Ideas for setting up 'Same-Four-Plus-Minus' ?

[tarrazu]

Per @bbstats "The year is 2016. We still don't have Same-Four-Plus-Minus, even though it's the most useful +/- stat I can think of."

For Player X, find each 4-man-combination that Player X has played with, and its +/- per 100.

Any ideas on setting up the framework within the glmnet RAPM style approach discussed in previous threads?

[J.E.]

You don't need to do anything with glmnet, nor do RAPM-type computations to derive same 4 plus minus
Also, it's a bit more complicated than finding each 4-man-unit and its' +/-

I would explain how to compute it, but I'm hoping someone can dig up the old thread

[Crow]

viewtopic.php?f=2&t=7887&hilit=same+4+man+plus+minus+JE