I may be a bit behind the curve in this. But this just popped onto my Twitter feed. A little light reading for me on holiday.x.com
2 Likes
So you had these experimental time trial data sets and used the 3 different models to construct the PD curves shown in table 1.
But as probably not all data points perfectly match the modeled curves you could use that
“MPA least absolute deviation regression.” (aka a breakthrough) to retrieve the adapted curves shown in Table 2.
Did I get that right?
And in “the model fit that included (PI) the intermittent data collected during the Pmax test resulted in a lesser adjustment of CP and Pmax in 3PMod”
That was a kind of 15s max test I guess where the hightest power reading was assigned to Pmax but all the surrounding readings are the intermittent data?
I would have expected some experiment to proof which model is the superior one, e.g. by predicting task failure in another TT test run or so. Why did you not include that?
Sorry if that´s too naive questions, I just tried to understand that starting from 0.
The complete scientific paper is yet to be published. This paper is just what was presented at the Science and Cycling conference. I suggest wait for that paper as it will answer many of the open questions you and others may have.
1 Like
On a slight different note do you know why my original post was flagged as advertising? I’m not bothered,just curious.
No. Might be because we added some more screening filters given all the junk that’s been spread on forums.
1 Like