Poker hand I'll never forget
Siona
went to Ceasars. I have to say that I'm not at all impressed with where
the poker is now. What used to be a very nice poker room is now smashed
between the horse betting, the keno, slots, and some table games. I had
the tv's close behind me and people who wanted to watch the races kept
backing up into my chair. At one point I had to go to the ladies room
and there was this little old lady standing in my way, in the poker area,
watching the horse races. I said,"excuse me" and she moved about an inch
over. I said, "I can't fit through there" and she about exploded with
"You don't see me interupting your gambling. What an ass". I didn't bother
to point out to her that she was in the poker area and her yelling for her
horse was bothering us, but that's not the point of this post.
Sunday, Monday and Tues I went to Harrah's. The hand that I'll never
forget (although I don't remember all the details like so many of you
seem to do) happened on Tuesday night. I was playing 7 card stud at
a 1-5 table. My first 2 cards in the hole are queens and my up card
was a 3. I started with $1 and everyone stayed in. My next card is an A,
I'm high and I bet $2. Everyone stays in. Next card - another A. I bet
$3 and eveyone stayed in. Last up card is a 4 and the only other thing
showing at the table is a pair of 5's. Nothing else at the table even
looks like a possible anything. I bet $5 and only 2 others stay in,
one being the pair of 5's. My last card and I pull the A. Now I've got
Aces full of queens. I bet $5, one guy folds and the guy with the 5's
raises me $5. The only thing that I can figure is he has 5's full and
thinks I have a straight. So I raise back $5. He raises me again.
If my husband hadn't walked over just then it might have kept going
till one of us was all in, but since he (my husband) tends to balk at
things like that I just called the guy. Imagine my reaction, the rest
of the table's reaction and the dealer's reaction when he turned over
the other 2 5's.
"Murphy was an optimist!"
Lack of planning on your part does not constitute an emergency on mine.
http://www.netaxs.com/~siona/
Bob Maher
We were at our weekly home game (a bunch of professors) playing a game
of lowball. Dr. Wu raised the max after seeing his cards, and about half
of us called. This was a somewhat surprising event, because this guy was
usually pretty conservative.
He stood pat, while each of the other callers drew 1 or 2.
The betting was checked around to him, and as expected, he bet the max.
Everyone else folded.
After the last competing hand had dropped, he showed his- Four Kings.
Before anyone could ask him if he knew what game we were playing, he
said, "I've never had four of a kind before, I may never have it again,
and I wanted to make damn sure that I won with it."
Steve Sembay
This is one of my favourite hands. I played a rather passive part in it,
and the one decision that I did make in it was probably an error, but it
raises a smile whenever I think about it.
I was playing in a pot-limit holdem tournament; unlimited re-buys for 90
minutes and 55 starters. It's half an hour after the re-buy period is
over and it's obviously not going to be my night. I've been in a lot of
action but I've been sucked-out by the river card four times in a row
when each time I've had the best hand after the turn (c'est la guerre). I
only have T200 left and the average chip position is about T5500. The
current blinds are T50 (equal sized blinds in UK events) and they've just
announced that the blinds are going up to T100 next hand. On this hand
the under the gun (UTG) player has posted an additional T100 blind. I am
seated one to the right of the first blind.
5 players call the T100 and I look down to see the massive hand of 22
(both red). In retrospect I probably should have played to maximise my
probability of survival here by simply calling and hoping to see the flop
at no extra cost. I can then fold if I don't make a set and still play
one more hand. What I did do was maximise the amount of chips that I
would win by going all-in, knowing that few would fold because of my
tiny raise. With half the field still left though, It's probably a moot
point.
First blind drops, everyone else calls so there are 8 players seeing the
flop which is somewhat unusual to say the least. The flop is 9s 5s 3d.
Checked to UTG who bets half the pot. One middle position caller (MP).
Man to my immediate right (IR) raises the pot. Remaining blind folds. UTG
thinks for what seems a long time, then folds. MP calls. The turn is the
Qc. MP checks, and IR bets the pot again. MP also thinks a long time,
then folds. I stand up preparing to make my exit as we turn over our
cards before the river. To the consternation of the rest of the table IR,
the real hero of the tale (or villain depending on your point of view),
shows Ts 4s ! I'm winning !! He'd been semi-bluffing on a T-high flush
draw. The river is the beautiful Kh and I've won the T1600 side-pot with
an un-improved pair of deuces starting against 7 other players.
