Going back to my fundamental model of an outcome (skill + random variance = outcome)
My issue is (and this is what's changed for me as a football observer over the years), I don't care how good the defense is, RV/"luck" is asymmetrical on the defensive side of the ball. When that skinny tail variance expresses itself on defense, the opponent scores. Springs slips.
As a general strategy (not talking about in game tactical uses of tempo) therefore, I would always optimize to play as few defensive snaps as possible if I had the better team. The surest path to victory in football is to have the per play points advantage because of your skill and to have more offensive plays relative to your opponent...not just absolute, volume more plays because you are actually increasing your odds of winning by limiting their chance at luck.
In general, every offensive snap is a chance for you to express skill and suppress luck. Every defensive snap is an elevated chance for the lesser skilled team to get lucky. It's like giving them more lottery tickets or playing Russian Roulette with more bullets, not fewer.
So anyway, you clearly understand it so that's my only actual point on the risk mitigation/constraint/game theory side of it.
My main argument is that people are confusing the undesirable outcome with the approach and this is the real mistake. Not once have I seen anyone who says "go faster" give equal mind share to the very real possibility that you fail and just end up punting faster.
It's a very common cognitive bias buy essentially it's if they just did this we would have won.
they get x = desired outcome anchored and give zero credence to the possibility of failure of x
It's why the Spinal Tap bit on "these go to 11" is so funny, it hits close to the mark on the truth of the human existence.
NT: "If we sped up, we'd have more chances"
Marty: "What if you sped up and just punted faster?"
NT (long pause): "If we sped up we'd have more chances"
