PLO50 2nd Nut Full

For me making a second 3bet in 160 hands is enough to make assumptions about the player. I mean for average TAG player there should have been about 15 3bets so far. It doesn't mean that he is 3betting tighter for sure. But lets say there is 50% chance that unknown player is reg/tag. How much would this 3bet stat change that number? If someone says it doesn't affect, that would be wrong. If someone says the number drops to 0%, that would be be wrong. But lets say that the stat and sample size weights at about 10% (probably bigger, but lets keep it simple). Now the chance of him being non-reg/tag is 40%.

Once we do the same for his preflop stats, they drop the number again. Lets say it drops to 30%. And then he min3bets, might be a misclick or not. Lets say it weights again 10%, so the overall probability drops to 20%. And then he uses those weird sizings on the flop and turn, lets keep it simple and say he might be tricky reg and the weight is another 10%.

Overall none of those individual things are THAT reliable information, but overall they sum up to estimation that he is not a good reg/tag to 90%. That is very reliable estimation.

Of course above is just a simplification, and we shouldn't start to numerize the probabilities like that. But we should gather all the information, process it and make estimations and assumptions. If those prove to be wrong, we calibrate and change them. It's all about probabilities and the weight of information. Like when we face a bet on the river, we don't know if it's a bluff or not. We estimate the probability based on all the information we have, and then we adjust our default strategy based on that. If we make good estimations in the long run, we make profit. If we make bad estimations in the long run, we lose money.

Its not marginal to call almost anything preflop versus this sizing from position and definitely not hand like high pocket pair that we can purely set mine and show profit. I have never said villain range is just AAxx. But it will get closer to it with decent sample and practically no 3bets at all. If we have 10 hands on someone and his stats are 60/60/40 can we not make any assumptions? Sure of course its possible guy has just hot deck but more likely scenario is that he has too wide ranges. Or lets say you join table and someone 3bets three hands in row without showdowns. You open and face 3bet from same guy. Could be four solid hands in row but we should really ask which is more likely, 4 solid hands in row or villain 3betting way too wide. This is example on the other extreme side of statistics on very small sample.

Not related to this hand but solver does lot of light calls in lot of spots because it assumes opponent to have very creative bluffs that most human players simply dont have at least in high enough frequencies. Good example could be x-r line in 3bet pot on J55r flop. 99% of players are way too value heavy in this spot when they x-r. Some might turn some Jx into bluff. But solver would x-r-f 45% of time in this example and x-r versus 1/3 pot cbet whopping 24% of time:

So hands like A322 with 3 high bdfd, A432 with two bdfd. Even more logical ones like 9876ds are very often just being x-f. I recently watched video on plomastermind about these paired flops that was on pool in PLO500. There data on pool showed that people are fast playing too much strong hands and not finding even closely enough bluffs making raise range way too value heavy. On those stakes pool in general is more creative and aggressive.

Originally posted by Mascherano95
I don´t think we have implieds odds. Villain is not going to bluff a paired river with air but will check-call his top-set when the river brings a flush? Makes no sense to me.

There is simple reason why this can be the case. That is not understanding relative hand strength and/or falling in love with the hand. There are absurd amount of examples of people not getting away from AA in holdem even if board runs very bad. Why? Because its aces, best starting hand available! When player get this idea of "now I am going to win big pot" it can be very hard to let hand go.

Why do we see people calling set or even two pair in omaha without any relevant blockers when board offer flush and straight? Likely because they dont like folding or they just think set is still strong hand. They might not think about any blockers etc. at all in first place. Why does some people play aggro with small flush and isolate themselves mainly versus better? Most likely again because they think flush is strong hand. E.g. in holdem two card flush on non paired board is very strong hand and lot of players come from holdem background to test omaha. So its not that rare to see bad calls.

Good example of overvaluing AA:

Originally posted by Mascherano95
Most fish 3-bet pot too. Strictly spoken the 3-bet sizing does not give us information about the players´ skill level.

