well i am still working on squeeze spots and the math behind it. does my math makes sense?

SB squeezed , initial raiser 4 bet, caller folded and sb is about to 5 bet jam.

3 - 3 - 13 - 28 -100

Before sb jams the pot is 45bb and he is about to risk another 87bb. assuming SB jams a hand with 30% equity vs my call Range i came up with this.
pot = 204bb - rake (10bb) = 194bb * 0.3 (V equity) = 58.2

villain share of pot 58.2bb - Villain risk 87bb = -28.8bb

that means villain actually risks 28.8bb to win 45bb.

does that sound correct ? because snowie is folding in this spot 67% of his range...

I'm only going to address the first part for now. I'm not very good at maths, but I think 3 + 3 + 13 + 28 = 47, not 45.
There's also another 1bb of dead money from the BB.

well one of the 3 is within my 28 and the BB is already add up

1(BB)+3(cold caller)+13(SB)+28(hero) = 45

It’s a bit confusing as to what question the math is trying answer and who villain is. The profitability of a 5 bet shove is based both on the equity of a two way allin and the percentage of a fold to the five bet. You seem to just be reframing what the risks/rewards are for an allin, which is still incomplete without estimated fold %.

Anyway, in either spot it seems wrong that snowie would fold 67%. That would make shoving pretty much any hand in your range profitable.

Villain is SB and Squeezing / 5 bet jams

Hero 4 bets. i am looking for MDF vs the 5 bet assuming villain bluff jams Axs.

according to my math villain must defend over 60%. snowie however is far from that, thats why i am asking if i might have missed something

Kinda the whole point of MDF is just drawing a lower bound so you don’t even have to pay attention to equities (pretend villain has 0%), but I guess a pseudo-MDF where you have 30%ish instead of 0% could be of use. Off the top of my head having at least 60% defense rate seems right. Not sure why snowie is so off. Keep in mind it’s not like Snowie is an actual GTO bot that solved holdem so this may just be a mistake. Another possibility is that the other player in the pot severely limits the correct 4bet percentage. MDF isn’t something that has strictly proven to be correct multi-way, but it still seems highly likely that it would be in this spot.

My eyes glaze over every time someone tries to use MDF to "solve" pre-flop in a multiplayer game.

The 4-bettor calls with hands that are profitable vs the jam. MDF has pretty much nothing to do with it. You just need to work out a jamming range for SB and see which hands in the 4-betting range make money as calls.

If you don't believe that MDF fails pre-flop, just look at some 10bb Nash push-fold charts. If the SB jams 6.66 x pot (10bb into 1.5bb), the BB still calls a LOT. Because it's profitable to do so.
If someone only 5-bet jams KK+ for 100bb, you don't calculate the MDF number when you've got jacks. You just fold.

Along the lines of what Arty was saying, you can construct your 4b/5b ranges from the top down.

My top range pre is KK+/AKs. I reserve this for regs that are known to jam or call with AKo.

Everyone else gets KK+/AKo.

You may notice there is no balancing steal combos.

The AK combos are weak enough, and get calls from JJ or worse, often enough.

When you face this range, there is no MDF. You probably call QQ+/AK and call it a wash.

Yeah, to put it another way, MDF is supposed to lead to a strategy that "prevents villain's bluffs from making money". But the thing is, 5-bet shoving ranges in multiway pots don't really contain any 'bluffs'. AK is sometimes in there as a balancer for QQ+, but AK is not really a "bluff", when it's got so much equity vs whatever calls.
If it was profitable to 5-bet jam ATC vs good players, everyone would be 5-bet bluff shoving all the time. But 5-bet jamming anything much worse than QQ+/AK just means you get owned when villain snaps you off with his monsters. i.e. your bluffs lose money even vs tight 4-bettors that 4-bet/fold about 50% of the time.

The whole point of using MDF for range building is that you're not trying to use villain range estimations. You're taking a GTO approach not an exploitative approach.

That's not MDF failing, its just you not using it. If someone 5bet jams on you for 100bb, you absolutely can calculate the MDF, but you can decide not to use it if you have some sort of read that your opponent is tight and only has KK+, or even if you have a general population read that people only shove something like QQ+, AKs+ in that spot. If you were *completely* readless (which is really impossible since population reads exist), you wouldn't want to construct a range that doesn't adhere to MDF (5bet shoving 27o is +EV).

But if there's a 67% fold rate, Snowie is effectively saying that there should be bluffs with your entire range. And a logical consequence of this is that there would be 0% fold percentage with your range since bluff shoving is always better than folding. That seems extremely unlikely to be true.

Its really just the simple question:

Why would you ever put yourself in a position where its profitable for your opponent to just shove his entire range?

And the simple answer is just you don't, there's something wrong here.

Its also a possibility that the discrepancy is based on how blockers work, and I think exact preflop positions might be relevant here.

