The heart did factor into my thought process on the flop.

He'd probably raise an A. I think a b/3! would be my default play, and it if gets 4! so be it.

In this specific post, the villain is a good to very good 15/30 player in the pool and a decent(probably winner in 40/although i'm not sure)overall. We also have a decent amount of history and he's definitely paying attention to things like ranges and who's in the pot with him.

Personally I don't want to k/c here because if I surrender initiative he can take the 6's to the river for 1.5 bets here and auto call pretty much any river.

There are hands I have that i'd be okay with not putting pressure on him, this specific one isn't one of those. like 7's or 8's come to mind maybe a non gutter ball/heart Jack.

this seems a bit like reductio ad absurdum. i never said ALWAYS k/r and b/c turn when he raises. i acknowledged most of the time we should be k/c, k/c ing here.

and you also forgot the other side of it when HE's semibluffing the turn and gives up on the river we win 3bb and not 1.5 or 2.5.

you also forgot to mention that he might fold pairs <J if anything like a 9, T, Q, A hit on the turn or the 3rd heart comes when we bet out after k/ring.

you picked the two most diametrically opposed scenarios for the lines and compared them solely on those, which i don't think is fair.

EDIT: also, i don't think the line can be "really really bad", period.

Getting raised on the turn here is not an extreme example, it's a very common and likely occurrence. I simply pointed out tou neglected to mention it while focussing so much on what ZOMG said, trying to fold out 6 outer in tiny pot.

And it's not really really bad in and of itself, like CR here prob doesn't lose to much overall; but when given the option of taking a vastly superior line, taking this one is definitely really really bad. And taking a line that will involve putting in 3 big bets and not showing down on a draw heavy board vs a tough opponent is just really really bad.

The bot I'm looking at

When defending BB vs BTN
across all J-hi, Q-hi, K-hi flops

xrs bottom pair 28% of the time
xrs mid pair 45% of the time
xrs top pair 90% of the time


When it xrs bottom pair, OTT
cbets 63%
donk checks 38%

When it xrs mid pair, OTT
cbets 67%
donk checks 33%

When it xrs top pair, OTT
cbets 84%
donk checks 16%

Across all hands that xr flop, it donk checks 25%.


Obviously not filtered for other flop texture considerations, but still useful. Based off this, x/r bottom pair isn't bad. Would also pair with a substantial donk check turn strategy.

so...... just because thats what some random bot does means it has to be a good strategy?

how is that any more useful than me just pulling me up some random player in a database and saying well this is what this guys does so it has to be good?

If you're going to keep posting your garbage no one will take you serious unless you can prove you're at least talking about a winning bot

and you know all bots have holes in them right? prove this is a spot the bot is playing correctly. You dont even say if the bot is winning in any of these lines. Just saying a bot does X so X must be good is beyond worthless

As BTN vs BB defend

Across Jhi Qhi Khi flops, after being xred OTF, OTT
BTN raises a cbet 18%
BTN folds to a cbet 18%
(small sample size)

Across all flops, After being xred OTF, OTT
BTN raises a cbet 21%
BTN folds to a cbet 17%


So, as BB, we do have to worry about being raised on the turn. But we also gain a decent chunk of FE on the turn.

I trust this bot. No, bots are not perfect, but I trust in it more than I would trust the so called pros. Pros have holes, too. Seeing how a bot plays may helps pros improve their holes, or at least open them up to considering alternate lines.

If you think its garbage, feel free to ignore. I think others may find it useful though.

its garbage because you offer no evidence besides "I trust this bot" that it plays well. Yes pro's have holes. Thats why no one is just looking up a players stats and just saying "trust me, this guy does it so it must be right"

You post numbers from a random bot. Have no evidence that anything it is doing is right and expect it to somehow be informative.

Please back up any of your claims. It's really dumb to say bot is doing X so X is correct when you arent even showing that every time the bot does X it isnt losing 10 BB's

I dont know why people keep harping on this. Why do you want FE? What hands exactly are you so hell bent on making him fold? If he folds the turn it means your c/r made him fold a 6 outer in a 4 bet pot, the kind of hands you'd actually like him to keep bluffing with, while at the same time value towning yourself or getting raised a lot.

Well, I'm sold.

Pot size being the main thing I overlooked here. k/c, k/c, decide seems like the best line in isolation.

This just means that I have to adjust my buckets a bit to accommodate more passive lines against tough opposition.