My good fortune continued because I went on to place equal first in
the tournament.
Steve
----------------------------------------------------------------------------
Steven Sembay | Department of Physics and Astronomy,
s...@star.le.ac.uk | University of Leicester, Leicester, UK
Darse Billings
>This is one of my favourite hands. I played a rather passive part in it,
>and the one decision that I did make in it was probably an error, but it
>raises a smile whenever I think about it.
This hand is an absolutely great example of the all-in advantage I wrote
about in a recent article. I think you played the hand very well.
First, I should mention that, apart from any other considerations, the
short-stack has a distinct advantage in percentage-payback tournaments.
In effect, each chip in a short stack is more valuable than a chip in a
taller stack. The peculiar nature of tournament mathematics which makes
this true is nicely illustrated in Mason Malmuth's book "Gambling Theory
and Other Topics". Beyond this fact is the practical and (probably)
theoretical edge enjoyed by the all-in player in any multi-player game.
>I was playing in a pot-limit holdem tournament; unlimited re-buys for 90
>minutes and 55 starters. It's half an hour after the re-buy period is
>over and it's obviously not going to be my night. I've been in a lot of
>action but I've been sucked-out by the river card four times in a row
>when each time I've had the best hand after the turn (c'est la guerre). I
>only have T200 left and the average chip position is about T5500. The
>current blinds are T50 (equal sized blinds in UK events) and they've just
>announced that the blinds are going up to T100 next hand. On this hand
>the under the gun (UTG) player has posted an additional T100 blind. I am
>seated one to the right of the first blind.
>5 players call the T100 and I look down to see the massive hand of 22
>(both red). In retrospect I probably should have played to maximise my
>probability of survival here by simply calling and hoping to see the flop
>at no extra cost. I can then fold if I don't make a set and still play
>one more hand. What I did do was maximise the amount of chips that I
>would win by going all-in, knowing that few would fold because of my
>tiny raise. With half the field still left though, It's probably a moot
>point.
Your last chips are precious, which suggests you should be unwilling to
put them in the pot unless absolutely necessary. The other tall stacks
are effectively getting an overlay when they match your paltry raise.
On the other hand, you are so far down that you need a big score to go
anywhere in the tourney. If you pick up a small pot, you'll still have to
go all-in a few more times to have a realistic chance of advancement.
Your hand of 22 is quite respectable under the circumstances, and as I
pointed out in the previous article, 22 gains a lot from being all-in,
because it gets to see all five cards in the hope of hitting the set.
Since there are so many callers, the pot is large enough that you should
feel committed to the hand. Even if you miss the flop, calling wouldn't
be a very large error. So raising all-in now when you are getting so much
action seems quite reasonable to me.
>First blind drops, everyone else calls so there are 8 players seeing the
>flop which is somewhat unusual to say the least.
Here is why the all-in player is getting such a big boost. Normally you'd
have to defeat all 7 opponents to survive. But because this is big-bet
(no-limit or pot-limit) poker, you expect that most of the players will
have to fold. Each person that folds is one less person you have to beat
at the showdown. Even if the best hand is still active, each folded
player donates a small amount of equity to your cause, because some of the
outcomes which would have won for them may now win for you. You don't
know how much you gain from any particular fold, but it will usually be
greater than zero. You are in a position to capitalize on this added
bonus, because you're all-in. This basic effect is at it's peak when
you're all-in in a family pot, and the game is big bet poker.
> The flop is 9s 5s 3d.
>Checked to UTG who bets half the pot. One middle position caller (MP).
>Man to my immediate right (IR) raises the pot. Remaining blind folds. UTG
>thinks for what seems a long time, then folds. MP calls. The turn is the
>Qc. MP checks, and IR bets the pot again. MP also thinks a long time,
>then folds. I stand up preparing to make my exit as we turn over our
>cards before the river. To the consternation of the rest of the table IR,
>the real hero of the tale (or villain depending on your point of view),
>shows Ts 4s ! I'm winning !! He'd been semi-bluffing on a T-high flush
>draw. The river is the beautiful Kh and I've won the T1600 side-pot with
>an un-improved pair of deuces starting against 7 other players.
