Just to prove a point, here is my 30s average power with Jacques last night on London Classique, you can see that even across the board the average ebbs and flows.
Where is that noise coming from?
That’s coming from people joining and leaving the RP, forcing me into the wind or braking me.
The final 1/4, or lowest average here, is when I was alone with Jacques completely. RP’s are not consistent, and people joining and leaving alter averages a significant amount.
I could break this data out lap by lap, but I won’t have a clear idea of how many were in the group or not… but it should be very apparent that the far right 1/4 of the image is quite a bit lower in power than the far left.
If I recall, the group at maximum only got up to maybe 5 people at a time, majority was with one othe person, but that final bit as stated was solo with the bot.
For the heck of it, I split out rough laps (this isn’t perfect but just for actual number’s sake) at roughly 8.5 minutes, just going off my spikes for the climb portion, so not actual lap points, but… close enough for our use case here. Normalized in parenthesis.
- 176 (177n)
- 175 (177n)
- 174 (175n)
- 169 (170n)
- 173 (175n)
- 175 (177n)
- 165 (177n)
- 162 (163n)
- 159 (160n)
- 166 (168n)
- 156 (161n)
Hopefully that breaks out how inconsistent RPs are, and how much pack dynamics affects … everything.
For clarity, this was also doing as good of a job as I could sitting with the bot while testing something else, as seen in my bug report this morning:
Robo Partners don't experience braking in U turns
This was as maximum consistency as possible with this robo partner as one can get in a public world for these data points.