Intervals.icu uses the differential algorithm. Here is the code that computes it:
let wb = wPrime
let pt = 0
for (let i = 0; i < data.length; i++) {
let t = time[i]
let secs = t - pt
let deltaW = cp - data[i]
if (deltaW > 0) {
for (let j = 0; j < secs; j++) wb += deltaW * (wPrime - wb) / wPrime
} else {
wb += deltaW * secs
}
data[i] = Math.floor(wb)
pt = t
}
I found a nice writeup here:
Good to know, thanks @david. This setting is configurable in the Garmin CIQ app so that now the W‘ Balance shown on the Garmin screen should closely approximate what I’m seeing in intervals.icu.
Great point, this could be the resolution to my query too. I also just realised I was on smart recording, I had to reset my device a while back and never changed it back to 1 sec.
Tested the new setting this evening. My icu chart is now pretty much identical to my Golden Cheetah one.
Sorry for bringing this back but wanted to avoid starting a new thread.
Is W’ calculated on the power page from the eFTP or from your inputted FTP? Or have I misunderstood and this doesn’t matter?
Thanks.
If you have the default power model selected (FFT curves + Mortons 3P) then the W’ number is based on the curve for your eFTP i.e. it is the same for everyone with the same eFTP. I am going to be doing some work on automatically estimating W’ from short efforts soon to go with eFTP.
If you have one of the other power models selected then W’ is the value derived from the model. You need to have the right inputs for this to work well.
which app do you use to visualize your W’Balance on your Garmin?
This is not true any more. W’ is individually estimated by looking at the most energy above eFTP on the athletes actual power curve for the selected duration.
Here is what I use.
Saying FTP should be “normally be higher than your CP” makes little empirical sense. Its been proven in the lab that CP and CP alone is the point that demarcates heavy from severe boundary, in other words it is the gold standard for the maximal oxidative point beyond which metabolic control is lost. FTP as a power output that can be sustained for approximately 1 hour cannot be greater than CP; you simply wouldn’t be able to sustain that wattage for that long and besides that empirical observation will not even honor the definition of FTP set out in the book TARWAPM (i.e “quasi steady state” maximal power output that can be managed for approximately 1 hour “WITHOUT” fatigue").
I posted a reply then realised I misread your comment!
So here’s a new one.
It all depends on how you define the parameters / test inputs for these two metrics. Depending on selection of test inputs for CP and means of estimating FTP, FTP may be higher than CP or may be lower. If one is defining FTP from a single test (be that 8min / 20min plus scale factor, or long form time trial / time to exhaustion effort), and includes that effort in their CP calculation, CP should be lower than FTP due to the asymptotic nature of the CP model (and you can’t get test data at infinite time). And of course if define CP from just short tests, and define FTP from one long test, they are likely to be quite different, and as to which comes out larger I think is dependent more on the athlete type (anaerobic or aerobic biased) than anything else - besides the poor selection of test values / model inputs that is!
I’ve read somewhere that a single 3 min max test is a very good value for CP but it needs to be max throughout, no pacing, so the power curve for the 3 minutes looks like a power curve, high to begin and fading away.
Yes I’ve read that too - a difficult test to do though as I think you need to be very motivated and “honest”! Would be easy to underestimate CP if back off too much after the initial strong effort, or overestimate if end up subconsciously pacing the effort / trying to minimise the inevitable pain!
You may have read in this open-access paper?
https://www.researchgate.net/publication/260375733_A_single-session_testing_protocol_to_determine_critical_power_and_W’
Yes, it would undoubtedly be very painful at the end.
The means of estimating something and the conceptual definition are two different things. We should not conflate the two. We have good evidence to suggest CP is the demarcating boundary between heavy and severe intensity domains. If you then introduce something called FTP (which has little research validation) to suggest that it is higher than CP, you’ve now just contradicted FTP and its definition laid out in the book. Riding in the severe intensity regime without fatigue is impossible. Frankly, my interest in FTP is minimal. I just picked up this conversation to correct the conceptual notion that there is a steady state exercise intensity that is even higher than CP. There isn’t one.
We can take further discussions on CP offline. Just msg me.
Might be good for us all to see this,as learning. 
Yes a fair point that the estimate / means of estimating is not necessarily the same as “the thing”. Rigour / openness of validation aside my view is that the measures of FTP & CP are trying to achieve broadly the same thing but as determined in different manners will give different answers. And of course there are a myriad of different ways to determine the estimate called FTP and the estimate called CP!
While not explicitly stated my remarks have been mostly on the means of estimating, and the most commonly used CP vs FTP estimation means, i.e. 2 or 3 all out tests for CP spanning 1min to 20mins duration vs an 8min or 20min all out test scaled by a factor for FTP.
Personally I also much prefer the CP concept to FTP, especially when used with W’ as an easy way to assess your over-threshold efforts and be able to track a more rounded definition of one’s cycling performance. And if done right you can learn a lot using CP/W’ and a calculation that can be done on a basic calculator - or by hand if you’ve got decent mental arithmetic skills. Although granted if one shells out for WKO the same can be done through FTP and FRC I understand.
I’ve done 30 second tests this way and a puke bucket was required. I can’t imagine the horror of trying a test this way for 3 minutes.
It’s not that bad you just kind of end up mashing the pedals in an ever slower way. You end up feeling as if you’re doing 10w for the last minute. Does help to have someone shout at you I can’t see a timer when doing it as it means I pace the effort.
This is an interesting paper looking at the validity of different tests. Comparison of Critical Power and W' Derived From 2 or 3 Maximal Tests - PubMed Conclusion being: shorter the test higher chance of overestimating W’. But as long as you’re consistent it ok.