Continuation of my bad Stud/8 thread, question

[Bartholow]

This came up in my hand post, but was so buried that it didn't get any responses.

In the hand, I had split kings and reraised an unknown player who completed first in with a 9 showing. It was asserted that I should have just folded, as anyone "reasonable" with a 9 showing should have (99)9, (AA)9 or (KK)9 to play. So what I'm wondering is how often an unknown player will be this "reasonable", specifically in the Party 10/20 but also in general?

In other words, when someone raises with a 9 up, doesn't Bayesian analysis tell us he's more likely to be a moron than have AA, KK or 99 down?

[DeadMoneyOC]

Well I can assure you the answer is not found through whatca-wanna-call-it analysis. It can certainly be determined given your read on any certain player though. If you think that this player doesnt need AA,KK,or999 to raise then re-raise. If you think he does need one of those three hands, then you should fold. Go with your first instinct and trust your read.

[Bartholow]

I had no read currently whatsoever, as stated. And I'm quite sure that we can use Bayes' theorem here.

[DeadMoneyOC]

OK, then you are going to need to explain this theory to me...

[Bartholow]

Here's a resource in technical language about Bayes Theorem. In reality the math is pretty trivial, and in most poker contexts you can just think of it as common sense.

In the context of poker, we use Bayes Theorem mostly to compute the relative likelihood of certain events, given some information. For instance, in the hand loosely in question, a 9 completed, there was another 9 dead, and I had split kings. So given those cards were out, IF we assume the raiser with the 9 had either kings, aces or rolled nines, we use Bayesian analysis
to determine the relative likelihood of those hands. Since there are only 2 nines unaccounted for, and two kings unaccounted for, there is only 1 way of having each of those hands. Meanwhile there are 6 ways of having aces buried. So there's a 6/8 chance he has aces, a 1/8 chance of kings and 1/8 chance of rolled. (These are relative probabilities given the cards we see, rather than the absolute probabilites of something like 1/(420*13) chance of being dealt rolled nines in general, for instance.)

So what I want to say, is given the cards out and the fact that the nine raised, but NOT assuming the player is good, what are the relative likelihoods of a) the player being any good b) the player holding various hands.

So as we just learned, a good player could raise 8 hands there. Let's say a bad player would raise any pair, but not other hands, for the sake of this example. Given the cards that are out that means something like about 70 hands. Now if P is the probability that an unknown player is good in general, and 1-P is the probability that an unknown player is bad, we can figure out from those probabilities PLUS the fact that the player raised with a 9 up what the conditional probability that this player is good is, like this:

8*P/(8*P + 70*(1-P))

I think I did that right.

So what I'm really asking for is the overall population ratio of "good" to "bad" players, approximately, and also what hands I can expect a "bad" player to raise with, on average. This is clearly going to be pretty unscientific, part of my point is simply that I shouldn't guess this player is reasonable right off the bat. But I'd also like to get some subjective guesses about all this from players who play in that game more often than I do.

[Bartholow]

Shameless bump! See my 2nd reply to DeadMoneyOC for more in depth discussion.

[DeadMoneyOC]

[ QUOTE ]
So what I'm really asking for is the overall population ratio of "good" to "bad" players, approximately, and also what hands I can expect a "bad" player to raise with, on average. This is clearly going to be pretty unscientific, part of my point is simply that I shouldn't guess this player is reasonable right off the bat. But I'd also like to get some subjective guesses about all this from players who play in that game more often than I do.

[/ QUOTE ]

I have only playing in that game exactly one time so I definitly dont have a good sense of what type of opening standards one could have with a 9 door card. I agree that a good player usually should only be raising in that spot with KK,AA or 999, but I also think there are probably a lot of bad players raising with QQ and JJ. So I guess I would probably re-raise given this exact situation. 999 seems unlikely and AA isnt that huge of a favorite. There are some other variables that may affect your desicion(in a generalized sense) that come into play but I am sure you are aware of them. If you are getting value in the long run against a random opponent, it is probably very thin. I would also work on getting read on the player in question !! [img]/images/graemlins/tongue.gif[/img]

I would also like to thank you for reminding me of the nightmare that was my Business Statistics class. [img]/images/graemlins/grin.gif[/img]

[randomstumbl]

[ QUOTE ]
In other words, when someone raises with a 9 up, doesn't Bayesian analysis tell us he's more likely to be a moron than have AA, KK or 99 down?

[/ QUOTE ]

I think your question is as follows... Against good playerss I'm a pretty big dog. Against bad players, I'm a favorite. Are there so many bad players that I have enough equity against a random player drawn from that population in this situation.

That's not an easy thing to answer without collecting a serious amount of statistics.

Luckily, it's very easy to just get a basic read on the 7 other players at the table. So, no one actually needs to answer this question.

[Bartholow]

Let's say it's your very first hand at a table. It is still your (or at leat my) job to maximize your expectation on that hand.

Obviously I don't want that "serious amount of statistics", I just want to talk about gut feelings related to this, with the understanding that seeing someone raise with a 9 should already give you some probablistic information about whether they are good or not.

[randomstumbl]

[ QUOTE ]
Let's say it's your very first hand at a table. It is still your (or at leat my) job to maximize your expectation on that hand.


[/ QUOTE ]

Generally, giving people respect until they show you they don't deserve it is going to be the right move. If they're bad, you're going to have plenty of shots at their money. If they're good, you're much better off folding.

I really doubt it's going to be optimal to commit a large amount of money to a marginal hand against an unknown opponent who is showing a lot of strength. You can probably invent a population of players that make it optimal, but even the 3/6 stud/8 population is going to be better than that.