If initial raiser is UTG 10 handed for example, it might be the case that the preflop 3betting and 4betting war is tight enough that we're mainly just valueing AA and using Axs as a semibluff with a blocker. So when you face the 5bet, you're mostly just calling with AA, occasionally folding with Axs, and perhaps rarely calling some good Axs, which might adhere to MDF. But when you actually look at statistics, since villain has AA a lot as well when he 5bets, the AA portion of your range is reduced to the case AA, meaning you mostly have and fold Axs when you end up in that spot.

Just a disclaimer, I'm not actually using Snowie here, I have no idea how Snowie actually displays range percentages or if OP is reading them correctly.

i just went trhough that scenario again.

CO raises 3bb, btn call 3 bb. SB 3 bets to 13 bb a total of 8 % (including any Axs). BB folds and it is back to CO. Snowie wands to Jam here as a 4 bet, but i force snowie to 4 bet to 28 bb (0.5pot) which it does with 19% of its range (calls 7% and fold the rest). after the 4 bet Btn folds and now the SB calls 18%, folds 18% and Jams 64% (including A5s).
Co now folds 67% of the time and only calls QQ+, AKs


A5s has 29% equity vs QQ+,AKs - making it a very profitable 5 bet bluff. right?

My argument is that MDF is not a GTO approach.
GTO is concerned with EV, not with frequencies per se. Pre-flop, you are not in a polarized nuts/air vs bluff-catcher scenario. You are not trying to making villain indifferent to bluffing. You're trying to maximise your own EV, while villain is doing the same.

Look again at the Nash push-fold charts. At 10bb deep, the SB is jamming almost 7x pot, and yet the BB calls about 37% of the time. MDF is meaningless there. (MDF vs a 667% of pot bet is just 13%).
SB jams T5s because it's profitable (or at least breaks even vs the best counter-strategy) and BB calls with Q9o for the same reason.

Prior to the 5bet shove there's 3+3+13+28 in the pot. So with a 5bet shove (villain, sb) is risking 87 to win 47. Even with absolute zero equity (when called), this shove is profitable against someone folding 67%.
.67 (47) - .33 (83) + 0 = 4.1

That's why 67% is violating MDF. It's pretty much impossible for that fold % to be correct. With equity (even 15% with 27o let alone 30% with Axs) its just much more obvious that that fold percentage is wrong.

So given that folding absolutely anything is incorrect, why is Snowie saying we need any fold percentage vs the 4bet at all?

There's something seriously wrong here either with Snowie or your interpretation of its ranges.

Apparently so. I've probably done the maths wrong, but when I tried doing an EV calc taking into consideration the pot size, fold equity and hand equity, I get an EV of +21bb, which sounds way too high. I'd guess Snowie itself thinks A5s is only worth about half a blind as a jam. What number does it show?

I don't get the point of your example since there's no violation of MDF. If you're saying MDF only tells you partially what you should do I absolutely agree.

Issue here is more like if some software tool is saying to only defend 8% vs the shove. We could point out that MDF isn't necessarily perfect, but I think the much more likely scenario is that the tool is wrong.

Pretty much.. 0.23bb for A5s.
And +21bb agrees with my calc

Imo, you are confusing mathematical methods of balance calculation (which have always been designed for river simulations) with the real world actual combinatorics of a deck of cards.

If you are known to 5b shove72o, they will just wait for QQ+ and snap you off, and the whole table will revolve around you and your preflop jams. Actually, they will likely only 4b these exact combos and flat the rest, arriving at the flop with a stronger range than what you are 3betting. Now, you are oop or multiway ip and holding 72o. Not going to end well.

Not sure what this means, I assume any poker calculation would use actual combinatorics of a deck of cards.

I mean yeah, 5bet shoving 72o is bad and very exploitable but that's kind of my point. There's no way a bot, which doesn't adjust to player tendencies, should ever be choosing a strategy that is so obviously vulnerable to such a spewy play.

the Problem is, as Long as i am overfolding (like snowie would be in that spot as far as i can tell) i really dont know what Villain 5 bet shoves here.

the reason for me to bother with that spot is a Hand i played against a "reg" the other i day. he opened from the BTN 2.5bb, SB calls, BB (Hero) Squeeze to 10bb, Btn 4 bets to 25bb, SB folds, BB jams and Btn calls with ATo.

I think he didnt really plan to 4bet-call ATo. i guess it was just a spur in the Moment tilt reaction. I also noticed that quite a few "regs" on my Limit (i Play 25z btw) squeeze a bit to wide and overfold to 4 bets. for now. so my thought here is, When villain has a lot of crap in his range he can either fold, call or jam. calling with crap sucks so it is mostly fold or jam. so when villain start with jamming too wide with his crappy range i dont want to overfold.

can we somehow agree on the fact that i should not 4bet-fold more than 40% in that spot in order to make 5 bet bluffs indifferent ?!

No. My reason is that discretized atomic ranges of cards that can be equal and have blockers do not behave that way, preflop.

MDF and indifference is a close approximation of later streets. Does not work preflop, or flop.

Sorry, I am not going to budge.