This is a perfect result for you. We might criticize the play of the
bettor, bluffing at a protected pot, but the fact is that he did add a
lot of equity to his hand by turning the 4's and T's into an extra six
winning outs, not to mention winning some cash from the side pot. It
may have been sub-optimal, but it can't be all that terrible a play.
Besides, these players are just imperfect humans, so they just might
act in a way that greatly enhances our winning chances.
>My good fortune continued because I went on to place equal first in
>the tournament.
C'est la guerre? Vive le short-stack :-)
Of course, we don't really *want* to be in a short-stack position,
because we eventually need the biggest stack to win. But the lesson is
that we shouldn't despair, because there are some compensating factors
that work in our favour. Don't just casually toss those last few chips
in just to get it over with. Fight on!
Cheers, - Darse.
--
char*p="char*p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}
James -jazbo- Beauregard
Regarding playing pat in lowball draw with quad Kings:
Long, long ago I played in a tiny limit (literally nickel & dime)
lunchtime game. I was dealt AKKKK. I didn't even know what snowing
was at the time & I probably couldn't have bought the pot anyway, but
I couldn't resist turning it over.
--jazbo
--
James -jazbo - Beauregard (mail to ja...@nova.bellcore.com only, please)
Kevin Cline
Actually, this hand is an example of sound strategy for draw poker.
In order to get called when you bet, it is necessary to bluff sometimes.
How often should you bluff? It depends on the size of the pot
and the size of the bet. The higher the bet in relation to the pot,
the less often you should bluff, and the less often your opponents
should call. The lower the bet in relation to the pot, the more often
you should bluff, and the more often your opponents should call.
How do you decide when to bluff (before the draw)?
Well, it is pointless to bluff with hands with drawing potential;
they are more profitable as calling hands than betting hands.
So you bluff with bad hands.
When you have a bad hand, you still have to decide whether to
bluff or not. You should make this decision based on some
random criteria; otherwise there will be a detectable pattern to
your bluffing that your opponents may figure out. At the card table,
you have built-in random information: the contents of your hand.
So, based on the size of the pot and the size of your bet, you select
a set of hands whose frequency matches the correct bluffing frequency.
For example, you might to decide to bluff a pat hand in lowball
whenever you were dealt trips, or a pat full house, or five cards
ten or higher.
There is an old book that gives a nearly optimal strategy for
five-card draw based on this idea. Perhaps some reader of the thread
can give the title; I am in the process of moving and can't dig it
out right now.
--
Kevin Cline
James Michael Hammond
>
>There is an old book that gives a nearly optimal strategy for
>five-card draw based on this idea. Perhaps some reader of the thread
>can give the title; I am in the process of moving and can't dig it
>out right now.
You might be talking about "Poker Strategy: Winning With Game Theory"
by Nesmith Ankeny. I took Prof. Ankeny's game theory class while an
undergrad at MIT. His book assumes you're playing jacks or better to
open, pot limit five card draw, and he lays out an approximation of
optimal strategy in most of the simpler situations and gives you the
fundamental information necessary to build the optimal strategy in
more difficult situations.
I don't know whether the book is still in print unfortunately. I
didn't ever buy a copy but read a friend's copy through long ago.
--JMike
-------------------------------------------
J Michael Hammond <ham...@cs.utk.edu> |
http://www.cs.utk.edu/~hammond/jmike.html |
Jay Sipelstein
>In order to get called when you bet, it is necessary to bluff sometimes.
>How often should you bluff? It depends on the size of the pot
>and the size of the bet. The higher the bet in relation to the pot,
>the less often you should bluff, and the less often your opponents
>should call. The lower the bet in relation to the pot, the more often
>you should bluff, and the more often your opponents should call.
>How do you decide when to bluff (before the draw)?
>Well, it is pointless to bluff with hands with drawing potential;
>they are more profitable as calling hands than betting hands.
