CHINESE POKER...? BIG MONEY TOP PLAYERS?
SMe
"The Chinese poker games were very large, and I witnessed some players take
huge swings. I watched Doyle Brunson go from over $100,000 loser, to about
$220,000 winner in five or six hours. Ted Forrest, Chao Giang, Doyle
Brunson, Jonny Chan, Annie Duke, Howard Lederer, and I were playing for
$1,500 a point and up, so you can understand how Doyle could take a $320,000
swing. "
What is this game that attracts the top players? Is it high stakes pure
"gambling" or there is a way to beat this game on the long run?
Can some one care to describe it in general terms?
Thanks
Sme
Terrence Chan
news:wEOM4.3959$JN3....@newsfeed.slurp.net...
> I read this in P. Hellmuth page:
>
> "The Chinese poker games were very large, and I witnessed some players
take
> huge swings. I watched Doyle Brunson go from over $100,000 loser, to about
> $220,000 winner in five or six hours. Ted Forrest, Chao Giang, Doyle
> Brunson, Jonny Chan, Annie Duke, Howard Lederer, and I were playing for
> $1,500 a point and up, so you can understand how Doyle could take a
$320,000
> swing. "
>
> What is this game that attracts the top players? Is it high stakes pure
> "gambling" or there is a way to beat this game on the long run?
The swings are a direct result of playing with people of Chinese descent,
who like to straddle, raise, three-bet and where applicable, cap it with
apparently random cards. Okay, just kidding.
> Can some one care to describe it in general terms?
Briefly, the game is played 4-handed, and the entire deck is dealt to the
four players. From here you make two five-card poker hands and one three
card hand with your thirteen cards. The "bottom" five-card hand must
outrank the top five-card hand, and both must outrank the three card hand.
The four players then show down, and you match up hands head-to-head...each
matchup (top vs. top, middle vs. middle, bottom vs. bottom) you win gets you
one point. To make this example easy because I'm feeling lazy, let's just
show your hand and mine:
My hand:
QQQ33, AK752 flush, 995
Your hand:
44422, QJT98, KK7
I win the first hand because my queens full beats your fours full, and I
also win the second hand because my ace-high flush beats your queen-high
straight. That's two points for me. But your kings beat my nines, so you
get a point there. Overall, I net one point. If we were playing for $1,500
a point like in Hellmuth's game, that's what you'd pay me for this round.
But all is not lost for you, however, because there's another player in the
game with the following hand:
KK55J, 99662, JJA
What an awful hand. You beat this player's hand on every sub-hand, and he
loses three points ($4500) to you. (Note however that he's actually an idiot
because if he looked he could make jacks full of something to beat you on
one hand.)
*At this point I'd like to note that I cannot actually verify that
everything that has been stated in my example is 100% accurate because I've
never played Chinese poker in a casino before, only in for-fun home games
when my parents taught me and I quickly got bored of it many many years ago.
As far as I know, this is accurate.
So you can see why there are such massive swings in CP, namely because
you're forced to play every hand. There's no folding; you're dealt 13 cards
and you just have to make do with what you have and hope you don't lose on
every sub-hand. Often when you're dealt a total junk hand in CP you try to
win one of the sub-hands so you only lose 2 out of 3 points to your
opponents.
--
Terrence Chan
http://www.sfu.ca/~tchand
"It profiteth the wise, to be deemed a fool."
-Oceanus, Aeschylus' _Prometheus Bound_
Daniel Negreanu
DANIEL:
Actually that's not completely correct. You win 1 point overall plus
one for winning the hand for a total of 2($3000).
> But all is not lost for you, however, because there's another player
in the
> game with the following hand:
>
> KK55J, 99662, JJA
>
> What an awful hand. You beat this player's hand on every sub-hand,
and he
> loses three points ($4500) to you. (Note however that he's actually
an idiot
> because if he looked he could make jacks full of something to beat
you on
> one hand.)
>
DANIEL:
This would actually be an awesome hand if played correctly, JJJ55,
99662, and KKA. That's a monster.
> *At this point I'd like to note that I cannot actually verify that
> everything that has been stated in my example is 100% accurate
because I've
> never played Chinese poker in a casino before, only in for-fun home
games
> when my parents taught me and I quickly got bored of it many many
years ago.
> As far as I know, this is accurate.
>
> So you can see why there are such massive swings in CP, namely because
> you're forced to play every hand. There's no folding; you're dealt
13 cards
> and you just have to make do with what you have and hope you don't
lose on
> every sub-hand. Often when you're dealt a total junk hand in CP you
try to
> win one of the sub-hands so you only lose 2 out of 3 points to your
> opponents.
> --
DANIEL:
That's not always true either, depending on the rules you've decided
on. In most game they offer a surrender option. Which means if you
have a bad hand you can pay each player 2 bets($3000) and avoid the
risk of getting scooped($6000) or facing royalties. Royalties being a
straight flush(5 points), four of a kind(4), two full houses(2), and
trips on top(3). You can also play with naturals, 6 pairs, 3
straights, or 3 flushes.
Playing with surrender adds a lot more skill, because than you have
to evaluate the strength of your hand, and your not a victim to
horrible cards.
The reason it's so popular(I think), was because sometimes the big
games aren't really worth playing if you know what I mean. Yet,
sometimes people just don't want to go home yet, or are expecting, a
how shall we say, "nice guy" to show up.
This game became so popular at the Bellagio, it literally destroyed
the poker room for over a month, closer to two. I would walk in there
and people were stuck playing chinese and didn't want to start a poker
game. Then, management stepped in and did the right thing, forcing
people to play chinese along with some other form of poker in a mixed
game. Thanks to the WSOP though, there will be no shortage of action
this month. Daniel Negreanu
Sent via Deja.com http://www.deja.com/
Before you buy.
Jonathan Kaplan
"SMe" <vl...@netusa1.net> wrote:
>
> ...or is there a way to beat this game on the long run?...
