cost of reorientation = 100*cos(pivot angle)
units = % of pre-reorientation momentum
Conversation
more accurately, if you include discontinuous scaling as a kind of pivot that forces you to resolve "solved" problems at a new level, you get
cost =100*cos(azimuth pivot angle)*cos(altitude pivot angle)
1
1
Made up this visualization of how momentum drains in pivots. The widening funnel has a logistic shape profile (s-curve) revolves around the current vector in 3d.
2
3
Mental model inspired by the thought that I pivot/reorient way too much probably, and thinking about how to get good, long runs where momentum can develop. How to find the longest "straights" you can when your operating mode is fundamentally dependent on maneuvering more.
1
1
Another way to think of it is: compound interest only works on the straights, not on the pivots. That causes arbitrary amounts of reset.
Replying to
Model is memoryless though, which is inaccurate. Going "back" to an old heading or scale should be somewhat more efficient.
