I haven’t found anything here on the forum so I make a request.
Will it be possible to have in the fields of the activity analysis an intensity field which was calculated with the normalized watts instead of the average? Similarly to the trainingpeaks one. I was quite used to that metric and I don’t find the avg/FTP as useful to that one. Specially for activities like mtb.
Isn´t that simply IF, intensity factor?
Yes it is. But I didn’t see it on the fields options.
NP on this site is calculated on a 30s rolling average from the start of the ride, not the start of the interval. Consequently you will get an accurate IF for an entire session but not an accurate IF based on NP for an given interval within that session
Yup, understood. But the value labeled “Intensity” for an interval is calculated using average power for that interval, not NP.
But you can see the NP for any given interval you chose so I guess it oculd be done the same for the intensity factor. It doen’t look odd or any diferent from trainingpeaks when you look at the NP of an interval.
The NP of the interval isn’t correct though. NP is not based on a 30s rolling average of just the interval on this site. It is on TP. Here it’s based a 30s rolling average from the start of the ride so no NP from any ‘isolated interval’ is correct on this site.
Seems like there are two potential fixes then 
But how big is the diference? The longer the interval the lower the difference you get for the calculation then I guess, shouldn’t be the same value almost for efforts >30sec? Is it a couple of watts for a 5min effort or what?
Very little. A couple of watts in a 10 minute effort.
It’s a minor annoyance for me is all. A perfectionist thing I suppose. If I’m doing a 5 or 10 minute effort I want the NP to match the average watts. That way I know I’ve done a pretty perfect effort. They won’t match on Intervals.icu because of that 30s average from a recovery interval affecting the ‘work’ interval NP. That’s all. Just so used to TP I guess.
Ah okay, then I don’t think it is an issue. The powermeter accuracy might be lower than that.
Yes, it’s just different. It doesn’t mean TP is better, you’ll get used to intervals! 
jaja
Hey @david , do you think this could be implemented? (including an intensity metric which is normalized watts/FTP in the fields section, additionally or instead of the already existing one which uses the average power, like the one in trainingpeaks)
I could add something. A question: How does Training Peaks calculate normalised power for an interval? How do they handle the 30s rolling average problem?
Perfect thanks!
I think they just start the 30s rolling form the beggining of the interval and don’t show any value for the first 30s of the interval and the NP values is lower than the average always for the first 1-1:30min of the intervals, after that the NP values start making sense (if that’s what you are asking)
Hmm … then the Intervals.icu approach of using the 30s of the activity prior to the start of the Interval is probably ok.
I think so, yes.
I don’t agree.
Look at it from this point of view -
4 x 5 VO2 Max efforts.
You’re going to need to hit the power bang on from the first few seconds of each interval. When you finish your session and look at your data on Intervals.icu you’re going to see massive differences in normalised and average power but you think in your head you’ve been steady and did a good session.
I looked at a session I did to compare both sets of data.
10 warm up, 4x5 w 5 rec + 20 mins SS at end. The difference in data is quite a lot.
Intervals.icu
Training Peaks
The power was consistent across each interval. I don’t have ERG mode so I want to look at the data after and see that my normalised = my average. That way I confirm what I felt was a good session (or the opposite ).
If I look at Intervals.icu and see such vast differences between NP and Ave power I’ll be left wondering if I actually did a good session or did it just feel that way.
Yeah, it seems like you have to choose between NP being undefined for the first 30 seconds or NP being inaccurate because it includes data outside the interval.
One time I added Lap NP to my Wahoo display and found that their calculations begin by including the 30 seconds prior to the start of the lap. If that preceding time is a rest interval, it makes NP useless for a short interval like VO2max. So intervals.icu isn’t the only place this issue comes up. 
If you are doing intervals on a trainer which are constant I don’t see the point in looking at the NP value. The NP is meant to be usefull when an effort is very irregular or with many non pedalling points like criteriums, CX, XCO, … or if done outside and on long intervals.
I don’t know what happend on the second interval with a 24W discrepancy, but in my experience I haven’t seen many times when NP<average in intervals.icu with intervals longer than 3’. I guess the interval on your trainer starts on low power and once it has begun starts ramping up. But as I said I would just look at the aversge in this case.