You wrote there's a 50/50 chance about the coach being as good or better. And there's no real rule on that. There's no magic rule that because things were one way that they'll fly back.
I'll quibble that I understand and agree with the "rule" although it's not a rule.
College sports being based prestige / recruiting, which tells you what your "baseline" talent level should be, means that a program average has a lot to do with how recruits view your school and how likely you are to land a specific quality of player. And then you're measured based on what you do with those recruits.
Almost by definition, if you're underperforming per that baseline, you're either recruiting poorly or coaching poorly, or both. And if you're overperforming that baseline, you're either recruiting excellently or coaching excellently, or both. (Medina would say that Holtmann is recruiting well but coaching poorly, I'd expect.)
I personally think that a heuristic (not a rule) for coaching performance is that if a coach is performing either above or below baseline, your next coach is
more likely to regress to the mean. I.e. not even 50/50, but actually more likely that a coach that follows a good coach will be more likely to be worse, and that a coach who follows a bad coach is
more likely to be better.
You can still get a coach who defies that likelihood. Purdue did it in football when they canned Danny Hope only to replace him with a coach that's in the running for the program's worst-ever.
But I think medina's point, which I agree with, is that if your coach is underperforming your program baseline, you shouldn't fear getting rid of him. If it were only 50/50 that the next coach would be better, I might argue that you might want to stick with a coach and wait it out to see it improve. But I think if you're underperforming the odds are better than 50/50 that the next coach will outperform Holtmann.
My own point that I bring up is that if you have a coach who is outperforming program baseline (i.e. Joe Tiller at Purdue), you shouldn't be ready to try to push him out to find the coach who can take the team to "the next level", because based on recruiting the "next level" is more likely to be regression to the mean than even loftier success.