Yeah, I think you’d want to throw a bunch of selected features at a model that predicts reported FTPs, using only data that are collected within a short time window of the FTP test. I could be wrong, but I suspect that such a model would perform fairly well. Would love to get my hands on some data – and some time on my hands – to try!
1 Like
Just set Xert’s Decay Method to No Decay and you no longer need to test.
2 Likes
Never used Xert and therefore I have a follow-up question: If you set it to no decay, wouldn’t that mean that your estimated FTP can only adjust upwards only? So it’s basically showing just my all-time PR for FTP and not my current one?!
1 Like
No, it will adjust both up and down accordingly.
Thanks for the answer. Guess I have to look into Xert to get more understanding of the method. Without decay, I don’t understand how it adjusts downwards. Like how does it know if I just don’t go hard or if I got weaker?
From intervals.icu I just have a similar experience to @AlphaDogCycling. When I follow a training plan without max efforts / no FTP tests / no races etc. my estimated FTP regularly went down and then got bumped up at the next FTP test.
1 Like
Thanks for your help @GeorgeAnderson!
That looks really interesting. It comes on my ‘try it out in the future’-list. Definitely a cool feature.
Personally, I don’t mind the ramp test and like the consistency in my testing protocol / knowing how exhausted or rested I go into the test. Since I do different sports irregularly in addition to cycling (running, tennis, climbing, hiking, skiing, weight training) my weekly cycling training load can vary quite a bit to my ‘body training load’. I’m curious to see how that works out with eFTP.
1 Like