Hi,
I wondering if it will be possible to add an other model of critical speed based on 3 run of 12/ 7 and 3’.
See these publication on critical power determination :
And an other suggestion, would it be possible to add ascending speed on graphics analysis ?
Many thanks
King regards
Could Intervals.icu be able to calculate Critical Speed based on a work out? Sorry for the question I’m exploring all the capabilities of the tool and it’s difficult to find a training guide or clear instructions.
Yes MedtechCD, but it’s a bit different to estimate with best performance on 1K, 2k, 3k, 4k and 5k that is realize within an other workout, than 3 predefined time workouts and select that particular workout to calculate critical speed.
Yes, I understand that you are proposing a new model to be integrated.
I was just pointing @ohmax to what is available now.
If the linked papers provide enough info on how to setup the model and are free to use, david is likely to put it on the todo list.
Many thanks
Kind regards
Maxime Chamoux
Thanks I had a look at the app is that only for cycling CS measurement? Or could I input running powers? I’m a Stryd user and usually have Auto CP enabled but when I want to calculate CP manually I use the Final Surge tool provided by Palladino. But I’m happy to try and use alternative tools for my own learning and development. In particular to compare values between CS and CP.
I use the same models for bikes and runs.
I am no expert, but to my understanding CP/CS-modeling is simply about finding a matematical model that uses your known power/speed at particular timepoints to predict power/speed at other timepoints, irrespective of the power/speed has been achieved by foot, rowing or bike. Even though the models are most often based on and used for cycling, Stryd has used same models for running. Still, different models give different results, so you have to use your own expertise and experience to choose the one with the results that seems reasonable for you. I think it is more difficult to predict running power, compared to ride power, at longer durations, due to the ground impact of running.
good to see some of our work used here 
now, re to your question @Chamoux74:
you can add a power curve using individual 3/7/12 min trials (or any other combination) by going to the power or pace tab and define your desired curve:
don’t forget to choose the model for the estimates under “options”:
you will end up with the estimates based on the model used and the P/t combination:
I agree with @Nikolai_Stenfors that CP/CS modeling is simply finding the best mathematical model.
It should be noted however, that finding the best model is not straight forward. It depends on trial duration as well as the model. Another study from my group used different combinations of durations and models to find the best individual fit. Here, we included a table with the prediction errors for each combo.
oops, forgot to paste the Link:
Thanks for sharing this publication and some important nuance in modelling @Nimmerichter 
Yes I see this for CP curve, but it doesn’t exist for CS, despite modelling is similar.
oh yes I see, this isn’t available under “pace curves”
anyway, based on the accuracy of the models the data points are taken from your runs. And if you perform tests regularly these data points are used to estimate CS/D’.
Its not necessarily required to define your own curve because your test-runs are used anyway.
Yes for sure,
But you know as well as I do, that base the model on the best pace of a certain distance extract from every runs of the year is a bit different than 3 or 5 run which are perform at the maximal pace during 12 min or 7min.
