damn.. can't edit that previous message.. (and to make things worse, this is on a different page..)
To clarify, I recall all hands were turned up preflop, and a small side pot, so seat 9 must have moved in too. Seat 3 was a big stack. So, actually.. :
"I was SB in seat 1. It went raise, 5 folds, Seat 9 moves in, Btn folds, I push, BB folds, seat 3 calls."

Yes it would seem near 0.25% for this double AK to happen. Now this is not the chance that at least 2 have AK, its the chance that exactly 2 have AK and the other 7 dont have TT-AA,AK,AQ. The chance that at least 2 AK exist regardless of the others is a bit higher of course.

But it is pretty much small anyway as you know its 1.3% per player and you need 2 to get it out of 8 etc with the second now having harder time to get AK ie 0.8% chance due to removal effects, times all the possible ways you can select 2 out of 8 and times the chance the others have outside the calling range to allow the 2 AK alone with you.

Of course since getting called in general is not easy thing more like 30-50% depending on situation, so say anything between 0.5-1 % of the time one pushes and gets called they see these 2 killing each other. But if you relax to AQ,AK its 5 times larger or so. The AK,AK vs 99 or TT say however is a nice 62-63% equity which is great surprise on a 3x pot.

Yupp finally somethin that makes sense up in here

ok.. Assuming all this has some predictive capability, lets take it a step further.

you've worked on the high end of the card scale. Should the same numbers and formula apply equally to any range of 5 or 6 card values, for instance the A through 5 or 5 to T?

I'm thinking along the lines of downcards being wheel cards, or paired (or unpaired) in a lowball stud or Stud-8 type game..

Yes the logic to calculate is the same. However i need to know what game you have in mind and what hand you have yourself and what benefits you them to have and what range they have when they call etc to make any adjustments. Are you talking about say Razz for example (or Omaha low) ? Also in these games its harder to exactly go all in in the same sense as holdem so you have no big fold equity (unless its preflop and very small stack near bubble say and they cant just call you light if its a pot bet raise that puts you all in or something but still its possible 2 call you (in such situation by the way you have also the case that one of them folds later in the hand as they bet each other in following rounds). Just describe the game situation to see what you mean.

And of course all those are small effects that probably are not taken into account in ICM calculators/simulators but are not going to alter the results significantly if at all , however they are all positive for the pusher. A typical situation like the AK i described might have the 10bb stack guy on a -1.5bb all in EV when called vs their entire range say if they have 99 but it jumps to +10bb EV (massive) if you have a 99 AK/AQ,AK/AQ 3way or one folds preflop. So tiny effect but big boost = possibly a small overall boost to equity giving further incentive to push a bit lighter in some situations if the opponents' stacks fit the idea.

I was only thinking of the potential to estimate what people's hole cards are more-or-less likely to be in, perhaps, Stud-8 Limit. Nothing more.

But if I get this correctly, your calculations are skewed by an all-in bet factor, and they wouldn't apply to Limit? Or were you just expounding on the calculation results with a follow up betting strategy...

Actually, I'd much prefer it if you described a stud-8 situation where the calculations might apply, and then I could study that. Otherwise, I'm likely to ask about some completely impossible situation, and then wouldn't learn a thing.