>
i first learned about this game years ago when a friend of mine (D.
Smolen) was writing the book. in discussions with him, it seemed to me
that this game was too easy to catch on.
guess i was wrong.
i think this game is so simple that a computer program could be written
to play optimally. i dont like playing games where a computer can play
better than a human, right on it's face. this seems (to me) to be that
kind of game.
no program can beat a great no limit holdem player. a computer program
can be written that can tie the best "chinese poker" player in the
world, right now. i think.
stupid game.
smile
Jonathan
--
no matter where you go, there you are...
Jeff Roberts
Terrence Chan wrote:
> SMe <vl...@netusa1.net> wrote in message
> news:wEOM4.3959$JN3....@newsfeed.slurp.net...
> > I read this in P. Hellmuth page:
> > [...]
> >
> > What is this game that attracts the top players? Is it high stakes pure
> > "gambling" or there is a way to beat this game on the long run?
>
> To make this example easy because I'm feeling lazy, let's just
> show your hand and mine:
>
> My hand:
> QQQ33, AK752 flush, 995
>
> Your hand:
> 44422, QJT98, KK7
>
> I win the first hand because my queens full beats your fours full, and I
> also win the second hand because my ace-high flush beats your queen-high
> straight. That's two points for me. But your kings beat my nines, so you
> get a point there. Overall, I net one point. If we were playing for $1,500
> a point like in Hellmuth's game, that's what you'd pay me for this round.
> But all is not lost for you, however, because there's another player in the
> game with the following hand:
>
> KK55J, 99662, JJA
>
> What an awful hand.
Not only that, but the hand is dead as there are five kings and five nines in
the deck.
hitcher
Terrence Chan
Thanks Daniel, for the clarifications and addendums. Good luck to you in
this big month of late April/early May.
maxer25
"
What is this game that attracts the top players? Is it high
stakes pure "gambling" or
there is a way to beat this game on the long run? "
There is a lot of skill in Chinese poker. Dave The Wave was one
of the first people who really studied it, so I understand.
Supposedly Howard Lederer offered Dave 10% of all his action in
the game for like two years if he would teach him how to play
expertly (which took 6 months).
I heard that they have both have done very well at this game.
* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!
Daniel Negreanu
Jonathan Kaplan <NutN...@aol.com> wrote:
> In article <wEOM4.3959$JN3....@newsfeed.slurp.net>,
> "SMe" <vl...@netusa1.net> wrote:
>
> >
> > ...or is there a way to beat this game on the long run?...
> >
>
> i first learned about this game years ago when a friend of mine (D.
> Smolen) was writing the book. in discussions with him, it seemed to me
> that this game was too easy to catch on.
> guess i was wrong.
>
> i think this game is so simple that a computer program could be
written
> to play optimally. i dont like playing games where a computer can play
> better than a human, right on it's face. this seems (to me) to be that
> kind of game.
>
> no program can beat a great no limit holdem player. a computer program
> can be written that can tie the best "chinese poker" player in the
> world, right now. i think.
> stupid game.
> smile
> Jonathan
>
chinese, a computer could NEVER beat the best chinese player in the
world. The best player in the world would know EXACTLY how the
computer would set his hands. That's a huge advantage. For example,
if you were playing with someone who NEVER split two pair and you had a
decision to make with 3's and 2's in the middle and K high on top, you
could play the 2's on top. Situations like these are very common, and
a computer program wouldn't be able to adjust to a thinking human.
Daniel Negreanu
Mark T Hachey
: chinese, a computer could NEVER beat the best chinese player in the
: world. The best player in the world would know EXACTLY how the
: computer would set his hands. That's a huge advantage. For example,
: if you were playing with someone who NEVER split two pair and you had a
: decision to make with 3's and 2's in the middle and K high on top, you
: could play the 2's on top. Situations like these are very common, and
: a computer program wouldn't be able to adjust to a thinking human.
: Daniel Negreanu
this is why a computer that played optimally would randomize optimally.
you can't beat a computer playing paper-scissors-rock, but you sure could
beat some dumb humans. of course, that computer can't beat you playing
PSR either. So, I would have to say the best player in the world would not
know how the computer would set his hands, because that computer would mix
it up based on a RNG that the human could not figure out.
let me also say that what I know about chinese poker comes from reading this
thread, but computers playing games optimally is different than what you
described. all games where you assume a random distribution of cards in your
opponent's hand are easy for computers to play, such as deciding what to
throw or keep in cribbage (of course pegging in cribbage is a different story
due to recognizing patterns in your opponents play, but yadda yadda yadda...).
soi the computer would not always split the two pair, because that is not
"BEST", whereas some humans might. But, maybe splitting two pair is always
best, and the human that doesn't do it is giving up an edge. The computer would
know this very easily.
mth.
--
Jonathan Kaplan
Daniel Negreanu <kidp...@hotmail.com> wrote:
> > no program can beat a great no limit holdem player. a computer
program
> > can be written that can tie the best "chinese poker" player in the
> > world, right now. i think.
> > stupid game.
> > smile
> > Jonathan
> >
> chinese, a computer could NEVER beat the best chinese player in the
> world. The best player in the world would know EXACTLY how the
> a computer program wouldn't be able to adjust to a thinking human.
i trust your assessment more than mine, particularly since i have only
actually played the game one time. but i have a question or two.
you say, "that's not quite true" and then, "...a computer could NEVER
beat the best chinese player in the world..."
does that mean the computer could beat those "not quite" at the highest
level, meaning, only the very best humans beat the computer?
on the same note, which do you think is closer in time (and easier to
achieve), a program that plays top notch chinese poker, or a program
that plays any of the more "regular" games to the same top notch level?
i gotta believe chinese poker would be easier to program than even a
"simple" game like razz.
but, it wouldnt surprise me to be wrong.
and i am very happy to see you big hitters playing chinese while you
are waiting for other games to happen. at least then, you arent
"marking time" in some mid level game (where i might be)....grin
Jonathan
--
no matter where you go, there you are...
