in which case HR is either lower or higher than normal.
1 Like
This is very cool. One thing I thought of and maybe it’s already factored in or doesn’t matter, but sometimes the entire ride is Zone 2, and sometimes the majority of the ride is 1 hour intervals with .5-1 hour zone 2. I find that Zone 2 heart rate after an interval session to be higher than just plain zone 2. Does this matter? Would looking at only complete zone 2 rides show a better indication of fitness increase?
No. You will fatigue more during the intervals, after which it is not uncommon to end up with a slightly higher than normal HR, compared to Z2 only activities.
You might have a point. I am going to add more filtering options to the custom charts soon e.g. on intensity for this use-case.
Is it possible to add Normalized Power/HR in Z2 as this is comparable to the Efficiency Factor calculation which is, according to eg. Friel the basis to calculate FAeT? So I am wondering why average power instead of normalized power is used.
That is already available as parameter for detected or manually set intervals.
For the entire workout you also have pwr/HR Z2 based on avg and EF based on NP.
I used average power because there is a linear relationship between power and HR (until about threshold anyway) so this seemed like a better option than NP.
Thanks but not sure if I got that. Therefore, please let me explain further:
I want to see the EF for zone 2 only and not for the entire workout in all zones. Especially, I would like to see EF for zone 2 over all workouts on my /Fitness page and not just for the single activity. Therefore, I hope that EF in zone 2 (= NP in zone 2/HR in zone 2) can be added as chart on the fitness page.
The already existing chart use average and not normalized power.
Thanks David. I am for sure not an expert in these questions. I mainly reference to what I read in Friels training bible. Personally, I think it does not matter if you use AvgP or NP in case of constant ERG activities where both values are very similar. However, if the real effort or stress is much higher (measured in NP) in course of some workouts; especially for outdoor activities or indoor races/non-ERG rides, it may no longer be possible to compare these ride types or better may lead to a wrong chart development and interpretation.
For example: The chart decreases and tells me that my aerobic fitness decreased, but actually the reason for that is that the NP of my last rides was significantly higher than the AvgP. Accordingly, my HR was higher than in course of a series of former constant ERG-rides with a similar AvgP and NP value. To eliminate this element I believe it’s better to use NP.
What do you think, makes that sense?
Hmm yes that does make sense. I hadn’t thought about it that way. Will have to see what I can do!
1 Like
EF and decoupling make not much sense if used for none-steady efforts. HR lags, NP is very dependent on power changes… And don’t forget that cadence plays a big role too! Same power at a higher cadence usualy gives a higher HR.
The concept EF and decoupling was based on a constant effort, unless I’m mistaking.
I use it to detect small performance gains over shorter periods. Even disgarding the first two minutes of a longer steady state effort, just to avoid influence of HR lag.
Typically, if you increase Power a bit, you first see a rise in EF (good). In many cases decoupling also rises (not good). Then, when keeping power equal, decoupling drops and you know you are ready to increase power again. If you use DFA Alpha 1, this is even more sensitive and a first predictor of the fact that you’re adapting or not.
AIEndurance proposes an introduction of two new parameters based on Alpha1 for tracking Aerobic fitness:
3 Likes
Hi david!
Neat little add on, this!
I have also noticed that Intervals.icu now sort of “predicts” Power/HR in Z2 in the Fitness tab.
How does this work?
Thanks!
I have fixed this. The moving averages on things like power/HR leave out days with no data. So if there are 30/42 days with data then the total for the last 42 days is divided by 30 and not 42. This was causing it to be “projected” into the future.
3 Likes
Oh I see, cool that I somehow helped make Intervals better 
1 Like
Hi David, any news on that to add a data field for NP/HR in Z2? Thanks and regards, Martin