a DERB hand

[elindauer]

I find it interesting that many people think DERB will DEFINITELY call a raise, while at least one person believes he is probably going to fold to a raise.

This is exactly the kind of confusion a mixed strategy would cause in a field of players who believe there must be one right answer. To take this further, based on the number of players who think he'll call and the size of the pot, I'll guess that DERB's mixed strategy here about coincides with the size of the pot, ie, he's handling a raise with the triple:

{10, 90, 0}

-Eric

[Paluka]

[ QUOTE ]
I find it interesting that many people think DERB will DEFINITELY call a raise, while at least one person believes he is probably going to fold to a raise.

This is exactly the kind of confusion a mixed strategy would cause in a field of players who believe there must be one right answer. To take this further, based on the number of players who think he'll call and the size of the pot, I'll guess that DERB's mixed strategy here about coincides with the size of the pot, ie, he's handling a raise with the triple:

{10, 90, 0}

-Eric


[/ QUOTE ]

No way man, I bet it is 12,87,1.

[Nigel]

[ QUOTE ]
Who thinks DERB calls the turn even if it misses him completely?

[/ QUOTE ]

He calls turn and river, even if both miss him. He does not fold A high here. He also would never fold to a raise on the river. He should know that his opponents know this and therefore could safely fold 100% with his current image, but he doesn't.

His river play is not mixed strategy. He doesn't even care what his opponents have, he only cares what he has - see his flop check on this hand for further evidence of that point.

[mike l.]

"Perhaps villain usually calls with ace high and worse, but bluff raises often enough for a call to be good. Is this so hard to imagine?"

that's not at all hard to imagine but it's part of what makes a bet-call here awful. if derb bet-calls here it means he's playing by the seat of his pants: he doesnt have a plan for each street, he just reacts when presented with the consequences of his most recent decision. that's how donks play. because of the one card straight on board and the Q on the flop and the fact it went to 3 bets preflop it's correct for derb to bet if he gets called enough by A high or worse pair, OR it's correct for him to check and let cutoff bet if he's the type to bluff a lot. but it's not both ways. there are other boards and other scenarios where it can be both but this is not one of them. if derb had something like TT then this might turn into a bet-call scenario depending on the opponent. but for derb to bet-call this board means he is playing on the fly and while his instincts may be good at times due to experience, he's basically a donk.

[mike l.]

"A good default line is to bet / fold"

i disagree. i think this is generally a bad play online where the player are more erratic and aggressive.

but i understand and agree with your mixed strategy to an extent. it's basically an extension of sklansky's concept of "game theory". except really that generally applies to total unknown opponents. and then we're back to defaults. which we all agree are wrong.

the game is situational. there are no defaults. there's always a best play and it's oftentimes detectable, and, more often than would seem to be possible, it's counterintuitive as well.

the thing is a mixed strategy works best on the flop and turn and when making a choice between calling or raising. at least that will instill the greatest confusion in your opponents, if that's your thing. if you make a mixed strategy fold the only one who knows about it is you. folds still need to be dictated by pot and implied odds and knowing your opponent.

what if we think about folding (and check-folding) always being the default play in any scenario and street and anything other than that being the entire mixed strategy matrix of possible check-calls, bet-folds, check-raises, call-reraises, etc, etc.

[NLfool]

I haven't read all the responses but I hate how the hijack played this. 22 in the hijack sucks in a heads ups or battle for the blinds. I mean you're better off with 89 as you're going to make a continuation bet regardless and you'll get the fold if he misses with 22 or air. You'll hit better draws pairs and etc. Plus 22 sucks against a player like DERB

If you get 3 bet with 22 well I would hope with position you'd cap it and take control for only a SB which is a bargain.

to me 22 is the worst hand in this situation against this player. Gives you just enough to stick around but really it's not enough

[mike l.]

agreed. the 22 is a ridiculous hand bound to be played ridiculously. which it was.

[elindauer]

[ QUOTE ]
[ QUOTE ]
I find it interesting that many people think DERB will DEFINITELY call a raise, while at least one person believes he is probably going to fold to a raise.

This is exactly the kind of confusion a mixed strategy would cause in a field of players who believe there must be one right answer. To take this further, based on the number of players who think he'll call and the size of the pot, I'll guess that DERB's mixed strategy here about coincides with the size of the pot, ie, he's handling a raise with the triple:

{10, 90, 0}

-Eric


[/ QUOTE ]

No way man, I bet it is 12,87,1.

