Ever fold KK preflop?

[benkath1]

Bookmark this thread. Keep playing. On December 4, 2006 open it up and re read it. Then go buy yourself something nice with all the money you've won in the past year with KK. [img]/images/graemlins/grin.gif[/img]

[bozlax]

[ QUOTE ]
No, folding KK preflop would be stupid. Even if you somehow "know" that your opponent has AA, your implied odds are huge if you flop a set.

[/ QUOTE ]

Well, if (I saw this happen live about 6 weeks ago in Vegas, standing on the rail at the Tuscany Garden's new poker room) your opponent gets so excited when it's raised and 3-bet to him preflop that he accidentally flashes the entire table his AA as he's pushing his raise in, your implieds just went in the sh1tt3r. It should be noted, however, that you should win the hand if you're the very next player to act after him (except that he's a big enough ass to flash his cards to the table, he's probably a big enough ass to not understand that if you're still playing after seeing his cards there's no way he can continue in the hand).

But "know"ing is much different than knowing. "Know"ing is just a euphamism for tilt, and is a bad thing.

[bozlax]

[ QUOTE ]
That's part of the reason I laid that hand down. Over that last 2700 hands, I've been dealt KK 14 times, raised it every time, but only won 28% of the time. That 28% would be even lower if I had called last night.

Now on the other side of the coin, Q's have been *very* kind to me. I win with QQ 80% of the time, and with AQ ~63% of the time.

[/ QUOTE ]

If you can't see how ridiculous the quoted thinking is, then there's really not going to be any way for us to help you. It is simply preposterous to think that QQ (and AQ, for the love of Shania!) are better hands to play than KK.

[ QUOTE ]
And based on the flop results, it certainly *was* the prudent thing to do.

[/ QUOTE ]

Let's make this simple: any decision that was the right decision based solely on anything at all that happened in the hand after the decision was made was the wrong decision.

[bozlax]

[ QUOTE ]
Do you really think you should conclude that the coin is weighted... only when you're holding it?

[/ QUOTE ]

QOTD

[ QUOTE ]
[ QUOTE ]

Quote:
Raising them PF has actually been -EV for me



This is not true.


[/ QUOTE ]

How is this not true? When a hand only wins only 28% of the time, raising should be -EV with. Actually, I guess with 4 players, you could argue that it is +EV. But only barely.

[ QUOTE ]


[/ QUOTE ]

EV is not relative to individual hand results (that's what makes it a useful concept to poker in the first place).

[ QUOTE ]
If you can't see how ridiculous the quoted thinking is, then there's really not going to be any way for us to help you. It is simply preposterous to think that QQ (and AQ, for the love of Shania!) are better hands to play than KK.

[/ QUOTE ]

No, that wasn't my point. My point was that the hands I've been getting are not falling where they theoretically should be. You're comment was exactly the point I was trying to make: because of variance/fate/luck whatever-you-wanna-call-it, my hand probability has run far outside what theory/models dictates. So my question was that since what I was seeing in reality was different than what theory dictates, do you compensate your play for that? I mean, everyone talks about this kind of stuff happening, so it's to be expected. So if you know you're in the middle of a downswing or whatever, it seemed like tightening up should soften that downswing, make it less severe so to speak.

As it turns out, that reasoning is flawed. You shouldn't change the way you play a hand simply because the deck is cold. I can accept that...

Now on a side note, there *is* a certain amount you can infer from the way things are running. If I know that the mean is 63% win rate for KK, and I'm currently at 28%, then eventually, I'm gonna go on a streak (all things being equal) where my KK will *not* get busted (just like I expect a "streak" where my QQ/AQ loose alot since they're running "hot" for me now). That *will* happen, it has to in order for win rates to reach the mean/expectation. Would it be wrong to alter your play based on that knowledge? Guess that question really belongs in the probability or variance forum, but it is something to ponder, no?


[ QUOTE ]
Let's make this simple: any decision that was the right decision based solely on anything at all that happened in the hand after the decision was made was the wrong decision.

[/ QUOTE ]

Try saying that 5 times with a mouth full of crackers [img]/images/graemlins/wink.gif[/img]

[ QUOTE ]
Do you really think you should conclude that the coin is weighted... only when you're holding it?

[/ QUOTE ]
I like that, very good analogy.

Now, I don't believe it's weighted for me only. Well, maybe a little bit [img]/images/graemlins/wink.gif[/img] <just kidding>. However, there is also a probability associated with how many times a head will come up in a row.

2 heads: 1/2 * 1/2 = 1/4.

And so on. So if I hit 5 heads in a row, well, the probability of that is very small. And the probability of 6 heads in a row is even smaller. Granted, each event is a unique event, but they can also be viewed jointly.

So if I was a gambling man, and I had 15 heads in a row, the probability of 16th head is still 1/2. But I know over the long hall, to even out, I'm gonna have to go on a streak of tails only to even this out. So at some point, you could shift your betting to "tails only", and come out ahead. Now granted, if you just bet on "heads only" or "tails only", you should be dead even at the end. But if you could pick up on when variance occurs and you hit a run of heads or tails, you could actually come out head. Hard part is: how do you know when you're in a run of heads or tails until after the fact?