Jim Geary
> not
I'll take Perry's PSR computer program(which didn't even win the
computer championship)[1] heads up against human. Race to 1000
for $1000. If I can stake iocaine, I'll even lay 11:10.
As this relates to DN's comment about optimizing CP against humans,
I suspect that not only can a computer be programmed easily to
optimize a random strategy for those situations he described, it
can also be programmed to learn and play exploitatively similarly.
Those that think I'm smoking dope should check out the links below.
Jim Geary
jimgeary.com - something to bore everyone
[1] http://www.cs.ualberta.ca/~darse/rsbpc.html
and the winner:
http://ofb.net/~egnor/iocaine.html
On 25 Apr 2000, Mark T Hachey wrote:
> : That's not quite true. There are some intangibles to playing
> : chinese, a computer could NEVER beat the best chinese player in the
> : world. The best player in the world would know EXACTLY how the
Daniel Negreanu
Jonathan Kaplan <NutN...@aol.com> wrote:
> In article <8e3irb$hl8$1...@nnrp1.deja.com>,
> Daniel Negreanu <kidp...@hotmail.com> wrote:
> > > no program can beat a great no limit holdem player. a computer
> program
> > > can be written that can tie the best "chinese poker" player in the
> > > world, right now. i think.
> > > stupid game.
> > > smile
> > > Jonathan
> > >
> > chinese, a computer could NEVER beat the best chinese player in the
> > world. The best player in the world would know EXACTLY how the
> > a computer program wouldn't be able to adjust to a thinking human.
>
> actually played the game one time. but i have a question or two.
>
> you say, "that's not quite true" and then, "...a computer could NEVER
> beat the best chinese player in the world..."
> does that mean the computer could beat those "not quite" at the
highest
> level, meaning, only the very best humans beat the computer?
Exactly. I'm sure a computer program could be developed to beat most
chinese players, but not all. Also, I agree, it is a lot easier to
develop a great chinese program than a dominant poker program, simply
because there are more variables.
Daniel Negreanu
mt...@jhunix.hcf.jhu.edu (Mark T Hachey) wrote:
> : That's not quite true. There are some intangibles to playing
> : chinese, a computer could NEVER beat the best chinese player in the
> : world. The best player in the world would know EXACTLY how the
example,
> : if you were playing with someone who NEVER split two pair and you
had a
> : decision to make with 3's and 2's in the middle and K high on top,
you
> : could play the 2's on top. Situations like these are very common,
and
> : a computer program wouldn't be able to adjust to a thinking human.
> : Daniel Negreanu
>
> this is why a computer that played optimally would randomize
optimally.
> you can't beat a computer playing paper-scissors-rock, but you sure
could
> beat some dumb humans. of course, that computer can't beat you
playing
tendencies that matter. Otherwise, if the computer was playing SO
randomly, it would be making a lot of fundamental errors. Personally,
I'm not convinced that a computer program could think better than a
human, I could be wrong. Still, if you find a program worthy, I'm
willing to bet on the best chinese player in the world over the
computer.
Now, if the computer were able to adjust to his opponents'
tendencies...hmmm, well then I wouldn't like my bet very much. I guess
if that were possible, the computer would have the edge.
Daniel Negreanu
Gary Carson
Jonathan Kaplan wrote in message <8e4c93$ae5$1...@nnrp1.deja.com>...
>In article <8e3irb$hl8$1...@nnrp1.deja.com>,
> Daniel Negreanu <kidp...@hotmail.com> wrote:
>> > no program can beat a great no limit holdem player. a computer
>program
>> > can be written that can tie the best "chinese poker" player in the
>> > world, right now. i think.
>> > stupid game.
>> > smile
>> > Jonathan
>> >
>> That's not quite true. There are some intangibles to playing
>> chinese, a computer could NEVER beat the best chinese player in the
>> world. The best player in the world would know EXACTLY how the
>> computer would set his hands. That's a huge advantage. ...and
>> a computer program wouldn't be able to adjust to a thinking human.
>
>
>i trust your assessment more than mine, particularly since i have only
>actually played the game one time. but i have a question or two.
Daniel may have played some Chinese Poker. But, he's clearly never done any
computer modeling -- or at least never done it well. His assertions about
what a computer implemented playing model could or could not do are just
silly -- and wrong.
Gary Carson
A. Prock
> That's not quite true. There are some intangibles to playing
>chinese, a computer could NEVER beat the best chinese player in the
>world.
I'm sure you're basing this on all the AI research that
you've done over the past 5 years.
- Andrew
Michael OMalley
response on here, but I think I may be able to explain Daniel's claims a little
bit.
I would like to think I am a chinese poker expert, although I am not. I have
played with some of the best and worst Chinese poker players in the world. I
have played with Stu Ungar, Puggy Pearson, Jim Ward, Brent Carter, Tom McEvoy
and many others. I learned the game from Linda Johnson and Denny Axel, who I
firmly believe are the two BEST Chinese Poker players in the world.
What Danny refers to is a players ability to set hands according to their
opponent. The most common trait of an average or beginning player is to either
always keep two pair in the middle, or always split those two pair between the
middle and top hand. So, the question is, can a computer program pick up a
players tendencies? And can a computer program take into account the cards he
holds that might affect his setting? In other words, if I dont have any Aces
in my hand, it is likely that my opponent holds two Aces, and I might play a
flush in the middle with no top, as opposed to 2 small pair in the middle and
kings up top.
Straight hi Chinese poker is what most of you have I presume refered to in this
thread, but I will tell you that most of the high limit chinese poker is played
2-7 in the middle. This form takes more skill, I believe.
One of my favorite all time poker related stories:
Denny Axel is playing Men "the master". They are playing 2-7 in the middle,
and after a few hands, Men suggests that they use the other half of the deck to
speed up play. In other words, the dealer deals 4 hands of 13 cards, they each
play one hand, then go to the other two hands that have been dealt. Those of
you that know Denny will understand why this suggestion immediately got Dennys
attention. Denny agrees.
