Razz - Working out the percentages - seventh street

[Mathew82]

[ QUOTE ]
I feel the same way, by the way... I always want to know how to do something, even if, in practice, I won't be doing it myself. Sometimes it really does help.

[/ QUOTE ]

Your right and there is alot of percentages on sceranio's that the simulator just can't provide... all it does is give showdown stats.

[RainierBob]

Often you can make reasonable inferences about hole cards. For example if someone is betting agressively against a threatening board, it follows that he has good cards in the hole. I thought of this in another posting where the calculation assumed all the "unknown" cards were random where it seemed to me that villain was almost certainly good in the hole and consequently his hole cards could be deducted from Hero's possible outs. Unless, of course, those good cards duplicated ones in Hero's hand.

[Praxising]

[ QUOTE ]
All I want to do is to find out how to work these things out. It might not be that useful other than by solving alot of these puzzles that you have a greater intuition about the percentages. I plan to do alot of this type of maths.


[/ QUOTE ]
I really was asking a question and not being critical, sorry if it sounded that way. I do similar stuff when I translate ancient Greek to modern English for Scripture study. I'm no scholar and I am not getting it perfect but it really helps me to get a flavor for the meaning. So doing this yourself has got to get ideas about how it al works into your head that are useful at the table. You have my rspewct, believe me.

But it was a real question. I want to know how you use this kind of info? I mean, AA in holdem wins 69% of the time against random hands. But we know hands aren't random,so the information is good for determining relative hand strength not absolute win expectation.

I seriously want to know how this helps you think about the game. Maybe it just helps hone instinct.

[Praxising]

[ QUOTE ]
When there is a card you are looking for, you never know if it's in the deck or in the cards that have already been dealt. All you know is that it is one of the cards that is "unknown" - this means already dealt or in the deck. But the thing is, there is basically no difference between a card being already dealt, and it being at the bottom of the deck, where it can never be dealt.

[/ QUOTE ]
I think there is, though. In all mathmatics that deals with the real world, we have to start with presumptions, right? Because we cannot know anything absolute about reality as it is in constant flux. (Except for macro situations, like somehow actually knowing where every cards is.)

If we treat the cards we have not seen as all being possibly in the set of undealt cards, we skew the results rather badly. Like when you make the probability calculation on the chance of getting a specific suit to make a flush draw by saying "we've seen four spades, there are 9 spades left in the deck" and so forth and come out with some 33.4% probability in two draws to hit the card.

The only reasonable presumption is random distribution which is what all of our poker hopes are always pinned on, anyway. You always have a 25% chance of drawing any suit. Of course, in reality, if you knew all the cards, this would vary, but your only reasonable presumption is of the cards dealt, 25% were of each suit and so, 25% of the cards left are each specific suit. The turn and the river have the exact same flush draw probability: 25%. This is, of course, a problem debated amongst mathematicians who deal with poker probabilities and different answers can be found in different books.

In Razz , the difference between the card being dealt and the cards being at the bottom of the deck is that you always have to make the calculations with the whole set of cards not in your hand or visible on the table. This number varies so much so often as the hand progresses, that it makes more sense to use the random probability numbers for each rank.

Of course, I even I can do 13 of 52 is 25% for the flush calculation. I am totally lost with the rest of this.

I can't do math myself, but I do know that all of the reliability of it is dependent on the initial presumptions being as reflective of the real situation as possible. And the usefulness of it is in how it relates to realworld situations.

I'm not challenging you by any means, not smart enough. I just want some numbers that are useful to me. Like:

How much difference does it really make to my A23 hand if one, two, or three of my cards are out? Of the same or different suits? How does the number of paint cards showing affect the probability that my 257 is going to make a five card 7 by the river? If I have a three card 7 and every other upcard on the board is 8 or less, what are the chances that any other one of the 7 players has a three card 8 or better?

In a sim, when my A23 should win 69% of the time, how many of those times is it winning with a better pair, K, Q or J? Based only on probability, I should call every three-card wheel to sixth st at least. But I fold A23KK. Should I?

[RustyBrooks]

When you do simulation you are not assuming that your opponent has random cards, you assign a range to him, like you do when you're playing, and simulate based on that range. How good the results are depend on how close your assumed range is to his actual range.