enzyne

My whole point was that

ICMV(call and win) + PrizepoolEV of having a big stack = ICMV(call and win) 

or at least very close to it. But suppose i'm wrong, and PrizepoolEV >0, and we do gain EV in subsequent hands. In that case you have to compare

A) ICM(V call and lose ) + ICMV(Call and chop) + ICMV(Call and win) + PrizepoolEV of having a big stack

and 

B) ICM(folding 77) + PrizepoolEV of having our old stack.

Only once you establish that A > B, you can say that calling 77 is better than folding.

Dear Ben. 

There is no need for such distinction. When we talk about frequency of +EV shoves, we naturally mean ICM, that's how we calculate EV. 

I think you are wrong in assuming "ICM ignorance of our changing stack size". In fact, stack size is the only thing ICM is taking in consideration. In other words, when we "gamble" and double up, our probability to finish first goes up, and probability of our opponents to finish first goes down, so ICM does take next hand in consideration in that regard. The changing of stacks as a result of all in is the main reason we have more +ICM EV shoves if we win.

My point is, when you calculate ICM EV, you have already taken into account the possibility of doubling up/crippling your opponent. 

Also, your implication that 

EV(subsequent hands if we win) - EV(we lose if we bust) > EV(subsequent hands if we fold now) 

is completely not obvious. In fact, I think if we calculate EV using ICM they are equal. (or, strictly speaking,  - 0.04 of prize pool in your example)

Stephen beat me to it. 77 study seems wrong to me. Your 80+% figure of EV shoves has to be weighted against EV we lose if we bust. Also comparing % of hands that have +EV shoves or calls is probably meaningless, we have to compare sums of all EVs of all hands we shove.