The best we can realisitically hope for is a coinflip here. We get 49.4% ev vs just the shover if he shoves a range of 22/77/AKs/AKo/KQs/KQo/AhXh/JTh/J9h/T9h. This might be conservative on the flush draws he shoves that we beat, but I don't play enough micro's to really range this 100%. I don't think he is ever just pure bluffing, and want to be really conservative on his flush shoving range.

Obviously, from a pure EV perspective, a coinflip is good and we should be putting the chips in. However, once you start to consider a GTO or a lifespan factor, it's absolutely horrible. We should almost never be looking to take 100bb coin flips in level 1 of a tournament. We really shouldn't be looking to take 100bb coinflips at any point in a tournament. If we have 20bb here, I'm all for shipping it in. Having 200bb here won't increase our chance of winning this tournament by a value significant enough to justify the risk of busting it. There is a tonne of poker left to play before we get anywhere near the money. 

Keep in mind, I think the range I've given him is quite loose. It was literally a “how good can we realistically be” scenario. The actual shove range I give V in this spot is much tighter, and drops us to about 44%. 

As played, I'd fold this and not loose any sleep over it. 

I'm not sure how much you know in regards to game theory, and I'll admit that I'm still at the early stages myself, so what I'm saying in this response is mostly subjective opinion.

Ultimately, I try to look at each decision I make inside the Meta Game we play, and how that can affect downstream play. Most of us do, because we are aware of how our opponents percieve us. Knowing that you a player is willing to take huge gambles early in a tournament makes them quite exploitable.

If I know you won't fold a flush draw here, then I'm only ever going to shove sets, AK, KQ and two pair combos(which we can assume I have none of on this board). This puts your equity at 42%. You are now willing to risk your entire tournament life, which at this point is 100bb(super deep), when you are only 42% to be able to continue in the tournament. 58% of the time, you bust out. Busting out = 0$. Doubling up however, does not garuntee us any money. So our real net result here, is a huge risk that has very little imperical gain. Especially if you are always taking a spot like this. Tournament poker is all about survival, and we do that by reducing our variance. Taking a 42% spot for 100bb is going to be insanely high variance. 

You also need to consider the fact that if you fold, your opponents won't magically think they can put moves on you. They don't know what you folded. For all they know, you could have QTss here. If they are even paying attention. 

I know I can be a bit scattered here, because all my thoughts on GT aren't fluid yet. What I'm essentially trying to get at is that you need to evaluate the potential gain versus the known risk, as well as the implications this will have/not have on future play. It isn't something mathmatically quantifiable. It will be different for everyone. It's a striking juxtapose for me, because I have very high risk tolerance with my bankroll, yet I don't like to take huge gambles in game. I guess for me, I just don't feel that the 42% of times I'm doubling up here will parlay into success often enough to offset the 58% of times I bust out – when in contrast to the success I expect if I just fold. 

What we need to consider is that there isn't any one tool that is perfect for making a decision when we are playing deepstack poker. Once you get down to 20bb, a lot of it can be simple math. At 100bb, there are a lot more variables involved, because losing the 10bb we have invested doesn't actually have a big impact on our expected return from this event. I guess a simple way to say what I'm thinking is that “2bb is worth a lot more to our 20bb stack is than 10bb is to our 100bb stack”.

I hope these rambling thoughts make sense, if not let me know and I'll try to sort them out better.

BionicApe said:

Oh, I see the real problem here. You're assuming the villain has a reasonable preflop calling range. This is probably where we diverge the most. I'm generally unwilling to give someone credit for a reasonable calling range when they make an unreasonable shove like the small blind did here. In my experience, aberrant betting often implies aberrant ranges. I feel strongly like I'm frequently going to have closer to a full 14 outs in cases such as this.

Sometimes they'll get me, but the type of player who does this with the nuts is few and far between. The type of donk who does this with a pair of 10s is much more commonplace.

The guy with TT, assuming he doesn't have the heart blocker, is still 52%. Meaning you are stacking in 100bb with a 48% draw. Slightly better than the assumed range above, but still not good enough for me to ever be calling off 100bb. 

Also, I'm not assuming he has a reasonable preflop calling range. I'm assuming he has a strong, not very bluffy flop overshove range. I guess I just don't run into the donks who are shoving T9s in spots like this. That is verly likely because the guys who do that go broke before they can make it to the buy ins I play. I also don't see many guys who snap off 100bb in this spot at my stakes either. 

The only hands I'm calling off in this spot are sets, and if I know the player to be spazzy, AK. I seroiusly just snap fold KQ. If I don't know much about the player, I'll look at folding AK as well. This changes if the shove happens on a non heart turn instead of on the flop.