So they play the first hand of the deck, then go to the second hand. Denny
plays the hand really strange. The hands are turned over, and Men goes nuts:
"How can you play that hand that way, you can make full house, nut low and A
hi, but you play KK, 8 low and 2 pair". Later that day I ask Denny about the
hand, he tells me he knew all the cards out, and knew his 8 low was good, and
that his kings were good. And he knew Men had Qs full. So if he plays that
hand the correct way he loses, and this way he wins.
Michael O'Malley
Michael OMalley
better now.
Denny had 2344455678JJK
Instead of playing the hand
JJK
23678
44455
He played it
K86
23457
JJ445
Because he knew Men's hand was
QQ
23567
AA
Maybe that makes the story a little better? LOL, ok the story was no good
anyway, I guess you had to be there. :)
Michael O'Malley
Gary Carson
Michael OMalley wrote in message
<20000426020427...@ng-md1.aol.com>...
>I am no computer programer, hell, i even have trouble cutting and pasting a
>response on here, but I think I may be able to explain Daniel's claims a
little
>bit.
Please.
His claims are just ignorant. He can ask his poker playing buddies all
night long, and he isn't going to understand AI or math modelling or the
computer implementation of either one.
>I would like to think I am a chinese poker expert, although I am not. I
have
>played with some of the best and worst Chinese poker players in the world.
I
>have played with Stu Ungar, Puggy Pearson, Jim Ward, Brent Carter, Tom
McEvoy
>and many others. I learned the game from Linda Johnson and Denny Axel, who
I
>firmly believe are the two BEST Chinese Poker players in the world.
>What Danny refers to is a players ability to set hands according to their
>opponent. The most common trait of an average or beginning player is to
either
>always keep two pair in the middle, or always split those two pair between
the
>middle and top hand. So, the question is, can a computer program pick up a
>players tendencies? And can a computer program take into account the cards
he
>holds that might affect his setting?
Of course it can. Keeping track of things, remembering them, and analyzing
stored history or conditional distributions is part of what computers do
best.
Gary Carson
Michael OMalley
In chinese poker, there is a correct play to every hand. If you made a
computer program to play this way, I would be able to beat it, because i know
how he is going to play. If I know the correct way to play every hand, I will
have an advantage over the computer program because i can fluctuate my play to
beat a correct setting.
Maybe I am missing something, but in chinese poker, the reason an average
player does not play well is because they miss a better setting, they dont look
at all the possibilities. If I am an expert player, and i am able to see the
correct way to play a hand, how does the computer have an advantage?
Michael O'Malley
Gary Carson
--
Michael OMalley wrote in message
<20000426042649...@ng-ch1.aol.com>...
>Ok, explain it to me.
>
>In chinese poker, there is a correct play to every hand. If you made a
>computer program to play this way, I would be able to beat it, because i
know
>how he is going to play.
If a play is exploitable by an opponent, then it's not the correct way to
play.
> If I know the correct way to play every hand, I will
>have an advantage over the computer program because i can fluctuate my play
to
>beat a correct setting.
What? It's either the correct way to play or it's not.
>Maybe I am missing something, but in chinese poker, the reason an average
>player does not play well is because they miss a better setting, they dont
look
>at all the possibilities.
And, looking at all the possibilities is exactly what computers do better
than you do. How do you think the chess programs work.
Humans do tend to better at heuristics, at making quick judgments about
large ranges of possibiliteis that don't need to be looked at. That's why
it's been so hard to develop chess programs -- because the number of
possibilities is so huge that you can't look at all of them
But, if you claim an expert chinese poker can look at all the possibilities
(I don't know if he can or not), then a computer has an edge because an
computer can look at all of them without ever making a mistake.
> If I am an expert player, and i am able to see the
>correct way to play a hand, how does the computer have an advantage?
I don't know. Maybe it's a tic-tac-toe or checkers game, where the only
advantage a computer has is you might get distracted and make a mistake.
Gary Carson
>Michael O'Malley
m...@acm.org
> Ok, explain it to me.
> In chinese poker, there is a correct play to every hand.
Stop. Either there is one correct way (including player adjustments)
or there isn't. If there is a correct way then a computer will do no
better then a draw. I'm guessing you mean a correct way assuming
shown cards (your example) or the players tendencies are not known.
Lets call that way basic strategy.
> If you made a
> computer program to play this way, I would be able to beat it, because i know
> how he is going to play. If I know the correct way to play every hand, I will
> have an advantage over the computer program because i can fluctuate my play to
> beat a correct setting.
OK, so you are say that you can beat anything (person or player) who
just plays basic strategy.
> Maybe I am missing something, but in chinese poker, the reason an average
> player does not play well is because they miss a better setting, they dont look
> at all the possibilities. If I am an expert player, and i am able to see the
> correct way to play a hand, how does the computer have an advantage?
I believe the thing that you are DN are missing is that the computer
can keep track of tendencies and changes in tendencies just as well as
a player. That is the computer can make the same changes from basic
strategy that you would.
I believe even DN said that if the computer program adjusted it wasn't
a bet he'd want to take. Well only the simpliest of programs wouldn't
adjust.
mph
PBO COP
Mark T Hachey
I think it would be possible to have a computer play awesomely without it
knowing anything about tendencies. Basically, you would program the computer to
play the game under this criterion : Given all possible arrangements of the
remaining 39 cards in the opponents hands, how to I arrange mine to minimize
my losses.
This would be a devastating player. It would find what is called in game theory
a maximin strategy, which might include playing the same hand differently on
two different occasions--just like a computer playing paper-scissor-rock
can minimize its losses by playing completely randomly. So, the computer
would sometimes split two pair, sometimes not, but you wouldn't be able to
outguess it. Just like you can't outguess the bluffer in the bluffing/game
theory section of Theory of Poker if he randomizes properly.
