This is one of those awkward spots which are pretty common in PKOs. Your shove is almost guaranteed to be profitable, but it will be an extremely high variance play with a 50bb stack behind you. They’ve already shown interest in the pot, presumably they know that BB will shove fairly often, and they should know that you’re likely to isolate with a relatively wide range. I see a lot of good players flat big pairs from early/mid positions when there are multiple shove stacks left to act. It will continue to get more popular as people realize how many insanely profitable scenarios it can create when everyone wants to isolate with marginal hands to hunt that short stack booty.

Hard to know how UTG+2 constructs their flatting range, but I suspect it might look something like: AA-KK(1/2 combos), TT-77, AQ-ATs, KQ-KJs, QJs, JTs, AJo.

If they call your shove with all of those hands you’re sitting at 44.4% equity (ignoring the shortstack’s hand). However, I’d expect their calling range to be 88+, AQs-AJs, KQs at the very loosest, which leaves you with 40.8% equity. Your equity goes down for every hand that gets dropped from that range (and nothing says they can’t also have QQ/JJ which makes our hand even worse) but, once UTG+2 is folding >50% of the time, your fold equity and your chance to win at least one bounty becomes more important than your diminishing equity when called.

We don’t want to ignore BTN’s range, but they’ll likely be folding ~75%+ of their flatting range and their chip stack is not as threatening, so they’re much less important. ATs is always doing well enough against the BB shoving range to more or less ignore it. Might be worth considering which cards are often blocked when BB shoves (ie. there’s often 1 less A or K in the deck for your other opponents), but that’s getting dangerously close to irrelevant minutiae.

Not directly relevant because of stack size disparities, but hands like KQs and ATs generally go up in value in multi-way all-ins because they have so many ways to make nutty hands. Against 3 all-ins playing the top 10% of hands, ATs and KQs run very close to AQo in equity. As those ranges get tighter, KQs becomes by far the best hand, while ATs eventually outperforms AQo against 3 ranges of the top <= 6.5% hands.

Combining the value of the bounties, fold equity and actual equity, it’s almost impossible for a shove not to be profitable. I can’t do all the math tonight, but remind me and I’ll show a method for approximating $EV of shoves in PKOs. It’s pretty elaborate and somewhat arbitrary, but I find the results for midgame scenarios are more convincing than the way ICMizer calculates PKO $EV.  

Is shoving the best possible play without a read on UTG+2’s flatting range? I don’t know… probably not, but it’s close. Some of the incentives to gamble in PKOs are illusory, a lot of your EV will still be survival-based (late game bounties get huge), but you still need to be fighting for bounties throughout the tournament.

I’m probably flatting and folding to UTG+2 shoves (calling BTN shoves if UTG+2 folds), but my strategy involves knowingly sacrificing $EV in some preflop situations because I rely on my opponents to make more mistakes than me postflop.

I think I oversold this. The $EV found with this approach is extremely arbitrary and absolutely not real. You’ll also never be able to do this in game, and rarely will you have accurate real data for doing it post game. Why the hell would you ever use it?

I use it to differentiate between hands that are otherwise similar (in terms of what you’d consider in a normal freezeout). The main crux of it is approximating how much our ‘future bounty EV’ changes after the hand. In general, the results encourage more aggressive gambling at stack depths 20-30bb (future bounty EV skyrockets after doubling our stack) and more conservative play at 50+bb (getting huge stacks in preflop is rarely much better than it would be in a freezeout).

It’s likely to be more confusing than helpful without an example, which I don’t have time to do properly right now (my other post took longer than I planned for). I’ll get it up Monday or Tuesday but I wanted to temper expectations in the meantime. It’s not something that will give you precise answers or a revolutionary way to figure out the perfect approach to PKO all-ins. It’s merely a tool that might be useful as a part of your poker toolbox. Working through it a few times has helped me have more confidence in choosing my spots to gamble for bounties (and better recognize when and why my opposition is making preflop mistakes), but it’s only a small piece of the puzzle.