Fit or Fold -- ouch

[jogsxyz]

[ QUOTE ]
jogsxyz,

Care to elaborate? I'm not quite sure what you mean, but it sounds potentially interesting.

~Mike

[/ QUOTE ]

In MOP the clairvoyant game is introduced.
Two players.
Each player starts with 185 units.
Each player antes 5 units.
Ignore the suits.
Only player B is dealt a card.
If it's an ace or king he wins.
2/13 chance of player B winning.
That's about 15%.
The game has 3 streets.
On each street player B may bet or give up.
When player B gives up, player A wins the pot.
If player B bets, player A may call or fold.

With both players playing best strategy,
player B is the favorite in this game.
This is the advantage of being the
clairvoyant. Player B after seeing his card
knows who's the winner. Player A doesn't
know. With nearly 85% winners player A is
the dog after 3 streets of betting.

This knowledge wasn't known when Mason wrote
his book.

jogs

[jogsxyz]

In fixed limit with only two streets of
betting being the clairvoyant is probably
worth 7 to 13%.

[uDevil]

[ QUOTE ]
You need to read the short handed section in HPFAP.

Best wishes,
Mason

[/ QUOTE ]

Hi Mason,

Not sure if this was addressed specifically to me, but it had been a while, so I re-read it. It was worthwhile. Some differences I noticed are that in the article you say (emphasis added):

[ QUOTE ]
Suppose you are the big blind in a short-handed game, it is raised by an aggressive player, and you call. (Notice that this is analogous to everyone passing to an aggressive player in a late position in a full game.)

[/ QUOTE ]

However, in the book, when discussing calling frequency and tallying up the frequency with which we flop a pair, straight draw, or overcards, you say

[ QUOTE ]
That brings us to 50 percent, which is not enough against a very aggressive player who is automatically going to try to steal.

[/ QUOTE ]

When you suggest changing the A[img]/images/graemlins/club.gif[/img] to 2[img]/images/graemlins/club.gif[/img] you say

[ QUOTE ]
If you do this you should be at approximately the right strategy for playing against super aggressive players who constantly take the pot odds on a steal or a semi-steal.

[/ QUOTE ]

These differences are significant. To me, "aggressive" means someone who will attempt a steal ~40% of the time. Super and very aggressive players who steal at every opportunity are a different class of player. In that case, calling is certainly reasonable.

[Mike Pemulis]

uDevil,

Actually, I'm not sure you're right here. How do you actually plan to profit by calling with QT?

The central issue here is that a play cannot be called correct only because it prevents your opponent from showing an automatic profit. Automatic profits happen all over the place in poker.

Calling this flop peel with QT "standard" is simply incorrect. It's worth pointing out, by the way, that short-handed experts have been criticizing the shorthanded section of HPFAP for years. The fact that it relies so heavily on preventing automatic profits might have something to do with that.

It's also interesting to point out that in Small Stakes Hold'Em, Ed Miller writes (and this must have gotten the Malmuthian thumbs-up) that any argument that does not mention the expectation of a play is deeply suspect. Note that "automatic profit prevention," without consideration of the actual profitability of the call, is an example of such an argument.

~Mike

[uDevil]

[ QUOTE ]

uDevil,

Actually, I'm not sure you're right here. How do you actually plan to profit by calling with QT?

[/ QUOTE ]

I'm no expert, so I'm not sure either. However, we're now talking about an opponent who has a very wide range of hands and will almost certainly bet the turn after we call on the flop. If we make a pair on the turn and call when he (almost certainly) bets, he'll also often bet the river. So assuming a pair will be good, we're getting maybe 9:1 to call the flop bet when we have more than 6 outs.

Of course, sometimes he'll have an ace (or better) and we're way behind. But this will be partly offset by the times we make our backdoor straight and we get paid off for extra bets.

We need to put some better numbers to this to see how wide a range is required before the call is profitable, but against a player who could be playing any two cards, it looks to me that calling should be profitable.

Note: I'm assuming this opponent will never fold a better hand than ours, so bluffing and semibluffing are not options. I think thats a good assumption against most players who fit the super aggressive description.

[Mike Pemulis]

uDevil,

OK, that's an explanation. But:

*I think you vastly underestimate the probability that an aggressive opponent has an ace. Remember, some of the extra hands that aggressive opponents add include aces.

*Remember that a pair will not always hold up, even when it makes top pair on the turn.

*Again, good shorthanded players have positively recoiled at the thought of making this play, or at least of making it standard.