[/ QUOTE ]

You seem to be making fun of me for trying to give numbers here. The real comedy is that so few people are willing to see that it can be something other than definitely fold, or definitely call.

Of course I don't know exactly. I don't even know for sure he's using a mixed strategy. I'm just trying to give an estimate of what he seems to be doing. I thought that would be obvious, but apparently not.

-Eric

[elindauer]

[ QUOTE ]
He calls turn and river, even if both miss him. He does not fold A high here. He also would never fold to a raise on the river...

His river play is not mixed strategy. He doesn't even care what his opponents have, he only cares what he has - see his flop check on this hand for further evidence of that point.

[/ QUOTE ]

Please give some evidence for how the hell you could possibly know any of this. You state this like you know it for a fact. I find this virtually impossible to believe.

I don't know that he's using a mixed strategy. I'm simply noting the following facts which suggest he may be:

- mixed strategies are good and profitable
- it would be hard to recognize someone using one
- there is lots of confusion about how DERB plays, including dissension about whether or not he would fold to a raise here. this confusion is consistent with a mixed strategy
- DERB seems to be winning consistently and we're having a hard time explaining why any other way


Somehow you just know this is not what's happening. You choose wording like he does not, he can't, and he never... Please explain how you've hit on this insight.

-Eric

PS. In my short sessions with DERB, I found that he folded the river in many spots that surprised me. My short experience is not consistent with your "always / never" analysis above.

[elindauer]

At the risk of sounding like I'm arguing just to argue, let me point out a few minor disagreements with what you wrote:

[ QUOTE ]
but i understand and agree with your mixed strategy to an extent. it's basically an extension of sklansky's concept of "game theory".

[/ QUOTE ]

It is game theory. I'm not so arrogant as to believe that I have invented anything new here. Just trying to give it a name and make it more easily used as an analysis tool.

If you research existing poker bots, you'll find that the best ones use this triple concept already. I'm just trying to point it out to the forum.

[ QUOTE ]
... except really that generally applies to total unknown opponents...

[/ QUOTE ]

Not at all. It applies very well to opponents who are good enough to adjust to what we are doing, which means it applies to lots of players that play mid and high limit poker. If we always call in a spot, they stop bluffing. But then we'd want to fold. But always folding causes them to bluff a lot... so we need a mixed strategy to hold them off.

[ QUOTE ]
there's always a best play and it's oftentimes detectable, and, more often than would seem to be possible, it's counterintuitive as well.

[/ QUOTE ]

I basically agree that at any given moment, there is one best play and that it is often times detectable. What you have to keep in mind though is that against good players, our plays change the way they play us. So, given a hand in a vaccuum, we have to consider that folding may cause them to raise more, making calling right next time, which makes folding right the time after that which... we can simplify this by writing {50, 50, 0}.

[ QUOTE ]
the thing is a mixed strategy works best on the flop and turn and when making a choice between calling or raising. at least that will instill the greatest confusion in your opponents, if that's your thing.

[/ QUOTE ]

I agree that mixed strategies have applications in many spots on the flop and turn when deciding to bluff / semi-bluff. I don't agree that it's not just as useful on the river when deciding both when to value bet / raise / bluff and when to call such bets.

[ QUOTE ]
if you make a mixed strategy fold the only one who knows about it is you.

[/ QUOTE ]

So? That can be a great thing. Say I call down with ace high when check-raised on the turn, but I only do it 20% of the time. Well, those showdowns provide a lot of deterrence against my opponent semi-bluffing me on the turn, right? If your theory is correct, he'll now assume that I always call down with ace high, and will bluff far too seldom. Since I'm folding 80% of the time, I'm taking quite a bit of advantage of this tendency, and would like to keep it that way. If he knew what I was really doing, he'd be better off bluffing more. But he probably won't. This may well be exactly how DERB uses his loosey goosey image to fold lots of other times with mediocre or even strong hands without detection.

[ QUOTE ]
folds still need to be dictated by pot and implied odds and knowing your opponent.

[/ QUOTE ]

Yes, of course you are right here. Pot odds should have a major bearing on the fold percentage in the triple. Just because the pot is big doesn't mean you should never fold though, right? Maybe you only fold 1% of the time in the big pots.

[ QUOTE ]
what if we think about folding (and check-folding) always being the default play in any scenario and street and anything other than that being the entire mixed strategy matrix of possible check-calls, bet-folds, check-raises, call-reraises, etc, etc.

[/ QUOTE ]

This seems to be just a rewording of what I've been saying all along, but perhaps I don't understand what you mean exactly.

-Eric