human battles bot

catrina m

Dormant account
PABaccred
Joined
Feb 8, 2007
Location
US of A
Card Sharks to Battle Computer at Poker


Jul 21, 11:56 PM (ET)

By MATT CRENSON



NEW YORK (AP) - Poker champion Phil Laak has a good chance of winning when he sits down this week to play 2,000 hands of Texas Hold'em - against a computer.

It may be the last chance he gets. Computers have gotten a lot better at poker in recent years; they're good enough now to challenge top professionals like Laak, who won the World Poker Tour invitational in 2004.

But it's only a matter of time before the machines take a commanding lead in the war for poker supremacy. Just as they already have in backgammon, checkers and chess, computers are expected to surpass even the best human poker players within a decade. They can already beat virtually any amateur player.

"This match is extremely important, because it's the first time there's going to be a man-machine event where there's going to be a scientific component," said University of Alberta computing science professor Jonathan Schaeffer.

The Canadian university's games research group is considered the best of its kind in the world. After defeating an Alberta-designed program several years ago, Laak was so impressed that he estimated his edge at a mere 5 percent. He figures he would have lost if the researchers hadn't let him examine the programming code and practice against the machine ahead of time.

"This robot is going to do just fine," Laak predicted.

The Alberta researchers have endowed the $50,000 contest with an ingenious design, making this the first man-machine contest to eliminate the luck of the draw as much as possible.

Laak will play with a partner, fellow pro Ali Eslami. The two will be in separate rooms, and their games will be mirror images of one another, with Eslami getting the cards that the computer received in its hands against Laak, and vice versa.

That way, a lousy hand for one human player will result in a correspondingly strong hand for his partner in the other room. At the end of the tournament the chips of both humans will be added together and compared to the computer's.

The two-day contest, beginning Monday, takes place not at a casino, but at the annual conference of the Association for the Advancement of Artificial Intelligence in Vancouver, British Columbia. Researchers in the field have taken an increasing interest in poker over the past few years because one of the biggest problems they face is how to deal with uncertainty and incomplete information.

"You don't have perfect information about what state the game is in, and particularly what cards your opponent has in his hand," said Dana S. Nau, a professor of computer science at the University of Maryland in College Park. "That means when an opponent does something, you can't be sure why."

As a result, it is much harder for computer programmers to teach computers to play poker than other games. In chess, checkers and backgammon, every contest starts the same way, then evolves through an enormous, but finite, number of possible states according to a consistent set of rules. With enough computing power, a computer could simply build a tree with a branch representing every possible future move in the game, then choose the one that leads most directly to victory.

That's essentially the strategy IBM's Deep Blue computer used to defeat chess champion Gary Kasparov in their famous 1997 match. No computer can calculate every single possible move in a chess game, but today's best chess programs can see an astounding 18 moves ahead.

Yet poker involves not just myriad possibilities but uncertainty, both about what cards the opponent is holding and more importantly, how he is going to play them.

"It's mandatory for you to understand how the other guy approaches the game. This is critical information in poker, and it's not true of any of these other games that we've studied in academia," said Darse Billings, a recent Alberta Ph.D. who has worked on the robot for 15 years - except for a three-year break to play poker professionally.

The game-tree approach doesn't work in poker because in many situations there is no one best move. There isn't even a best strategy. A top-notch player adapts his play over time, exploiting his opponent's behavior. He bluffs against the timid and proceeds cautiously when players who only raise on the strongest hands are betting the limit. He learns how to vary his own strategy so others can't take advantage of him.

That kind of insight is very hard to program into a computer. You can't just give the machine some rules to follow, because any reasonably competent human player will quickly intuit what the computer is going to do in various situations.

"What makes poker interesting is that there is not a magic recipe," Schaeffer said.

In fact, the simplest poker-playing programs fail because they are just a recipe, a set of rules telling the computer what to do based on the strength of its hand. A savvy opponent can soon gauge what cards the computer is holding based on how aggressively it is betting.

That's how Laak was able to defeat a program called Poker Probot in a contest two years ago in Las Vegas. As the match progressed Laak correctly intuited that the computer was playing a consistently aggressive game, and capitalized on that observation by adapting his own play.

Programmers can eliminate some of that weakness with game theory, a branch of mathematics pioneered by John von Neumann, who also helped develop the hydrogen bomb. In 1950 mathematician John Nash, whose life inspired the movie "A Brilliant Mind," showed that in certain games there is a set of strategies such that every player's return is maximized and no player would benefit from switching to a different strategy.

In the simple game "Rock, Paper, Scissors," for example, the best strategy is to randomly select each of the options an equal proportion of the time. If any player diverted from that strategy by following a pattern or favoring one option over, the others would soon notice and adapt their own play to take advantage of it.

Texas Hold 'em is a little more complicated than "Rock, Paper, Scissors," but Nash's math still applies. With game theory, computers know to vary their play so an opponent has a hard time figuring out whether they are bluffing or employing some other strategy.

But game theory has inherent limits. In Nash equilibrium terms, success doesn't mean winning - it means not losing.

"You basically compute a formula that can at least break even in the long run, no matter what your opponent does," Billings said.

That's about where the best poker programs are today. Though the best game theory-based programs can usually hold their own against world-class human poker players, they aren't good enough to win big consistently.

Squeezing that extra bit of performance out of a computer requires combining the sheer mathematical power of game theory with the ability to observe an opponent's play and adapt to it. Many legendary poker players do that by being experts of human nature. They quickly learn the tics, gestures and other "tells" that reveal exactly what another player is up to.