>So you bluff with bad hands.
It is possible to be more precise than this. You can work out the proper
game theoretic frequecy to bluff, which depends on the factors mentioned
above. Suppose you should bluff a fraction p of your hands. Then, the
hands you should bluff are precisely the BEST fraction p of the hands which
you would otherwise fold.
Why? Simply because you might get called by someone who can only
beat a bluffing hand. In that case, you might as well have the best
possible hand that you could show down.
>So, based on the size of the pot and the size of your bet, you select
>a set of hands whose frequency matches the correct bluffing frequency.
Right, and I maintain that the set of hands which to bluff should be
fixed, and not picked at random.
>For example, you might to decide to bluff a pat hand in lowball
>whenever you were dealt trips, or a pat full house, or five cards
>ten or higher.
I think these are a bad choice of bluffing hands.
One other factor to consider as well (I'm not sure how to take this into
account in the above, anyone else?) is the cards your hand removes from
the deck. From this consideration, I'd FAR prefer to make a bluff with
2222 than with KKKK. Might there be someway to compute the set of hands
to have that minimize the liklihood of yuor opponents having a playable
hand, and then bluff the proper percentage of these, instead of what
I stated above? Would this improve your win rate?
>There is an old book that gives a nearly optimal strategy for
>five-card draw based on this idea. Perhaps some reader of the thread
>can give the title; I am in the process of moving and can't dig it
>out right now.
I forget the title, but the author is Ankny (sp?) a Prof at MIT. The book
gives a game theoretic treatment of Jacks or better, high only draw
poker. I read it about 8 years ago, but would probably get alot more
out of it now. Does anyone know the exact title, and whether its still
in print?
-- Jay Sipelstein
Michael Maurer
Don't know if it's still in print. From our library catalog:
AUTHOR: Ankeny, Nesmith C.
TITLE: Poker strategy : winning with game theory / Nesmith C. Ankeny.
IMPRINT: New York : Basic Books, c1981.
xv, 189 p. ; 22 cm.
TOPICS: Poker.
Game theory.
NOTES: Includes index.
Language: English Year: 1981
ISBN 0-465-05839-6 : $11.95
-Michael
--
______________________________________________________________________
Michael Maurer Center for Radar Astronomy (415) 723-1024
mau...@nova.stanford.edu http://www-star.stanford.edu/~maurer/
Darse Billings
>In article <3u630e$5...@sun132.spd.dsccc.com>, kcl...@sun132.spd.dsccc.com (Kevin Cline) writes:
>>In order to get called when you bet, it is necessary to bluff sometimes.
>>How often should you bluff? It depends on the size of the pot
>>and the size of the bet. The higher the bet in relation to the pot,
>>the less often you should bluff, and the less often your opponents
>>should call. The lower the bet in relation to the pot, the more often
>>you should bluff, and the more often your opponents should call.
In theory, the ratio of bluffs to legitimate bets should be *larger* for
small pots, and *smaller* for large pots. When betting the size of the
pot, the optimal bluff ratio is 1:2 -- one bluff for every two legit
bets. The optimal call frequency of a pot sized bet is 1/2. When there
are N bets in the pot, the optimal bluff ratio is 1:N+1, with a call
frequency of 1/(N+1).
>>How do you decide when to bluff (before the draw)?
>>Well, it is pointless to bluff with hands with drawing potential;
>>they are more profitable as calling hands than betting hands.
>>So you bluff with bad hands.
>It is possible to be more precise than this. You can work out the proper
>game theoretic frequency to bluff, which depends on the factors mentioned
>above. Suppose you should bluff a fraction p of your hands. Then, the
>hands you should bluff are precisely the BEST fraction p of the hands which
>you would otherwise fold.
>Why? Simply because you might get called by someone who can only
>beat a bluffing hand. In that case, you might as well have the best
>possible hand that you could show down.
Wrong. With no more cards to be dealt, you should bluff with your weakest
hands, because they have the least chance of winning a showdown. With
stronger hands you check, hoping to win the showdown, or pick off a bluff.
When there are still cards to be dealt, however, you should semi-bluff
with the strongest fraction of hands not otherwise bettable, because of
the free-card danger. The strength of your hand is a combination of the
probability of having the best hand and the potential for improvement in
the case where you do not.
