I'm downvoted, but you can go look at the literature. Jump discontinuity studies for placement, testing comparison, etc. - none of them show the effect that most people on HN want them to show.