Sunday, April 12, 2020

I got some splaining to do for Rich Lowery

In any observable group, there is going to be an unobserved distribution of contact frequency, and it would seem potentially first order to account for that.

Hospitals have learned and hedged the unobserved distribution.  The first thing they learned was their lack of capacity for the super spreaders and called for shutdown.

But in the general case this is what Hayek did not know, we can uncover the unobserved distribution by segmentation and finding the maximal bipartite matching pattern.  This is the value chain model, he one Hayek was dealing with.  When we dump time and find the equilibrium by adaption we restructure the distributions by segmentation. This is channel packing done right and is the thing that fills the Hayek triangles. Coinbase and  Learning Tree are all about structured queues that create optimum bipartite graphs.

Hospitals channel pack, they are like hedge funds and reserve beds for the expected semi-repeatable sequences over the depreciation cycle of the virus. I decomposed the process on Kling's blog among the comments and will ot repeat it here. Go find someone to write the book on self-sampling systems, there is no book yet the math guys are just now reading the proofs (which are not my doing). The proofs tell us when channel packing will not work, and the hospitals naturally knew this when they called a shutdown of the super spreaders.

No comments: