ICM Spot, Can I ever call?

S

Semprini

Rock Star
Bronze Level
Joined
Jun 26, 2026
Total posts
235
CA
Poker Chips
98
Casino Coins
0
  • #51
monkeytilter said:
Still wrong.
We can calculate the EV of bluffing any flop - simply assuming we just lose any time we get called - if we assign a call/fold %age to villain*.
Hand equities do not need to come into this as we can just assume worse case and assume we no equity when called!
It a simple expectation calc. and a very standard approach.

*If you don't want to assign a flop continue %age (across all flops) to villain then we can work the other way around and just do the MDF calculation:

MDF = POT SIZE / (POT SIZE + BET SIZE)

This tells us how often villain needs to continue (call/raise) on the flop to prevent us auto-profiting with a 100% stab frequency**

**I've simplified the "favourable flop" strategy to "any 3 cards" as again this is worse case, we can examine how villain must react this very naïve strategy and our likelihood of it's profitability maybe?

And what number do you come up with? Like the teacher says, show us your work.
 
monkeytilter

monkeytilter

Legend
Loyaler
Joined
Apr 12, 2020
Total posts
2,000
Awards
4
Poker Chips
438
Casino Coins
45
  • #52
Semprini said:
And what number do you come up with? Like the teacher says, show us your work.
No offence but I sense I'm teaching you :unsure:
So you do the calc. and we'll go from there one step at a time - that way we can hopefully keep this from spiralling into the weeds(y)

I'll give you the numbers to plug in to save some time (please feel free to check them and put me right if I've gone wrong).

The pot size is 12,600 (bb=6000 + sb=3000 + antes=3600)
The 'bet' cost us 9000 (3000 to complete SB + 6000 'stab').

MDF = POT SIZE / (POT SIZE + BET SIZE)
 
S

Semprini

Rock Star
Bronze Level
Joined
Jun 26, 2026
Total posts
235
CA
Poker Chips
98
Casino Coins
0
  • #53
monkeytilter said:
No offence but I sense I'm teaching you :unsure:
So you do the calc. and we'll go from there one step at a time - that way we can hopefully keep this from spiralling into the weeds(y)

I'll give you the numbers to plug in to save some time (please feel free to check them and put me right if I've gone wrong).

The pot size is 12,600 (bb=6000 + sb=3000 + antes=3600)
The 'bet' cost us 9000 (3000 to complete SB + 6000 'stab').

MDF = POT SIZE / (POT SIZE + BET SIZE)

You are the one who has to prove that limp stab has better odds of winning than jamming pre. I did my math. You do yours. I'm no fool. All that you are doing is calculating pot odds, (assuming that you don't get raised pre).
 
Lodestone

Lodestone

Rock Star
Platinum Level
Joined
Apr 16, 2026
Total posts
412
CA
Poker Chips
164
Casino Coins
20
  • #54
fundiver199 said:
Ok so GTO Wizard limp 13,4% of hands in Chip EV mode with 7BB, but realistically how much value does that actually add, when the program is jamming 64,3% of hands? And are you implementing this correctly, which mean limping 99+, a few suited connectors and a little bit of junk? My guess is, you are probably not. So you will likely be playing better by making a very small simplification and either push or fold.

Besides this is not Chip EV, and for that reason it might be profitable to jam basically any two cards, if the Villain fold as much, as he is supposed to. Whereas if you limp and get jammed on, you need to fold anything other than those limp "traps" with 99+. I will say though, that if the Villain is very passive, maybe you can get away with limping any junk and then stab the flop and pick it up a ton. But this only works against someone, who is completely clueless on how to play in ICM-spots.
This is why I default to shoving or folding. There is no way I'm going to:

1. Remember the specific 1/6th of combos to limp with

2. Play GTO in further streets given how niche of a scenario it is when jamming 75% of range is just so much simpler, likely without giving up much EV.
monkeytilter said:
I think a shove is an ICM torch with the stack configuration, my only option is to fold if a limp isn't viable.

I disagree with the payouts rewarding aggressive play here. It's very flat, 1st is only 2.5x a 'min' cash?
You are probably right about payouts. The thing that jumps out most to me is that 3rd and 4th are equal and 2nd is only 50% more than last paid.

The adjustment to go aggro should come AFTER cashing since it's very first or last once the bubble pops.
 
  • Like
Reactions: Tammy and monkeytilter
monkeytilter

monkeytilter

Legend
Loyaler
Joined
Apr 12, 2020
Total posts
2,000
Awards
4
Poker Chips
438
Casino Coins
45
  • #55
Semprini said:
You are the one who has to prove that limp stab has better odds of winning than jamming pre. I did my math. You do yours. I'm no fool. All that you are doing is calculating pot odds, (assuming that you don't get raised pre).
Now, now you pressed me on this and I've indulged you so please return the favour and tell me what the calculation shows.

How often does villain have to continue to prevent us from auto-profiting?

(I am trying to deliver a proof step by step so we don't spiral again as I said.)
 
F

fundiver199

Legend
Loyaler
Joined
Jun 3, 2019
Total posts
16,989
Awards
2
Poker Chips
1,264
Casino Coins
0
  • #56
So after all this discussion I finally got a chance to ICMize this hand. From the action provided by OP and the size of the pot, blinds must be 6.000/3.000 with a 600 ante. Its worth nothing, that this is a very flat payout structure, where 2/3 of the money is distributed on the bubble. Once its down to 4 players, $40 has already been allocated, and there is only $20 more left to play for. For comparison in a normal MTT, its only around 30% of the money, that is distributed on the bubble. So this is a format, where you really want to cash before playing for the win.

