see link below:
http://news.yahoo.com/s/afp/20070725/lf_afp/lifestylecanada
According to the article: chess, checkers, backgammon can already beat
humans consistently...
xvp: Humans narrowly beat computer in poker battle. Could vp be next?
What would be the point of a person competing with a computer in video
poker? The computer could never lose if it is programmed correctly -- it
will always make the optimal play.
···
On 7/25/07, gilbert_616 <gilbert_616@yahoo.com> wrote:
see link below:
http://news.yahoo.com/s/afp/20070725/lf_afp/lifestylecanadaAccording to the article: chess, checkers, backgammon can already beat
humans consistently...
[Non-text portions of this message have been removed]
--- In vpFREE@yahoogroups.com, "Ken Kirschner" <ken.kirschner@...>
wrote:
What would be the point of a person competing with a computer in video
poker? The computer could never lose if it is programmed correctly --
it
will always make the optimal play.
Absolutely, there are some games that are "solveable". One of the
easiest is Tic-Tac-Toe which we can all play perfectly. VP is somewhat
more complicated as is Checkers and Chess. However, given enough
computing power these games are all solveable as well. Once this is
done the computer will never lose and two computers would always end up
in a draw.
Poker is a little different animal since there is no way to program
bluffing in a deterministic way. Therefore, even two computers
programmed with all the math rules and a set of algorithms to simulate
bluffing would not necessarily end up in a draw. However, I agree that
they will eventually beat humans as these algorithms get better and
better. Will the next WSOP winner be a computer? Or, a human with an
imbedded chip? Who knows?
Dick
As a matter of fact, checkers has fairly recently been "solved":
http://www.technologyreview.com/blog/editors/17660/
Unless there is an error in the article, it would be impossible for
any human to ever beat this checkers computer program in a series of
games in which first move alternates.
One game in which computers lag well behind top players is Go, the
Asian board game. There is no luck in this game, but the board
remains too big for a chess or checkers approach that relies heavily
on brute force. Strong amateur players have no problem beating the
best computer Go programs.
--Dunbar
--- In vpFREE@yahoogroups.com, "Ken Kirschner" <ken.kirschner@>
wrote:
>
> What would be the point of a person competing with a computer in
video
> poker? The computer could never lose if it is programmed
correctly --
it
> will always make the optimal play.Absolutely, there are some games that are "solveable". One of the
easiest is Tic-Tac-Toe which we can all play perfectly. VP is
somewhat
more complicated as is Checkers and Chess. However, given enough
computing power these games are all solveable as well. Once this
is
done the computer will never lose and two computers would always
end up
in a draw.
Poker is a little different animal since there is no way to
program
bluffing in a deterministic way. Therefore, even two computers
programmed with all the math rules and a set of algorithms to
simulate
bluffing would not necessarily end up in a draw. However, I agree
that
they will eventually beat humans as these algorithms get better
and
better. Will the next WSOP winner be a computer? Or, a human with
an
···
--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:
imbedded chip? Who knows?
Dick
Yes - the computer could lose because the person
playing against it would hold the 10h and draw the
other 4 to the royal.
What would be the point of a person competing with a
computer in video
poker? The computer could never lose if it is
programmed correctly -- it
will always make the optimal play.>
> see link below:
>
http://news.yahoo.com/s/afp/20070725/lf_afp/lifestylecanada
···
--- Ken Kirschner <ken.kirschner@gmail.com> wrote:
On 7/25/07, gilbert_616 <gilbert_616@yahoo.com> > wrote:
>
> According to the article: chess, checkers,
backgammon can already beat
> humans consistently...
>
>
>[Non-text portions of this message have been
removed]
____________________________________________________________________________________
Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more.
http://mobile.yahoo.com/go?refer=1GNXIC
Ken wrote: What would be the point of a person competing with a computer
in video poker? The computer could never lose if it is programmed
correctly -- it will always make the optimal play.