Think of this another way. You sit down at a poker table and start talking to the guy two to your left, who is eccentric in that lovable, thought-experiment sort of way. He offers to sign a contract that, for the rest of his life, when you open the button and he calls out of the blind and checks it to you on a Axx flop and you bet and he has QT, he will either (a) always call or (b) always fold. What would you prefer? I'll happily choose (a).

~Mike

[uDevil]

[ QUOTE ]

*I think you vastly underestimate the probability that an aggressive opponent has an ace. Remember, some of the extra hands that aggressive opponents add include aces.


[/ QUOTE ]

Maybe so. Hands with aces in them are 15% of all hands. How many hands will a "super aggressive" opponent raise? 80%? If so, aces make up ~19% of his range. But when a ace flops, the probability he has an ace is decreased to ~16% (need to check this). In this case we're in big trouble.

[ QUOTE ]

*Remember that a pair will not always hold up, even when it makes top pair on the turn.


[/ QUOTE ]

Sure.

[ QUOTE ]
*Again, good shorthanded players have positively recoiled at the thought of making this play, or at least of making it standard.


[/ QUOTE ]

I think that's mostly because they are thinking of an aggressive player as someone who plays like they do. So they assume villain is stealing ~40%. In that case, he's twice as likely to have an ace in his hand as a "super aggressive" player and the flop is terrible.

[ QUOTE ]
Think of this another way. You sit down at a poker table and start talking to the guy two to your left, who is eccentric in that lovable, thought-experiment sort of way. He offers to sign a contract that, for the rest of his life, when you open the button and he calls out of the blind and checks it to you on a Axx flop and you bet and he has QT, he will either (a) always call or (b) always fold. What would you prefer? I'll happily choose (a).

[/ QUOTE ]

If I'm raising 40%, I choose (a). If I'm raising every hand, I choose (b).

[Mike Pemulis]

uDevil,

That's all very good thinking, and I'm pretty sure you're right, or at least approximately so. But I don't think that the hypothetical opponent in Mason's article should be assumed to steal with hands like 92o (and even 100 percent of the time with something like Q4).

And the point, really, is that you decide to call or not by figuring out your expectation, not by trying to prevent your opponent from having an automatic bet. (Again, consider a standard high-limit hold'em game with experts in it. They make automatic bets all the time, and not just on the flop. Anyone who believes that there's something inherently wrong with that is, I'm sure, more than welcome to sit in these games and teach those guys a lesson. But I repeat myself.)

~Mike

[uDevil]

[ QUOTE ]

That's all very good thinking, and I'm pretty sure you're right, or at least approximately so. But I don't think that the hypothetical opponent in Mason's article should be assumed to steal with hands like 92o (and even 100 percent of the time with something like Q4).

[/ QUOTE ]

The argument is rough. But it does seem that there is a threshold of steal percentage above which calling is good and below which it is bad. I would like to know what Mason's estimate of this threshold percentage is.

[ QUOTE ]
And the point, really, is that you decide to call or not by figuring out your expectation, not by trying to prevent your opponent from having an automatic bet. (Again, consider a standard high-limit hold'em game with experts in it. They make automatic bets all the time, and not just on the flop. Anyone who believes that there's something inherently wrong with that is, I'm sure, more than welcome to sit in these games and teach those guys a lesson. But I repeat myself.)

[/ QUOTE ]

I agree that the idea of preventing an automatic profit is an unconvincing argument. However, there's probably a way to define "automatic profit" to make it consistent with EV calculations. I assume the argument is put this way for the sake of readers who faint at the sight of numbers.

[Mike Pemulis]

[ QUOTE ]
I agree that the idea of preventing an automatic profit is an unconvincing argument. However, there's probably a way to define "automatic profit" to make it consistent with EV calculations. I assume the argument is put this way for the sake of readers who faint at the sight of numbers.

[/ QUOTE ]

uDevil,

There's a problem with that sort of assumption. The only way that "prevents an automatic profit" and "has positive expectation" mean the same thing, once we fix the standard meaning of "has positive expectation," is to perversely redefine the words in "prevents an automatic profit" to mean things other than "prevent," "automatic," and "profit." It's like getting on a bus to Times Square, being dropped off at Central Park, and being told that for bus-signage purposes "Times Square" is a name of Central Park.

Just about any reader familiar with poker would know that Mason meant that the goal is to prevent a bettor from getting folds more often than the break-even point for profitability (as determined by the pot odds). And any clear thinker familiar with poker, who had time to consider the situation and the arguments, would realize that this argument holds no proverbial water.

~Mike