From time to time, there will be fish who start open jamming and jamming vs minraises with most of their hands. Usually they continue doing it for the entire match and not rematch. I called such an all in with K 9o recently but I have a feeling, the calling range should be tighter, since villains keep on shoving and u dont see a rematch often. How tight would you be willing to call it off at 75bb and at 50bb, in case you are carddead during the 1st level?

Have you read Chapter 16 'Win-rate Maximizing Play' of Vol. 2 of 'Expert HUNLH' by Will Tipton? There, he specifically addresses the case of playing shove/fold vs a maniac who stacks off with exactly any two, and Will estimates that, when the sum of the stacks is 100 bb, Hero should be getting it in with a ~12% range if the stacks are about even (50 bb both) but wider if the stacks are uneven, e.g. ~30% if the eff stack is 20 bb (that's partly due to the pot odds and partly because the eff stack will be bigger, creating a more lucrative situation, if the stackoff doesn't end the game immediately).

75 bb eff (with 150 bb in play in total), the range is even narrower, ~6% (Hero's probability of eventually winning the SnG is then as big as 63%). K9o was definitely too weak to stack off.

The range doesn't depend much on who has the bigger stack (as opposed to the mirror situation) nor on Hero's position.

Or are you asking how the growth of the blinds affects the ranges (Will was assuming no growth)? I don't think that the prospect of the blind growth loosens the ranges up significantly, as it's not a big threat - 50 bb eff would be still deep enough to get a massive edge.

Thanks for the thorough answer. I did an Equilab calculation and a 6 percent range of 88 plus, A 10s plus and A Qo plus, has 71% equity vs any two cards. Does the 63% win % account for the times you have to fold, when waiting for the 6 percent range?

Well, I'm not sure if the range was 6%, 7% or 8% - in the book, there's just a blurry plot of the calling and shoving frequencies on the Y-axis vs Hero's fraction of chips in play on the X-axis. But anyway, yes, folding while waiting for a good hand does reduce the equity dramatically. Indeed, if you have to wait for 16 hands (the average) for a playable hand, then you'll have posted and folded 12 bb (out of your 75) before stacking off.

Depends on which variant of future game you apply. If the fish is infinitely rematching e.g. you should call with any hand that is +EV.

It was stated clearly in the OP that such fish usually doesn't rematch, otherwise there'd be no problem.

True, my bad.