I strongly disagree with this.
Making the optimal play is guaranteed to work if you're playing millions
of hands or more. If you're playing 200 hands in the match, anything can
work. Assume you're playing 9/6 DDB and the hand dealt is 'AKT5'4. The
computer will quickly hold 'AKT5'. It's not hard to see that in a short
match, 'AKT' could work much better. Another simple example from the
same game is do you hold the kicker from AAA45? If you're playing
forever, you should quickly hold AAA. In a short match, holding AAA4
could work out very nicely.
Checkers has been solved absolutely. There is a logic involved and
probabilities are not involved. For every move there's a countermove and
you can say absolutely that this play is either the best or tied for the
best. Every time. In video poker, however, probabilities ARE involved.
You cannot say always say that one play will beat another every time,
with every possible draw of the card. You can only say that one play
will beat another ON AVERAGE or OVER THE LONG RUN.
Even with probabilities aside, there are tie hands in video poker. From
an unsuited T8764, in several different games holding T876 (and hoping
for a 9) is equally desireable as holding 8764 (and hoping for a 5) are
exactly tied in EV. (I'm limiting the discussion for now to games where
these types of inside straights ARE held. Obviously there are games
where these inside straights could be thrown away.) The computer will
hold one of these. You could hold the other. Four times out of 47, your
play will turn out better than the computer's --- even though the
computer is playing perfectly. If the rest of your plays are perfect (as
is definitely possible in relatively simple games like 10/7 DB), if this
one hand was the only difference between your play and the computer's,
you're going to win the match sometimes.
In fact, if you KNEW which pair the computer holds (as is possible to
figure out from AUTOHOLD on games like Video Poker for Winners and
WinPoker), you would make the "counter" plays until you were ahead, and
then switch to exactly the same plays. This would allow you to win more
often. It would not be difficult for the computer programmer to
randomize the selection process to preclude you doing this, but the
machine could technically be considered playing perfectly whether or not
it included this randomizing process.
Bob Dancer
For a 3-day free trial of Video Poker for Winners, the best video poker
computer trainer ever invented, go to //www.videopokerforwinners.com
I think the point of the question assumes that enough hands would need
to be played to discount the luck factor. If not, anyone one could win
including a monkey randomly pressing keys. I doubt that was the intent
of the question.
OTOH, maybe we can described a VP-like game where one person selects a
10 card sequence using some rules that require certain card sequences
and the other person tries to select the cards that he thinks will
provide the best results. Then, the roles are reversed. After a fixed
number of hands are played the scores are added up and a winner
declared. Where is Milton Bradley when I need him 
Dick
···
--- In vpFREE@yahoogroups.com, Wild Bill <wcimo@...> wrote:
Yes - the computer could lose because the person
playing against it would hold the 10h and draw the
other 4 to the royal.
Bob,
We have a functional disagreement on what it means to beat the computer.
Luck of the draw is a pretty crappy way to rate things. So you have to look
at the long-term in my opinion. Or you look at the total EV of all plays.
Or some other luck-free way of measuring things.
There is a key piece of information that was omitted from the poker article
that removes luck from the test. Each hand was played twice in different
rooms. In one room, the human played hand A against the computer's hand B;
in the other room, the hands were reversed. This completely removes luck of
the hands being dealt from the test and it is the only fair way to compare
human play to computer play.
The reason other games have not had similar needs is because luck plays
absolutely no role in chess, checkers, tic-tac-toe, etc. But luck plays a
role in the results in games such as poker and video poker, so you have to
remove luck from the equation to truly know who/what is the better player.
If you leave luck in for video poker, then it is a completely bogus test.
Ken
···
On 7/25/07, Bob Dancer <bob.dancer@compdance.com> wrote:
Ken wrote: What would be the point of a person competing with a computer
in video poker? The computer could never lose if it is programmed
correctly -- it will always make the optimal play.I strongly disagree with this.
