wedrifid comments on How can we get more and better LW contrarians?

[wedrifid]

You seem to have misinterpreted me (my intentions) on a few (relatively minor) points (which is probably why you don’t have a reliable model).

Doesn’t the causality mostly go the other way there?