It's not easy to measure, as if you win the first half by 20 pts, the 2nd half may not be a good sample. And coaching to to hold a lead may not be the same as coaching to run up the score. That means 1st half pt differential vs. 2nd half isn't necessarily the story.
Has someone come up with a corrected metric?
I remember this even being talked about during his Cincy years.
Also, being well-prepared going into a game does not imply a coach is a "bad second half coach" - just as the 80s Lakers actually were a good half court team, contrary to bad takes implying "great full court team implies bad half court team!"