Making the optimal play is guaranteed to work if you're playing millions
of hands or more. If you're playing 200 hands in the match, anything can
work. Assume you're playing 9/6 DDB and the hand dealt is 'AKT5'4. The
computer will quickly hold 'AKT5'. It's not hard to see that in a short
match, 'AKT' could work much better. Another simple example from the
same game is do you hold the kicker from AAA45? If you're playing
forever, you should quickly hold AAA. In a short match, holding AAA4
could work out very nicely.Checkers has been solved absolutely. There is a logic involved and
probabilities are not involved. For every move there's a countermove and
you can say absolutely that this play is either the best or tied for the
best. Every time. In video poker, however, probabilities ARE involved.
You cannot say always say that one play will beat another every time,
with every possible draw of the card. You can only say that one play
will beat another ON AVERAGE or OVER THE LONG RUN.Even with probabilities aside, there are tie hands in video poker. From
an unsuited T8764, in several different games holding T876 (and hoping
for a 9) is equally desireable as holding 8764 (and hoping for a 5) are
exactly tied in EV. (I'm limiting the discussion for now to games where
these types of inside straights ARE held. Obviously there are games
where these inside straights could be thrown away.) The computer will
hold one of these. You could hold the other. Four times out of 47, your
play will turn out better than the computer's --- even though the
computer is playing perfectly. If the rest of your plays are perfect (as
is definitely possible in relatively simple games like 10/7 DB), if this
one hand was the only difference between your play and the computer's,
you're going to win the match sometimes.In fact, if you KNEW which pair the computer holds (as is possible to
figure out from AUTOHOLD on games like Video Poker for Winners and
WinPoker), you would make the "counter" plays until you were ahead, and
then switch to exactly the same plays. This would allow you to win more
often. It would not be difficult for the computer programmer to
randomize the selection process to preclude you doing this, but the
machine could technically be considered playing perfectly whether or not
it included this randomizing process.Bob Dancer
For a 3-day free trial of Video Poker for Winners, the best video poker
computer trainer ever invented, go to //www.videopokerforwinners.com
[Non-text portions of this message have been removed]
--- In vpFREE@yahoogroups.com, "Ken Kirschner" <ken.kirschner@...>
wrote:
If you leave luck in for video poker, then it is a completely bogus
test.
I thought that was pretty obvious from the beginning.
Dick
If you don't have a random draw then you don't have video poker.
But, maybe I'm wrong about that?
Cheers
.
--- In vpFREE@yahoogroups.com, "Ken Kirschner" <ken.kirschner@>
wrote:
> If you leave luck in for video poker, then it is a completely
bogus
···
.--- In vpFREE@yahoogroups.com, "mroejacks" <rgmustain@...> wrote:
test.
I thought that was pretty obvious from the beginning.
Dick
The point of these competitions is to see whether or not the computer
program is _better_ than the human. IMO, you cannot be considered better if
you simply win some portion of the time due to random luck. None of the man
versus computer competitions work that way currently, so why would we create
a video poker competition that allowed pure luck to be a determining factor?
If we wanted to know whether or not someone can win through random luck,
there would be no real point -- anyone can get lucky in the short-term. In
the long-term, skill will out. And that skill is what these man versus
computer competitions are all about.
Ken
···
On 7/26/07, whitejeeps <whitejeeps@yahoo.com> wrote:
If you don't have a random draw then you don't have video poker.
But, maybe I'm wrong about that?
Cheers
.
.--- In vpFREE@yahoogroups.com <vpFREE%40yahoogroups.com>, "mroejacks"
<rgmustain@...> wrote:
>
> --- In vpFREE@yahoogroups.com <vpFREE%40yahoogroups.com>, "Ken
Kirschner" <ken.kirschner@>
> wrote:
> > If you leave luck in for video poker, then it is a completely
bogus
> test.
>
> I thought that was pretty obvious from the beginning.
>
> Dick
>
[Non-text portions of this message have been removed]