Use It or Lose It

It's been remarked upon that popular positions are often supported with weak arguments, because people aren't in the habit of having to defend them. I think there's a distinct but related time-dependent effect on advocates of sufficiently unpopular positions. At first, the advocate of the unpopular position grows more sophisticated over time as they refine and elaborate their case against the orthodoxy—until they eventually notice that arguing doesn't work, at which point their argument quality undergoes a sharp and sudden decline: if there's literally no way you can win (because advocates of the orthodoxy are just going to confabulate a series of ever more ridiculous bullshit objections to waste your time), why bother putting in all that effort?

If "Because while you can select a sample from a different multivariate distribution to match a sample from another distribution along one or a few given dimensions, the samples are going to differ in the variables that you didn't select" is just going to be ignored anyway, the temptation to flip a table and just say "Because fuck you, that's why" may become nigh overwhelming.