It can be proved that this computer can not be beaten (in the long run).
Here's an example. Stop reading now if you really don't what to think about it.
Consider the following table
-1 10
4 3
where location (i,j) represents computer's gain if he plays strategy j and the
human plays strategy i. That is, if you choose 1, and computer chooses 2, the
computer wins 10 from you, etc.
The computer can always win if it plays strategy 2
all the time, but realizing this, the human would play his strategy 2 so he
only loses 3, instead of 4, so the
computer might switch to strategy 1 to get the gain of 10. It turns out that
optimal strategy for the computer would be to play a certain percentage of
each strategy, and to choose randomly (the exact percentage can be found
easily). The human would also have a strategy randomly split up between
the two.
If the table were
-1 10
2 3
the human should always play strategy 1. He has no incentive to deviate from
it (see why). The computer should therefore always play strategy 2.
In Chinese Poker (remember that game), the computer makes up 52choose13 of
these tables, each of which has 13choose5 X 8choose5 rows (for his strategies),
and even more columns (for its opponents' strategies).
It can then select an optimal strategy based on these tables in such a way
that it can not be beaten (on avg). Now, if the human knew it was playing
such a strategy, the best he could do would be to break even. And that would
be playing optimally for him. A normal human would be wrecked.
Now, I'm not an expert, but I've studied a little game theory. I'd really like
to see someone who has studied game theory give their opinion about this
analysis. Or at least come up with a more easily understood example.
Has Sklansky studied game theory? He talks about it.
The difference between this and normal poker is that betting and knowledge of
other cards give you a lot of information. I think I'll stop here.
mth.
--
Gary Carson
PBO COP wrote in message <20000426133038...@ng-fa1.aol.com>...
>Gary is drunk again..
Sometimes even when I'm sober I'm not as tolerant of ignorant fools as I
should be.
Gary Carson
PBO COP
should do some research before you open your own ignorant mouth !! Loser..
Gary Carson
PBO COP wrote in message <20000426143546...@ng-fa1.aol.com>...
>Ha......well, the person that posted that is much smarter than u. So maybe
you
>should do some research before you open your own ignorant mouth !! Loser..
How did you know about the loser part? I lost my socks? Did you find them?
It's really hot today and I hate wearing shoes with no socks when it's hot.
I guess I could buy a second pair though. Hey, do you think you could stake
me to a second pair of socks?
Gary Carson
A. Prock
>So, the question is, can a computer program pick up a
>players tendencies?
Yes.
- Andrew
PBO COP
Jim Geary
and switching to expoitative in the meta-roshambo context of adjusting.
Jim Geary
jimgeary.com - something to bore everyone
NChrisH
<<
NOOOO...the stock market is down...can't back you anymore.. >>
You can make money in a down market as well as a bull market. Just like
changing your table strategy depending on game conditions.
Chris
Daniel Negreanu
> Straight hi Chinese poker is what most of you have I presume refered
to in this
> thread, but I will tell you that most of the high limit chinese poker
is played
> 2-7 in the middle. This form takes more skill, I believe.
It may take more skill, but most of the high limit chinese games are
NOT played deuce in the middle. The Bellagio is the new chinese
capital of the world, and they play straight high, no surrender, with
bonuses and naturals. Daniel Negreanu
Daniel Negreanu
average
> > player does not play well is because they miss a better setting,
they dont look
> > at all the possibilities. If I am an expert player, and i am able
to see the
> > correct way to play a hand, how does the computer have an advantage?
>
> I believe the thing that you are DN are missing is that the computer
> can keep track of tendencies and changes in tendencies just as well as
> a player. That is the computer can make the same changes from basic
> strategy that you would.
>
> I believe even DN said that if the computer program adjusted it wasn't
> a bet he'd want to take. Well only the simpliest of programs wouldn't
> adjust.
>
my post and replied on a previous paragraph I wrote. I never claimed
to know anything about computers, and at the end of my post I was
conceding that a perfect chinese poker program could probably beat the
best chinese player in the world.
program that read players tendencies and play accordingly. I
mentioned in the post that Carson rudely snipped and took out of
context, that if that were possible, I'm wrong. Yet he still called me
an ignorant fool. I didn't reply to that post, because nothing good
ever comes out of having a conversation with Mr.Carson. I made a post
conceding that I may be wrong, yet I'm an ignorant fool? Hmmm...If I
were ignorant wouldn't I have been pig-headed and stubborn about it? I
don't know, maybe I'm so ignorant, that I don't know the definition of
ignorant.
Sincerely, the ignorant fool, Daniel Negreanu desperately trying to
learn this whole poker thing from the great poker mind, Gary Carson,
I'm in awe of his greatness. Keep plugging away Gary, you'll beat that
15-30 game someday. Another 15 years of poker study, and you may
actually get into a 30-60 game. Lookout poker world, Gary's steppin'
it up! OOhhh, aahhh, not Gary! We're scared of Gary, he writes
books!!!!!!(That wasn't to poke fun at any other authors...just Gary)
Gary Carson
Daniel Negreanu wrote in message <8e98kl$kii$1...@nnrp1.deja.com>...
>program that read players tendencies and play accordingly. I
>mentioned in the post that Carson rudely snipped and took out of
>context, that if that were possible, I'm wrong. Yet he still called me
>an ignorant fool. I didn't reply to that post, because nothing good
>ever comes out of having a conversation with Mr.Carson. I made a post
>conceding that I may be wrong, yet I'm an ignorant fool?
Yes, you are.
>Hmmm...If I
>were ignorant wouldn't I have been pig-headed and stubborn about it? I
>don't know, maybe I'm so ignorant, that I don't know the definition of
>ignorant.