How often regs use min3bet?
Pot size 3bet = no information
Min3bet = probable information

Not related to this hand but solver does lot of light calls in lot of spots because it assumes opponent to have very creative bluffs that most human players simply dont have at least in high enough frequencies.

I know that human have less bluffs in certain big pot spots and there are deviations. But if you think everyone check-folds a fullhouse otr to such a tiny bet, that´d might be mirror-thinking.

I am not going to read that J55 example since I do make effort to differentiate that and an Ace-high flop. I have a 225 bb/100 over 100k+ in 3bp´s ip. Since my GTO studies of 3bp recently that graph is skyrocketing over 4k hands so potentially I am doing sth right here when 160 hands is supposed to be a big sample. I am not going to stop keep working on my solver-based strategy and not going to replace it for a nut-peddling fit-fold strategy that indeed worked well on PLO10.

I am not telling you what to do. Just keep folding those spots like many other regs, just another eez triple barrel for me then. If it is a 40/10 player indeed like over a bigger sample, he is likely not 3-betting much more than 2%. The non-AAxx part might be JT98-7654ds. And you can check out how such a range composition is weighted in combos. Accordingly as I already said when you are sure villain is having AAxx we should fold the flop because of pot odds. Or you show me a relevant hand where a player with same profile pays you off on 3 betting streets with top set vs flush. Otherwise I have to persist on your assumptions being purely speculative.

Few points:

160 hands is not a big sample and no one has said so. But that doesn't mean it's meaningless sample. As Ghaleon stated, even few hands can give useful information. But again, that doesn't mean it WILL give useful information.

Neither Ghaleon or I have ever said that we would fold in the original hand. The discussion has been about meaningful sample sizes and weight of information.

Originally posted by Kyyberi
Few points:

160 hands is not a big sample and no one has said so. But that doesn't mean it's meaningless sample. As Ghaleon stated, even few hands can give useful information. But again, that doesn't mean it WILL give useful information.

Neither Ghaleon or I have ever said that we would fold in the original hand. The discussion has been about meaningful sample sizes and weight of information.

Agreed. "big sample" was ironically. 50-200 hands are uselful already for Vpip and PFR to some dregree for 3-bet as well. It´s still not a call flop call turn fold river not going to discuss this further, we can keep discussing this when we get the same scenario again in another hand.

Second Nut Full on the river with 117 BB at the Button... O-MA-HEART, OH-MY-SOUL

I wouldn't fold river. To make hero folds I would need to have solid reads on opponent. If he has AA, then he has and we make notes, especially on flop and turn sizings. And I don't think that has been the issue here for a long time. Discussion has moved to information reliability, which is SUPER important topic.

Even 200 hands is not bullet proof info on preflop stats. I once found myself sitting on a table with 8/4/0 stats after 100 hands. And as Ghaleon said, sometimes just few hands gives useful information.

Correct me if I am wrong, but I think what caused a bit confusion here is that Mascherano95 thought we are discussing about the original hand, while for me this is the first time I even comment that original hand. Could have been more clear to move this information topic to another thread.

probably leveled myself there^^
just seemed to make some sense at that moment with the stats and the sneaky line
but its also true that these guys can show upwith random stuff and especially 3way think its not good to fold that
funny that this hand lead so such a big discussion and especially @ghaleon nice posts as always:f_thumbsup:

Not going to search hand example without tracker support about specific bad calls but I am pretty surprised if people have not three barreled straight board or flushing runouts to see being called by very bad calling hand. In this case we also talk about bd flush hit possibility. I personally found myself underestimating risk of bd flush when I started omaha with long holdem background.

Originally posted by ghaleon
Not going to search hand example without tracker support about specific bad calls but I am pretty surprised if people have not three barreled straight board or flushing runouts to see being called by very bad calling hand. In this case we also talk about bd flush hit possibility. I personally found myself underestimating risk of bd flush when I started omaha with long holdem background.

Risk of Backdoorflushes kinda dependant on the preflop ranges in play like a tight range is more likely to hit that because you have more double-suited hands.