What's Wrong With This Statement?

[M.B.E.]

I can't understand brigt's answer. I think Pzhon got it right. If you check, you will still win the pot sometimes (specifically, when your opponents all check behind you [if they are behind you], nobody has a full house or quads, you hit the trey or deuce, and nobody else hits the full house or quads).

[Cerril]

First, Ed's post does give some reasons that show checking has some EV, and while that and another post put me on the path that was how I came to see it.

Second, obviously we're talking about where the breakeven point lies, I -thought-. Since the initial post says 'if the chance is greater than 20%, bet; otherwise fold' 20% is just taken to represent the zero point.

It's an interesting question and an intricacy that obviously I'm not getting yet. It's definitely been shown that you can reduce the value very slightly and still have it be +EV; it's also been shown that you can increase the value slightly and still have it not be the best decision.

It's possible that the correct answer revolves around making an assumption many of us are somewhat loathe to make given the highly detailed nature of the example (which implies 'everything you need to know is here'). If that's true then I guess I am stumped.

The only message I've seen that didn't seem to be aiming in roughly the same direction was the one you replied to in the first place, which to be honest I'm not sure I understand - the language seemed a little vague or I just missed the point (more likely).

[Brineboitano]

I think i might have it (sorry if i'm duplicating someone, i didn't read all the responses)

Just because a play has positive EV doesn't mean it's the best play. If you check, it's possible your opponent will check down the best hand. Then either you lose and cost yourself nothing in the process, or you might draw to a split. In other words: if you check, there is NOT a 100% chance you lose the entire pot, so your EV is positive, but if you bet with exactly a 20% chance of winning on a bluff, your EV is zero.

But now I'm second guessing myself, cause that doesn't account for the possibility that they bluff you out if you check, but I'm going to post this anyway, screw it.

[ilya]

Let's say you figure that there's exactly a 20% chance of success on the flop. What this means, really, is that you think there is exactly a 80% chance that someone has 44 or higher. Now moving in is 0 EV.
If you check, however, and it is checked behind, the turn may bring a J (as Gabe said), counterfeiting the pair, or alternately any card higher than your opponent's pair. So, by checking, you give your opponent's hand a chance to become weaker, on average. Though the chance that your opponent has a pair will have increased to approximately 82.5% (he will make a pair about 12.7% of the 20% of the time that he is unpaired), the chance that your opponent calls your all-in will have dropped way below 80%, as he is less likely to call with an overcard to his pair on the board.
You then gain EV by checking even if your chance of success is somewhat greater than 20%, since your opponent's ability to call on the turn is likely to be *drastically* reduced, and since he will sometimes indeed check behind on the flop fearing a trap.
I am too tired to try to work out the numbers, but I think the idea makes sense to me -- it's a sort of reverse slowplay.

[top6]

I agree, but for a different reason. Mr. Sklansky said the correct answer was in a post that nobody had talked about. Near as I can tell, Gabe's is the only possibly correct post that nobody has talked about.

This is the kind of high-level thinking we expect from law school graduates.

[ilya]

[ QUOTE ]
I agree, but for a different reason. Mr. Sklansky said the correct answer was in a post that nobody had talked about. Near as I can tell, Gabe's is the only possibly correct post that nobody has talked about.

This is the kind of high-level thinking we expect from law school graduates.

[/ QUOTE ]

Hehe...I initially followed this line of thinking myself...perhaps I should go to law school after all... [img]/images/graemlins/grin.gif[/img]

[rgreenm90]

I'm officially changing the topic from "interesting principle" of EV to the merits of monotheistic religions, mainly to get Sklansky back here to answer this friggin question.

[David Sklansky]

Because you may hit a two or three. Now tell me why that means you need more than a 20% chance to steal when you can't win when you are called.

[rgreenm90]

Because you can win in a showdown if you hit the two or three, whereas there is simply no way you could if you don't.

Does Gus Hansen's style of poker make you cry, Mr. Sklansky?

[ilya]

Oh yeah, that makes sense if your opponent bets no matter what. You then call. Your chance of hitting a 2 or 3 on the turn or river is somewhere around 24%. Subtract from that the small chance that you make your full, but your opponent also makes his, and you still end up with over 20%. Thus, unless your chance of stealing successfully exceeds that over-20% percentage, you should check and call.

I still think my other reason is the right one if there's some chance your opponent will check, though I suppose it depends on what that chance is, exactly.