In holdem, you might semi-bluff AK overcards on the flop or turn, but
would check on the river. With 32 (nut low) you discard on the flop or
turn, but pure-bluff on the river if given two free rounds.
But _optimal_ (game theoretic) bluffing is *much* inferior to _maximal_
bluffing, aimed at exploiting your opponent's weaknesses. In practice,
you should bluff when you think it will be profitable (chance of success
greater than the pot odds). Often this means bluffing at large pots,
because weak opponents are often oblivious to the size of the pot.
You should not bluff a calling station, but should bet more often for
value. [One of my favourite questions is: "How could you bet that shit?",
to which I *never* reply "Why did you call and lose?" :-]
Game theoretic bluffing is reasonable against unknown opponents, or
against much stronger players who may be able to read you well -- but
then an even better play may be to quit the game...
>>So, based on the size of the pot and the size of your bet, you select
>>a set of hands whose frequency matches the correct bluffing frequency.
>Right, and I maintain that the set of hands which to bluff should be
>fixed, and not picked at random.
The set of hands can be fixed, but will *occur* at random.
>>For example, you might to decide to bluff a pat hand in lowball
>>whenever you were dealt trips, or a pat full house, or five cards
>>ten or higher.
>I think these are a bad choice of bluffing hands.
>One other factor to consider as well (I'm not sure how to take this into
>account in the above, anyone else?) is the cards your hand removes from
>the deck. From this consideration, I'd FAR prefer to make a bluff with
>2222 than with KKKK. Might there be some way to compute the set of hands
>to have that minimize the liklihood of your opponents having a playable
>hand, and then bluff the proper percentage of these, instead of what
>I stated above? Would this improve your win rate?
Yes, this is a well known principle in lowball. As noted, you should
occasionally run the "snowing bluff", in order to get more action from
your legitimate pat hands. I would bluff KKKK just on principle -- I
hate to ignore a sexy hand. :-) But it *isn't* the best choice for
snowing. Top priority should go to bad hands with five low cards
(eg. 22337, 444A6, etc), then those with four low cards, because these
reduce the chance of your opponent drawing to a hand worth calling.
But if the time is right, it just doesn't matter what your hand is...
>>There is an old book that gives a nearly optimal strategy for
>>five-card draw based on this idea. Perhaps some reader of the thread
>>can give the title; I am in the process of moving and can't dig it
>>out right now.
>I forget the title, but the author is Ankny (sp?) a Prof at MIT. The book
>gives a game theoretic treatment of Jacks or better, high only draw
>poker. I read it about 8 years ago, but would probably get alot more
>out of it now. Does anyone know the exact title, and whether its still
>in print?
You are thinking of the "Winning Poker with Game Theory" book by Nesmith
Ankeny (sp?) (full reference given elsewhere), dated around 1975, I think.
Better, in most respects, is "Winning Poker Systems" by Norman Zadeh.
It was written a few years before (1972, I think), as a follow-up to
his PhD thesis, and covers a broader range of variations and concepts.
Zadeh himself is a game theorist of the highest order -- his writings
on backgammon and other games are highly recommended.
These books are both hard to purchase, but the Zadeh book can often be
found in your local library.
Cheers, - Darse.
--
BlddLDfddFdbfRuuruBuubUF
Darse Billings
>In theory, the ratio of bluffs to legitimate bets should be *larger* for
>small pots, and *smaller* for large pots. When betting the size of the
>pot, the optimal bluff ratio is 1:2 -- one bluff for every two legit
>bets. The optimal call frequency of a pot sized bet is 1/2. When there
>are N bets in the pot, the optimal bluff ratio is 1:N+1, with a call
>frequency of 1/(N+1).
^^^^^^^
Oops, typo. This should read N/(N+1) -- you should nearly always call when
the pot is large.
jac...@xmission.com
>In article <3u630e$5...@sun132.spd.dsccc.com>, kcl...@sun132.spd.dsccc.com (Kevin Cline) writes:
>>In order to get called when you bet, it is necessary to bluff sometimes.