Maybe you are right, and I am incorrectly blindly putting my faith in a bot. I just love bots. I believe bots to be generally stronger than humans, even weaker bots. Its a bot that has solved HU LHE. It will be a bot that solves 3-handed LHE. So, I'm going to try to learn from the bots. What I can tell you is that this bot beat every other bot it's been put up against for the last 5 years.

Even if this bot actually is a limit holdem wizard, no human is going to be able to duplicate his strategy anyways, making the comparisons more pointless.

Bots will never solve live holdem, becuss humans will be incapable of duplicating their strategy/decision making

So if you copy the bots flip strategy and it's actually correct, your going to screw up the turn/River/and other actions on flop with other holding enough to make whatever line the bot takes significantly less profitable than it is for him.

It doesnt even matter if you're correct if you're going to present the information in such a meaningless way. In order for the data to have any merit you have to back up thats it's at least a winning strategy for the bot.

You claim that because the bot is doing play that way it must be right. At least show that the bot is showing a greater profit using this line than an other. I've played a ton against bots. The easiest way to beat them is to find their "holes." Prove the bot plays this spot better than any one else and that you're not just blindly following in a spot where the bot doesnt play well.

Folding out a 6 outer is worth ~+1/7th of the pot. If we are drawing to a 5-outer, value towning yourself is ~-.88BB. So, you look correct, FE doesn't matter that much. We still get to charge hands that would free card the turn, but it does look like taking a more passive line is better.

I am intrigued by a x/r, donk check line though, only barreling if we improve or a heart comes.

Pros:
-Might get immediate FE on the flop (the good kind of FE)
-Equity-wise, we're a small favorite on the flop
-Allows us to continue on favorable cards
-Don't get value owned
Cons:
-Villain gives up on bluffs
-Awkward line that might be hard for us to play (or balance)

If he folds to the flop c/r you didnt want him to fold. If you're concerned that a flop call will let him free card, why then would you want to c/r, ch and put him in a spot where he can now play pretty close to perfectly and bet for value or still freecard you? Why are you worried about being freecarded in a 3 bet pot but not a 4? Villain will make very few mistakes unless you're now going to try and find a c/r, c/r range.

PP,

i too enjoy studying bot play. however, i find that it's mostly irrelevant when it comes to play vs. humans at stakes up to 40/80. here's why.

the bot you're looking at is playing vs. other bots and it's ultimate goal is not to be exploited. we're playing against humans that are very exploitable and our ultimate goal is to win the most money we can.

from the bb in a spot like this, one element of the bots strategy is to make the btn indifferent to betting more than a certain percentage of the time. vs. humans we're looking to exploit the fact that humans cbet at a frequency that is too high. with this hand, we do that by letting him bluff too often on the flop, turn, and river.

from the bb after k/r'ing the flop, the bots have a turn donk checking strategy that balances their range vs. range equity. hero likely will blindly bet all turns. in fact, he chose this line because he wants to b/3b when an A comes. if i'm the btn and hero b/3bets the turn with this exact hand, he's going to be in bad shape a lot of the time. the times he's ahead, i'll have 13 outs to improve. i think wanting to b/3b the turn is just compounding the initial error of k/r'ing the flop.

bots are going to be betting the flop/turn/river at the correct frequencies. winning humans often fail at this miserably. we exploit that by bluff catching with this hand.

these are just some of the reasons why i think bot play is somewhat irrelevant to most situations in live poker right now.

as an aside, do you know what the 2nd board card is on the K, Q, Jhi boards the bot is k/r'ing bottom pair? also, what is the bots kicker when it k/r'es bottom pair?

Agreed

@rodeo
I saw various board cards, various kickers. Some low, some high.

Agreee that exploitative strategies are more useful. In my gut, Ah4x feels too strong to relegate as a bluffcatcher. It can improve and handle heat. Compare with the static 55-TT, which I'd rather play passively.

???????????

Look I actually sympathize with the general idea of learning from bots. But if we take this statement at face value, I'm sorry it looks like you completely misunderstand what a bot is. A bot that has "solved" HULHE (of which I know of only one in existence) by definition cannot also be the same as one that solves 3-handed LHE. We're talking about computer programs here, not cyborgs sent from the future to kill sarah connor.

just want to chime in saying I appreciate phunkphish's posts ITT and do not understand why he was getting heat from a mod for asking good questions.