Analyze sticky behavior

[jkbhagatio]

In social experiments, we'll probably go ahead with sampling patch distributions completely randomly for each new patch at the end of a block, such that the same patch could have the same distribution (e.g. H-H) in successive blocks.

Given this, it's important we try and analyze/quantify sticky behavior, and the effects of sticky distributions on behavior (e.g. in a time period where the same patch was consecutively H across blocks, was performance better than in other non-sticky block sets?)