I was also thinking about this being potentially good bluff candidate there, however given ICM its more complicated.
Im thinking from human perspective because of icm opponent will give this all in quite alot of credit. We dont block anything strong since i dont think opponent would play T9s this way much however we unblock KK QQ, like idk about bluffing QJs here 99 seems more likely to be good but yeah interested to see bluffing combos on river vs that line.
Its an interesting spot playing already but will run it in piosolver with ICM after session tonight.

For CEV he needs to have ~25% with a bluff catcher to be indifferent. If I have 21 combos of value that means he is indifferent calling for chips if I bluff 7 combos, not 40 combos. TT/99 are credible bluff candidates here, but T9 is a very small part of his value range and my blocker effects are negligible. I'd rather have a pair on board + straight card; JT,J9,T9,T8 seems like the best candidates.

I just ran the sim. Now one difficulty were the exact preflop ranges because while i have monker cev ranges for 100bbs, my monker icm ranges dont include stackdepth as deep as 100bbs so i had to make some manual adjustments, basically i tightened our calling range and took off all stuff that is close to break even in cev. I also tightened 3bettors ranges slightly and changed the weights a bit, like i didnt let him 3bet medium strong suited hands as much and made it a bit more polar weighted mroe towards suited kings and offsuit and suited aces (the bluff portion i mean). I still didnt make dramatic changes here though still kept in some 3bets such as J9s that are 3bets in 100bbs cev but just lowered frequencies and generally made range like 1.9%ish tighter for 3bettor and for caller i made it 3.7% tighter the calling range.

3betting range I used:

calling range i used:

the sim set up:

First thing to note is that bb is not supposed to use that turn sizing he used. Out of 33%, 49% and 75% betting options, on this turn oop bets with the 33% sizing essentially exclusively and does so 41% of the time, the rest checks (other sizings still being used at small fractions of a percent at point where i stopped sim, less than 0.2%).

Now if we still follow the in game line, vs the 10% sizng on the river (which btw now is being used 22.43% of the time total and oop jams 40.94%, checks 36.63%, so now this is a thing given options) 99 is a very sensitive hand and depending on the suits does do a mix of call and fold, call and raise and call/fold/raise....9s9c will be folded almost always, when we have a 9d in our hand we do some bluffraise all in shoves. Here are the frequencies for 99:

this is what our range in general does in this specific line used and Sam is right that JTs J9s are overall more preferred bluffjams by pio than:

now here is what we would do if on turn oop instead would have used the sizing option that pio likes, 33% and would then proceed to bet 10% on river (he does this now only 5% of the time):

in this line we would bluffraise specifically the 99 some more while generally we shove less

one more thing to note is that in both lines he is supposed to check QQ on river not block. And vs check in both lines we do jam all in with 99 a small percentage of the time (between 10-20% depending on suits again in both lines)

Hey Mathias

Thanks for ran this sim !

1. I didn’t expect a tiny blocks on the river. It’s surprising that he doesn’t do it with KK, QQ. Also, I don’t think people are 3b as much here , or maybe I’m wrong and people at 1k + are quit balance, but seems a bit too larger 3b range for me (T9s 3b near 0% for me). I don’t think we have access to Pool tendencies 100bb deep final table, unfortunately !
   Also, I think if we call 76s 100% , we are calling 98 87s 100% of the time , but It wouldn’t be as different as it is now I think

2. If we are raising, what is his calling range ? I think it’s the most interesting thing here, because I m convinced that pop would never respond to what solver recommend actually.
   My thought was that if he’s not calling the Jx , and Kk , QQ region (Which for me would happen most of the time even if I don’t play mtt) , we were an able to bluff a crazy amount of time .
   Also, nobody had <30bb. (Which reduce the icm pressure isn’t ?)

3. Yeah as You and Sam said it, JT J9 98s are better hand to bluff, according to Pio. 33% turn is curious , I think we tend to play passively in general in such ICm environment , especially on those stacks depth. Why 33% / not block and 75%/ block,
   IN CG in tend to be the opposite.

I am interested to have river range vs raise and block bet range if you still have the sim. (vs 33%)
Thanks for being open to discuss and for showing your thinking process.

Cheers.

I still have the sim but the problem is that both lines are seen so rarely, even when we look at him betting 33% turn which pio does, he then almost never blockbets river 10%, if i uncheck square size proportional to weight, its not a good representation of his actual range/the weights on river. A hand like A6dd for example he folds mostly but calls 30% of the time, but thats almost the only suited ax other than AQs that he gets to this point with, also they mostly take other lines. AQss calls now almost always, KK QQ fold the few hands that played like this overall I wouldnt bet again human overfolding if he gets to river with the hands i see here.
I think its more likely that a human will overly blockbet hands like QQ KK and other weak hands in that spot and that this is the reason why we may consider bluffing a bit more with a hand like 99. Again though its an ICM spot if we guess/exploit wrongly here we hurt ourselves so thats the other side of the coin.

how do you figure out how often a bluff needs to work? for this example 31%, how do you get this number, and then how many bluffs are allowed..

thanks

Ihaveyounow691

Hey!

As Sam said , I did a mistake here : it s 25 and not 31.

Here is what I Did :

We are putting 240 to win a 500 pot

Pot is now 240+500= 740

Vilain has to complete ~200 to win a 740 pot.
So hero needs to be good (in ChipEV)

: 200/ (740+240)~= 21%

————-
Imagine vilain has 1 to get into a pot of 4

If he won 20% of the time, he is at equilibrium.
If he play 5 times and is good one time :

1. He looses (-1)
2. He looses (-1)
3. He looses (-1)
4. He looses (-1)
5. He win (4)

In this case, bluffs allows for hero is

1/(4+1) = 20% In Order to put vilain s bluffcatcheur indifferent between folding and calling

—————

So at Sam shoes, against a perfect opponent, we are allowed to have ~21% of bluffs. (I said 31% because I inverse hero and vilain

Hope I answered your question, tell me if it’s not ! And sorry for tongue mistakes I’m not a native aha

For chips J8s is a close to bottom of range continue OTB here. In cash games you should continue because it's +CEV, but in a tournament you can make money folding because someone might get knocked out on this hand and they're more likely to get AI in a HU pot than a three way pot.

In this specific hand it's close enough that I could see either argument being made.
Eg.
1. Remi is opening tighter because it's a FT so I should call tighter preflop here.
2. Remi is gonna play more passive postflop because it's a FT so I should call looser preflop here.
3. The blinds cover both of us and will be squeeze happy so I should call tighter preflop here.

In the actual hand it's possible I made a mistake, it's a fringey spot and all three options should be close.

Generally speaking when stacks are grouped closer together at a FT there is less deviation from CEV strategy.