I turned on FGS (Future Game Simulation), which take into account future hands, and this actually narrowed SBs jamming range a bit. FGS3 was the maximum number possible meaning next 3 hands, and when this narrow our range, we probably want to not jam the most marginal hands, since taking even more hands into account would likely decrease the EV of jamming further. I am guessing, that the main reason for this is the two short stacks, that have a decent chance of busting, when they get hit by the blinds next.

Anyways the results. At the Nash Equilibrium SB should be jamming 82% of hands (likely a bit less as discussed above), and BB should call with 22% of hands. J8s is a clear jam winning us 0,53% of the price pool. The hand perform well, because its suited and connected and very "live" against the calling range. For comparison K2o only win 0,2% of the price pool, and we need to go all the way to K9o to find a hand, that perform better than J8s winning 0,54% of the price pool.

So if BB defend correctly, it would be a massive blunder to walk him with a hand as strong as J8s. However we also want to check, if the 22% calling range is realistic, and in my opinion it is. BB is supposed to call with 55+, A5o+, KTo+, QJo, A3s+, K8s+, QTs+. If anything I would actually argue, that many players might not find all those calls.

BB is the second largest stack, but if he call and lose here, he is left with 1BB and essentially out of the tournament. So is he really finding a call here with 55, A5o or QTs? Maybe. But I would not be super worried about getting called to wide, so I would just go with the Nash range here and maybe pass on the very marginal hands like Q2o, J3o or 32s, that are only winning 0,02-0,04% of the price pool.

However if BB actually is calling to wide like 30% of hands, then jamming J8s becomes unprofitable losing us 0,26% of the price pool. But this means, he is now calling with any ace, K8o, K5s and Q8s. And I just dont see many players doing that, when they can so easily fold here and still be in a good situation to cash. But if we have a reason to think, this guy dont care about cashing and just will not fold that kind of hands, then the best play is to walk him.

I would not limp and assume, that he is unwilling to fold hands like K8o or K5s preflop but at the same time wont fight back at us, if we limp. He does have position, and he does have us covered although only by a tiny amount. So if we go postflop with him, he basically has all the cards. And a scenario like, what happened in this hand, will be very common in my opinion. Or he jam on us preflop, and then we obviously need to fold.
 
  • Like
Reactions: Lodestone and monkeytilter
monkeytilter

monkeytilter

Legend
Loyaler
Joined
Apr 12, 2020
Total posts
2,000
Awards
4
Poker Chips
438
Casino Coins
45
  • #57
fundiver199 said:
So after all this discussion I finally got a chance to ICMize this hand. From the action provided by OP and the size of the pot, blinds must be 6.000/3.000 with a 600 ante. Its worth nothing, that this is a very flat payout structure, where 2/3 of the money is distributed on the bubble. Once its down to 4 players, $40 has already been allocated, and there is only $20 more left to play for. For comparison in a normal MTT, its only around 30% of the money, that is distributed on the bubble. So this is a format, where you really want to cash before playing for the win.

I turned on FGS (Future Game Simulation), which take into account future hands, and this actually narrowed SBs jamming range a bit. FGS3 was the maximum number possible meaning next 3 hands, and when this narrow our range, we probably want to not jam the most marginal hands, since taking even more hands into account would likely decrease the EV of jamming further. I am guessing, that the main reason for this is the two short stacks, that have a decent chance of busting, when they get hit by the blinds next.

Anyways the results. At the Nash Equilibrium SB should be jamming 82% of hands (likely a bit less as discussed above), and BB should call with 22% of hands. J8s is a clear jam winning us 0,53% of the price pool. The hand perform well, because its suited and connected and very "live" against the calling range. For comparison K2o only win 0,2% of the price pool, and we need to go all the way to K9o to find a hand, that perform better than J8s winning 0,54% of the price pool.

So if BB defend correctly, it would be a massive blunder to walk him with a hand as strong as J8s. However we also want to check, if the 22% calling range is realistic, and in my opinion it is. BB is supposed to call with 55+, A5o+, KTo+, QJo, A3s+, K8s+, QTs+. If anything I would actually argue, that many players might not find all those calls.

BB is the second largest stack, but if he call and lose here, he is left with 1BB and essentially out of the tournament. So is he really finding a call here with 55, A5o or QTs? Maybe. But I would not be super worried about getting called to wide, so I would just go with the Nash range here and maybe pass on the very marginal hands like Q2o, J3o or 32s, that are only winning 0,02-0,04% of the price pool.

However if BB actually is calling to wide like 30% of hands, then jamming J8s becomes unprofitable losing us 0,26% of the price pool. But this means, he is now calling with any ace, K8o, K5s and Q8s. And I just dont see many players doing that, when they can so easily fold here and still be in a good situation to cash. But if we have a reason to think, this guy dont care about cashing and just will not fold that kind of hands, then the best play is to walk him.

I would not limp and assume, that he is unwilling to fold hands like K8o or K5s preflop but at the same time wont fight back at us, if we limp. He does have position, and he does have us covered although only by a tiny amount. So if we go postflop with him, he basically has all the cards. And a scenario like, what happened in this hand, will be very common in my opinion. Or he jam on us preflop, and then we obviously need to fold.
Thank you for running the ICM numbers.

I think he will call jams closer to Nash. without some of the junkier 'looking' hands. So lots of Kx/Kxs, all Ax and most likely all pairs. Could be wrong but that's my read.

As I said I don't expect good ICM folds, see it too often in these games. Players see sexy cards that need playing not the ICM😆
 
  • Like
Reactions: fundiver199
Top