[Aaron W.]

I know I'm a little late to be jumping in, but there seems to be something worth saying that hasn't been said:

[ QUOTE ]
I'm an engineer, and a data driven guy. So based on the data I have available, it seemed like a good laydown. It's kind of Bayesean thinking, but given that KK hasn't been paying off in the past, and all the action preflop, it seemed like the prudent thing to do. And based on the flop results, it certainly *was* the prudent thing to do.

I guess what it really boils down to is do you base your decisions on what the "theoretical" results should be or what "actual" results are. Theoretically, I should have played KK. Based on actual performance, I should have laid it down. I made my decision based on actual performance, and it turned out to be right. But over the long haul, this should hurt me.

[/ QUOTE ]

"Based on actual performance" and "based on the flop results" are meaningless statements and you acknowledge this in the last sentence.

You don't have the foreknowledge that an ace is coming on the flop. The evaluation of a hand is done with only the data available at the time of analysis, not future data. Before the flop falls, you have a large space of possible flops and hands that villains can be holding.

If you really want data, you can actually compute it. Compute the chances of an ace flopping (with no king) assuming that someone out there holds an ace. You'll be surprised to know that this happens rarely. I got a number on the order of 15% of the time. Worrying about the ace flopping is a minor concern. When you consider how much you stand to gain the 85% of the time an ace doesn't fall (especially for the 12% of the time you catch a set), your fear of an ace on the flop simply melts away.

Also as an engineer, I would expect that you are aware of statistical sampling. If you were making a product and wanted to do some spot checking, you wouldn't look at 14 of them and declare that you've got enough data to conclude that your operation involving millions of your product must be running smoothly/failing. You would want to test at least a few hundred, or even up to a thousand or so, depending on how tight of a control you need on the finished product.

[ QUOTE ]
Something else just occurred to me... You don't hit KK all that often. I think it was Ed Miller in one of his posts here that said something along the lines of "a mistake on hand that doesn't happen very often is a small leak". So if you make a decision based on texture of the game, and the way the cards have been running for you, is that really wrong? I mean, it's not like I lay down KK every hand. First time I've ever done so, in fact. And it turned out to be the right thing to do [img]/images/graemlins/smile.gif[/img]

[/ QUOTE ]

Yes, it's really wrong. Is it a leak? No. Folding KK when it's capped to you is not a leak since it rarely happens. But it's wrong.

My concern is that you're not looking at your "leak" properly. Here's the link that I think you've got in the back of your mind. It's not that you happened to fold this hand, or that you might have folded QQ-22 in this spot with gigantic implied odds staring you in the face.

The leak that you have is that you undervalue your powerful hands. This is a gigantic leak. You get powerful hands 3-5% of the time (depending on which hands you count), and they account for a substantial portion of your winnings. From my 10k at 1/2, AA-QQ and AK won $1038. Since my NET winnings over the same 10k was $445, you can see that failing to take advantage of these monster hands is a *HUGE* leak. When these hands are *THAT* important and they come so rarely, folding them preflop is simply out of the question.

[ QUOTE ]
That *will* happen, it has to in order for win rates to reach the mean/expectation. Would it be wrong to alter your play based on that knowledge? Guess that question really belongs in the probability or variance forum, but it is something to ponder, no?

[/ QUOTE ]

Not really. Each hand is an independent event. You can look back over 1000 or however many hands, and see that you were on a "hot" or "cold" streak, but there is no way to predict what is going to happen on any particular hand. So you have to just play optimally for each hand, which includes never folding KK preflop.

I think as you play more and more hands you will look back at this and laugh. 14 times is nothing. I've been playing limit holdem semi-seriously since about may and I have gotten KK 423 times, and like everyone else (including you after you play more hands), it is a big winner in the long run and I have never once folded it.

[Aaron W.]

[ QUOTE ]
So if I was a gambling man, and I had 15 heads in a row, the probability of 16th head is still 1/2. But I know over the long hall, to even out, I'm gonna have to go on a streak of tails only to even this out. So at some point, you could shift your betting to "tails only", and come out ahead. Now granted, if you just bet on "heads only" or "tails only", you should be dead even at the end. But if you could pick up on when variance occurs and you hit a run of heads or tails, you could actually come out head. Hard part is: how do you know when you're in a run of heads or tails until after the fact?

[/ QUOTE ]

This is completely false. You will not gain by shifting to a "tails only" strategy. You cannot predict variance. Regressing to the mean is not a matter of "opposite swings" balancing out.

If the first 15 times were all heads, how many do you expect there to be if you flop the coin another 30 times? You expect the next 30 to be half heads and half tails, so there would be 30 heads overall. So then you would see 30/45 heads = 2/3 of the time.

Starting again with 15 heads in a row, how many heads do you expect if you flop the coin 2,000,000 more times? You would expect 1 million heads and 1 million tails. So in all, you would see heads 1,000,015/2,000,015 = .50000375% of the time. THIS is how you return back to average. Over a very long sample, the numbers slowly slide back to average, not because of any predictable swings, but because the sample size gets large enough so that signal wins out over noise.