>
>Sincerely, the ignorant fool, Daniel Negreanu desperately trying to
>learn this whole poker thing from the great poker mind, Gary Carson,
>I'm in awe of his greatness. Keep plugging away Gary, you'll beat that
>15-30 game someday. Another 15 years of poker study, and you may
>actually get into a 30-60 game. Lookout poker world, Gary's steppin'
>it up! OOhhh, aahhh, not Gary! We're scared of Gary, he writes
>books!!!!!!(That wasn't to poke fun at any other authors...just Gary)
Gee. Did I hurt your feelings?
Gary Carson
Daniel Negreanu
Honestly, I'm just sick of your bullshit. You talk this great game,
like you are the God of poker, yet I don't know anyone that's seen it.
Heck, you know where I am, if you want to play with me so bad I'm not
to hard to find.
For a long time I didn't respond to your posts, but obvioudly you
didn't get the message. We have our differences and I was willing to
bury them, all you want to do is dig them up again. Well, I wont be a
complete pushover. You want to consistenstly trash talk everything I
say, fine. Question is, will you back up all that lip at a poker
table. I didn't think so. Daniel Negreanu
Gary Carson
>> Gary Carson
>
> Honestly, I'm just sick of your bullshit. You talk this great game,
>like you are the God of poker, yet I don't know anyone that's seen it.
LOL. Do I know anyone that's played with you? Who cares?
>Heck, you know where I am, if you want to play with me so bad I'm not
>to hard to find.
Why would I care?
> For a long time I didn't respond to your posts, but obvioudly you
>didn't get the message.
What was the message?
> We have our differences and I was willing to
>bury them, all you want to do is dig them up again.
What differences? I just think you're funny.
> Well, I wont be a
>complete pushover. You want to consistenstly trash talk everything I
>say, fine. Question is, will you back up all that lip at a poker
>table. I didn't think so. Daniel Negreanu
I don't trash talk everything you say. Just your really stupid
assertitions.
Get a girlfreind.
Gary Carson
PBO COP
Lone Locust of the Apocalypse
>Gary...learn how to spell girlfriend and maybe you will find one !!
This has got to be one of the most pathetic attempts at a flame that
I have ever seen.
Maybe those of you with personal vendettas against each other could
take it to e-mail instead of polluting the group.
PBO COP
Tom Weideman
pa...@cs.Stanford.EDU wrote on 5/2/00 10:27 PM:
> Daniel Negreanu (kidp...@hotmail.com) wrote:
>
> : It doesn't matter if the computer plays randomly, it's simply the
> : tendencies that matter. Otherwise, if the computer was playing SO
> : randomly, it would be making a lot of fundamental errors. Personally,
> : I'm not convinced that a computer program could think better than a
> : human, I could be wrong. Still, if you find a program worthy, I'm
> : willing to bet on the best chinese player in the world over the
> : computer.
>
> Are you serious? Pick whoever you want, give me a month to write the
> software (in consultation with some other r.g.p'ers, of course), and we'll
> play. $1000 a point sound good? I might be willing to play higher once I
> do the analysis and maybe pool some cash.
The code has already been written by mutliple people, and I wouldn't be
surprised if Patri could repeat the task. In fact, what demonstrates how
truly uninformed Mr. Negreanu is on this issue is the fact that the best
Chinese Poker players in the world play great because they have learned
proper strategy from (and are the best at emulating) - you guessed it -
computer programs.
I've really enjoyed reading rgp over the last year or so, since it has given
the chance for so many people who are "elite inner circle" of poker to
clearly demonstrate for us how little they know about so many things
involved with gambling in general and poker in particular. Until they
discovered rgp, most high-profile tournament r00lerz had to go to the
trouble of publishing a book to blow their mystique out of the water. Now
they can do it with a 5 minute post, and then prove their cluelessness
wasn't a fluke with additional posts. The irony is that these folks
probably saw rgp as a vehicle for expanding their fame. Heh.
Tom Weideman
Patri Forwalter-Friedman
: It doesn't matter if the computer plays randomly, it's simply the
: tendencies that matter. Otherwise, if the computer was playing SO
: randomly, it would be making a lot of fundamental errors. Personally,
: I'm not convinced that a computer program could think better than a
: human, I could be wrong. Still, if you find a program worthy, I'm
: willing to bet on the best chinese player in the world over the
: computer.
Are you serious? Pick whoever you want, give me a month to write the
software (in consultation with some other r.g.p'ers, of course), and we'll
play. $1000 a point sound good? I might be willing to play higher once I
do the analysis and maybe pool some cash.
With all due respect, you have strayed into an area where you are not an
expert, but many r.g.p'ers are. Computer programs (AI's) have their
strengths and weaknesses in various areas of game playing. They could
kick ass at Chinese Poker. If you don't believe me, I will take your
money to prove it.
Explaining why to a layman (no insult) is difficult. I can try to
do it in person, but I'm not gonna type it here. For those who are
versed in AI, the adjustments Daniel is talking about are all deviations
from minimax strategy, and so a computer playing minimax strategy will
beat these opponents. Since minimax strategy for chinese poker is pretty
complicated, a computer will stick to it better than a person. So human
opponents will in general not be playing minimax, thus they will lose to a
computer.
Knowing how the computer sets its hands is irrelevant, since they are set,
by definition of minimax, in non-exploitable ways. The best you can do is
imitate the strategy and break even, but it is complicated enough that
this will be hard for a human to do.
Since poker is a game of exploiting weak strategies, and the existence of
many different strategies means your own strategies must be tailored, the
ideas of minimax and non-exploitability are not natural for poker players
to understand.
Chinese poker as Daniel describes it is being played a little like real
poker, winning by adjusting to ones opponents strategies. This may be the
way good poker players beat bad poker players in Chinese Poker, but there
are ways to play which will be less exploitative (win less from bad
players), but will beat the good players, and that is how the computer
would play.
: Now, if the computer were able to adjust to his opponents'
: tendencies...hmmm, well then I wouldn't like my bet very much. I guess
: if that were possible, the computer would have the edge.
Of course they would adjust to their opponents.