>>How often should you bluff? It depends on the size of the pot
>>and the size of the bet. The higher the bet in relation to the pot,
>>the less often you should bluff, and the less often your opponents
>>should call. The lower the bet in relation to the pot, the more often
>>you should bluff, and the more often your opponents should call.
>
>>Well, it is pointless to bluff with hands with drawing potential;
>>they are more profitable as calling hands than betting hands.
>>So you bluff with bad hands.
>
> It is possible to be more precise than this. You can work out the proper
> above. Suppose you should bluff a fraction p of your hands. Then, the
> hands you should bluff are precisely the BEST fraction p of the hands which
> you would otherwise fold.
>
> Why? Simply because you might get called by someone who can only
> beat a bluffing hand. In that case, you might as well have the best
> possible hand that you could show down.
>
>>a set of hands whose frequency matches the correct bluffing frequency.
>
> Right, and I maintain that the set of hands which to bluff should be
> fixed, and not picked at random.
>
>>whenever you were dealt trips, or a pat full house, or five cards
>>ten or higher.
>
> I think these are a bad choice of bluffing hands.
Hmm. I'm really shocked to hear you say something like this, since
I *saw* you bluff with 2-7 offsuit in the BARGE hold'em tournament
last year. You're just damned lucky that you paired the board to beat
my A-Q. I'll be gunning for you this year.... ;-)
--
Steve Jacobs (jac...@xmission.com) \ Warped by OS/2
"Expectation isn't everything..." \
Jay Sipelstein
First: I'm even more shocked to hear myself say that, since as Darse Billings
points out, I'm flat wrong. (Thanks Darse, I've been playing in too many
local games where you're not allowed to check. Really.)
Second: I had position on you that hand, and besides, you sucked out on me
with AQ vs my A9 and a 789 flop to knock me out later.
I'll be gunning for YOU this year,
-- Jay Sipelstein
Johnathan Jeffrey Rouse
Darse Billings <da...@cs.ualberta.ca> wrote:
>sip...@VEGAS.SCANDAL.CS.CMU.EDU (Jay Sipelstein) writes:
> >In article <3u630e$5...@sun132.spd.dsccc.com>, kcl...@sun132.spd.dsccc.com
[discussion of game theoretic bluffing and maximal (intuitive) bluffing]
>But _optimal_ (game theoretic) bluffing is *much* inferior to _maximal_
>bluffing, aimed at exploiting your opponent's weaknesses. In practice,
>you should bluff when you think it will be profitable (chance of success
>greater than the pot odds). Often this means bluffing at large pots,
>because weak opponents are often oblivious to the size of the pot.
>
>You should not bluff a calling station, but should bet more often for
>value. [One of my favourite questions is: "How could you bet that shit?",
>to which I *never* reply "Why did you call and lose?" :-]
I just have to relate a hand I saw played this week at a $5-$10
Hold'em game.
I was standing behind the rail scouting for juicy tables when I
saw a guy look at his freshly delt 54off. There is a preflop
raise from early position, and a bunch of callers including our
hero. Flop comes Ax 9x 6y. Preflop raiser pops it, two callers
and of course our hero. Turn is a 7z. Same guy goes for it, one
caller and then hero has a draw. (He calls). River comes a Jack
with no flushes makin' it. Same guy continues to run it through
and everyone folds to our hero who calls and raiser shows AJs. He
mucks (disgustedly)....
"Ahem! Floorman, could you put me on the list please? Oh no
list? You are gonna seat me where? Oh the smoke is really bad
at that table could I get a transfer to #8? Thanks!"
> >I forget the title, but the author is Ankny (sp?) a Prof at MIT. The book
> >gives a game theoretic treatment of Jacks or better, high only draw
> >poker. I read it about 8 years ago, but would probably get alot more
> >out of it now. Does anyone know the exact title, and whether its still
> >in print?
>
>You are thinking of the "Winning Poker with Game Theory" book by Nesmith
>Ankeny (sp?) (full reference given elsewhere), dated around 1975, I think.
I managed to pick up a copy of this book at a used book store for
a few bucks. What a steal! (Great play!)
-Jeff Rouse
jjr...@acs.ucalgary.ca