?????

Interesting idea! If I can spare the cycles, that would be interesting to do. Are there segmentations that are useful against adaptive movements but utterly superfluous against non-adaptive movements? (other than data age?) Vice versa?