Patri
Creepshow
I am no poker or computer person, but I have followed enough chess to firmly
believe that if you can program a computor to take the measure of Garry
Kasparov in a match, a feat not performed by any human opponent, then Chinese
poker would be the last thing you would actually want to wager $ vs. a
computer. sorry about run-on sentence.
I think a computor program would however, have a more difficult timebeating in
"pass the trash", my speicialy.
Best of luck,Gary
Marc Gilutin
Forwalter-Friedman) wrote:
>Daniel Negreanu (kidp...@hotmail.com) wrote:
>
>: It doesn't matter if the computer plays randomly, it's simply the
>: tendencies that matter. Otherwise, if the computer was playing SO
>: randomly, it would be making a lot of fundamental errors. Personally,
>: I'm not convinced that a computer program could think better than a
>: human, I could be wrong. Still, if you find a program worthy, I'm
>: willing to bet on the best chinese player in the world over the
>: computer.
>
>Are you serious? Pick whoever you want, give me a month to write the
>software (in consultation with some other r.g.p'ers, of course), and we'll
>play. $1000 a point sound good? I might be willing to play higher once I
>do the analysis and maybe pool some cash.
>
>With all due respect, you have strayed into an area where you are not an
>expert, but many r.g.p'ers are. Computer programs (AI's) have their
>strengths and weaknesses in various areas of game playing. They could
>kick ass at Chinese Poker. If you don't believe me, I will take your
>money to prove it.
>
I've only been playing CP for a little while but I've played with a
LOT of players who have been playing this game for so long that they
play perfectly...can't get much edge on them.
Cheers
Marc
And about that $1k/pt, there are stories of side games at Binions that
are going for considerably more per/unit kondition.
Gary Carson
Patri Forwalter-Friedman wrote in message
<8eodcs$dn5$1...@nntp.Stanford.EDU>...
>Daniel Negreanu (kidp...@hotmail.com) wrote:
>
>: It doesn't matter if the computer plays randomly, it's simply the
>: tendencies that matter. Otherwise, if the computer was playing SO
>: randomly, it would be making a lot of fundamental errors. Personally,
>: I'm not convinced that a computer program could think better than a
>: human, I could be wrong. Still, if you find a program worthy, I'm
>: willing to bet on the best chinese player in the world over the
>: computer.
>
>Are you serious? Pick whoever you want, give me a month to write the
>software (in consultation with some other r.g.p'ers, of course), and we'll
>play. $1000 a point sound good? I might be willing to play higher once I
>do the analysis and maybe pool some cash.
Say something rude about his mother. It'll piss him off and he'll lay you
10-1.
Gary Carson
>
>: Now, if the computer were able to adjust to his opponents'
>: tendencies...hmmm, well then I wouldn't like my bet very much. I guess
>: if that were possible, the computer would have the edge.
>
>Of course they would adjust to their opponents.
Well, now you screwed up. If you'd have snipped that paragraph then you'd
have pissed him off. Now he'll never offer the 10-1 challange.
Of course, he had previousl made an assertion akin to "The sky is green".
This last paragrpah simply says, "But if the sky isn't green, then it's some
other color". I'm not sure what it adds, but Daniel thought it was a very
important paragraph, with an important statement. Must have something to do
with Canadian money.
Gary Carson
Jim Geary
be something I'd hazard a computer would be especially adept at among
the continuum of poker game conquerability.
Jim Geary
jimgeary.com - something to bore everyone
David E. Wallace
Jim Geary <jay...@primenet.com> wrote:
>> beat some dumb humans. of course, that computer can't beat you playing
>> PSR either. So, I would have to say the best player in the world would
>> not
>
>I'll take Perry's PSR computer program(which didn't even win the
>computer championship)[1] heads up against human. Race to 1000
>for $1000. If I can stake iocaine, I'll even lay 11:10.
If the human can roll dice while playing, I think you might
get several takers at those odds.
--
Dave Wallace (wal...@netcom.com)
It is quite humbling to realize that the storage occupied by the longest
line from a typical Usenet posting is sufficient to provide a state space
so vast that all the computation power in the world can not conquer it.
David E. Wallace
Patri Forwalter-Friedman <pa...@cs.Stanford.EDU> wrote:
>Daniel Negreanu (kidp...@hotmail.com) wrote:
>
>: It doesn't matter if the computer plays randomly, it's simply the
>: tendencies that matter. Otherwise, if the computer was playing SO
>: randomly, it would be making a lot of fundamental errors. Personally,
>: I'm not convinced that a computer program could think better than a
>: human, I could be wrong. Still, if you find a program worthy, I'm
>: willing to bet on the best chinese player in the world over the
>: computer.
>
>Are you serious? Pick whoever you want, give me a month to write the
>software (in consultation with some other r.g.p'ers, of course), and we'll
>play. $1000 a point sound good? I might be willing to play higher once I
>do the analysis and maybe pool some cash.
With all due respect, I think it would take more than a month to code
up a world-class chinese poker player from scratch. If you start from
a really good chinese poker playing program, then you could probably
add some significant enhancements in a month that could make it better.
I do agree that given sufficient programming effort, a program will
probably beat a human a human player.
>With all due respect, you have strayed into an area where you are not an
>expert, but many r.g.p'ers are. Computer programs (AI's) have their
>strengths and weaknesses in various areas of game playing. They could
>kick ass at Chinese Poker. If you don't believe me, I will take your
>money to prove it.
>
>Explaining why to a layman (no insult) is difficult. I can try to
>do it in person, but I'm not gonna type it here. For those who are
>versed in AI, the adjustments Daniel is talking about are all deviations
>from minimax strategy, and so a computer playing minimax strategy will
>beat these opponents. Since minimax strategy for chinese poker is pretty
>complicated, a computer will stick to it better than a person. So human
>opponents will in general not be playing minimax, thus they will lose to a
>computer.
Ah, but do you know the true minmax strategy for Chinese poker so
that you can program it? If not, you have a fair amount of research
to do before you can implement it. Heads-up limit hold'em also has
a minmax mixed-strategy solution, but it's so complicated that
no one (to my knowledge) has been able to figure it out, even
with the help of a computer.
It's also not necessarily the case that a minmax strategy will
always beat a non-minmax strategy. It won't lose to it (assuming
that the game is fair), but it won't beat an opponent whose pure
strategy components are part of some optimal minmax strategy,
even if the opponent combines those components in a non-optimal
(and theoretically exploitable) way. For example, in Roshambo,
the minmax strategy (choose each option randomly with probability 1/3)
won't beat an opponent who always chooses "rock."
>Knowing how the computer sets its hands is irrelevant, since they are set,
>by definition of minimax, in non-exploitable ways. The best you can do is
>imitate the strategy and break even, but it is complicated enough that
>this will be hard for a human to do.
>
>Since poker is a game of exploiting weak strategies, and the existence of
>many different strategies means your own strategies must be tailored, the
>ideas of minimax and non-exploitability are not natural for poker players
>to understand.
>
>Chinese poker as Daniel describes it is being played a little like real
>poker, winning by adjusting to ones opponents strategies. This may be the
>way good poker players beat bad poker players in Chinese Poker, but there
>are ways to play which will be less exploitative (win less from bad
>players), but will beat the good players, and that is how the computer
>would play.
It will not lose to the good players, but it won't necessarily beat them.
>: Now, if the computer were able to adjust to his opponents'
>: tendencies...hmmm, well then I wouldn't like my bet very much. I guess
>: if that were possible, the computer would have the edge.
>
>Of course they would adjust to their opponents.
With sufficient programming effort, such adjustments could be
incorporated. But you are moving well beyond minmax strategies
at this point, and I wouldn't assume that any program thrown together
in a month or two would incorporate such features.
Daniel Negreanu
pa...@cs.Stanford.EDU (Patri Forwalter-Friedman) wrote:
> Daniel Negreanu (kidp...@hotmail.com) wrote:
>
> : It doesn't matter if the computer plays randomly, it's simply the
> : tendencies that matter. Otherwise, if the computer was playing SO
> : randomly, it would be making a lot of fundamental errors.
Personally,
> : I'm not convinced that a computer program could think better than a
> : human, I could be wrong. Still, if you find a program worthy, I'm
> : willing to bet on the best chinese player in the world over the
> : computer.
>
> Are you serious? Pick whoever you want, give me a month to write the
> software (in consultation with some other r.g.p'ers, of course), and
we'll
> play. $1000 a point sound good? I might be willing to play higher
once I
> do the analysis and maybe pool some cash.
>
an
> expert, but many r.g.p'ers are. Computer programs (AI's) have their
> strengths and weaknesses in various areas of game playing. They could
> kick ass at Chinese Poker. If you don't believe me, I will take your
> money to prove it.
>
> Explaining why to a layman (no insult) is difficult. I can try to
> do it in person, but I'm not gonna type it here. For those who are
> versed in AI, the adjustments Daniel is talking about are all
deviations
> from minimax strategy, and so a computer playing minimax strategy will
> beat these opponents. Since minimax strategy for chinese poker is
pretty
> complicated, a computer will stick to it better than a person. So
human
> opponents will in general not be playing minimax, thus they will lose
to a
> computer.
>
set,
> by definition of minimax, in non-exploitable ways. The best you can
do is
> imitate the strategy and break even, but it is complicated enough that
> this will be hard for a human to do.
>
> Since poker is a game of exploiting weak strategies, and the
existence of
> many different strategies means your own strategies must be tailored,
the
> ideas of minimax and non-exploitability are not natural for poker
players
> to understand.
>
> Chinese poker as Daniel describes it is being played a little like
real
> poker, winning by adjusting to ones opponents strategies. This may
be the
> way good poker players beat bad poker players in Chinese Poker, but
there
> are ways to play which will be less exploitative (win less from bad
> players), but will beat the good players, and that is how the computer
> would play.
>
> : tendencies...hmmm, well then I wouldn't like my bet very much. I
guess
> : if that were possible, the computer would have the edge.
>
> Of course they would adjust to their opponents.
>
Obviously I'm a little computer illiterate, and that's why I
retracted my original statement, with "I wouldn't like my bet very
much." I also finished the post with, "the computer would have the
edge". so I'm agreeing with you guys, since I guess it is possible, I
didn't know.
What I guess I was trying to say is this:
If a computer set up his hands with a common point system, using the
highest point total available, a human would crush it. That's all.
Daniel Negreanu
Tom Weideman <zwi...@pacbell.net> wrote:
> in article 8eodcs$dn5$1...@nntp.Stanford.EDU, Patri Forwalter-Friedman at
> pa...@cs.Stanford.EDU wrote on 5/2/00 10:27 PM:
>
> >
> > : It doesn't matter if the computer plays randomly, it's simply
the
> > : tendencies that matter. Otherwise, if the computer was playing SO
> > : randomly, it would be making a lot of fundamental errors.
Personally,
> > : I'm not convinced that a computer program could think better than
a
> > : human, I could be wrong. Still, if you find a program worthy, I'm
> > : willing to bet on the best chinese player in the world over the
> > : computer.
> >
> > Are you serious? Pick whoever you want, give me a month to write
the
> > software (in consultation with some other r.g.p'ers, of course),
and we'll
> > play. $1000 a point sound good? I might be willing to play higher
once I
> > do the analysis and maybe pool some cash.
>
be
> surprised if Patri could repeat the task. In fact, what demonstrates
how
> truly uninformed Mr. Negreanu is on this issue is the fact that the
best
> Chinese Poker players in the world play great because they have
learned
> proper strategy from (and are the best at emulating) - you guessed
it -
> computer programs.
>
It can give you the best strategy against players who make all types of
common errors. I didn't want to discuss the program here, because I
wasn't willing to give it to anyone. Obviously I didn't write it, but
I'm guessing you know who did, and I know exactly who you are talking
about. Daniel Negreanu