A computer can't detect those, but it can keep track of how an opponent plays the game. It can observe how often an opponent tries to bluff with a weak hand, and how often she folds. Then the computer can take that information and incorporate it into the calculations that guide its own game.

"The notion of forming some sort of model of what another player is like ... is a really important problem," Nau said.

Computer scientists are only just beginning to incorporate that ability into their programs; days before their contest with Laak and Eslami, the University of Alberta researchers are still trying to tweak their program's adaptive elements. Billings will say only this about what the humans have in store: "They will be guaranteed to be seeing a lot of different styles."

Even so, Laak and Eslami are top-notch players with a deep understanding of poker's mathematical fundamentals. They should be able to keep up with the computer - this time.
 
I do think that tournaments are probably the truest test of this issue rather than cash games for a range of complex reasons I wont bore you with.

Ultimately though I do not think that it will be possible to make this a true test of man v machine in the truest sense in which man would lose.

Take chess for example. It is simplely not true that Computers are better than the best humans at chess. It has only just reached the point where draughts has reached the point where PC's programed rightly are unbeatable. The sample size of Pro v PC games are way to small to read anything clearly from and for the following reasons it is not a fair contest

a) the pc has a huge wealth of opening theory pre saved in its memory banks. this is all heavily "human influenced" so if this info is used it is no longer man v machine. Also the human does not have immediate access to the same library so contest is inherently unfair

b) humans are falable - should they be judged for this or should assesment be based on a mistake free human (seeing which has greater capabilities) this is the weakest of the arguments I suggest

c) much of the choosing of which possible moves to discard is often human (programer) influenced so not fair comparison

d) PC processors are much quicker than a human can focus on a chess board at. Also PC's do not need breaks. What would be fair timelimits for each side?

With poker the situation is even more complex. The combination of incomplete info, PC's basically acting in predictable manners and the element of bluff will always give the human the edge in my view.

A human could mimic a style of play and then deviate from it in a way that a computer would always fall for the trap at least once. It is true that randomness can be incorporated into a PC program but when there are sufficient factors for a decision a PC would generally have to act in a predictable fashion.

Until AI develops into the level of Data from star treck then computer poker programes would have to either act in predictable ways (working on the basis of odds and historic behaviours of oppoenent) there will be an element of the "human" in a programer directing the machine as to how to play the hand. I am sure that PCs programs can be got good enough to challenge ametuers but doubt that humans will be challenged in a fair contest for some time.

Also the whole thing is irrelevant - Poker HU is an intensively competative game in which one mind and ego are pitted against another.

Playing a pc is like racing a car or trying to outstare a fridge it is ultimately meaningless except for software developers, AI developers and those who would prefer to purchase a pc program to play poker for them rather than play the game in its full glory.
 
Update

MAN VS. ROBOT POKER

A draw and a win in first two rounds of Alberta University project to pit poker pros against a robot called Polaris

Attracting hundreds of column inches in the mainstream press around the world this week is the unusual University of Alberta project to pit man against its Polaris poker-playing research robot.

So far the results are not exactly conclusive, with Polaris pulling a draw in the first match Monday against poker pros Phil "Unabomber" Laak and his partner Ali Eslami. Although the robot won 7 small bets, anything less than 25 small bets is considered a draw in the "tournament" rules.

In the second round, it appears that Polaris had a more convincing performance, beating both players despite a run of good cards on Laak's part.

Session 3 begins at noon Alberta time Tuesday.

Polaris is described by its university research team creators as the world's most advanced poker-playing computer program, replete with artificial intelligence (AI) and capable of 'learning' or adapting to the play of its opponents as a game develops.

Put together by an award-winning team of researchers from the University of Alberta, the robot is under extreme test in the project by putting it into contention with two highly experienced and respected poker aces, Laak and Eslami in the first real money series of matches.

The scientists have chosen a high profile time and venue - the annual conference of the Association for the Advancement of Artificial Intelligence (AAAI) - for the test now underway in Vancouver, Canada. In order to motivate the human players, a $50 000 prize is reserved for the overall winner.

The research team hopes to have more success this time around than they did back in 2005 when Laak defeated the state-of-the-art poker robot t that time, Poki-X in a much publicised encounter at the World Series of Poker (see previous InfoPowa report) These hopes are based on improvements in technology which the team claims have enabled them to invest the robot with more human qualities - controversially the ability to bluff, trap, check-raise bluff, big lay-down and to read an opponent's style of play and adjust to it.

"This match is extremely important," University of Alberta computing science professor Jonathan Schaeffer told Associated Press, "...because it's the first time there's going to be a man-machine event where there's going to be a scientific component."

The set-up consists of four 500-hand duplicate matches, with Eslami in one room and Laak in another. In each match the same series of cards will be dealt, with teammates playing the opposite hands in each game. So whatever cards Eslami gets in one hand will be the same the computer gets against Laak, and vice-versa. Community cards will be the same for both, and no communication will be allowed.

At the end of each session, the combined bankroll of the human team will be compared to the combined bankroll of the bot team to determine the winner.

If the human team wins by more than 25 small bets in a session, they'll take $5 000. If it's 25 or less (a statistical tie), the players will get $2 500 per session.

The four separate sessions will be played over two days, allowing both teams to learn more about their opponent and adjust their strategy accordingly.

Keep up to date here:
You do not have permission to view link Log in or register now.
 
Update

MAN VS. MACHINE (Update)

Supremacy still undecided after three sessions

With one more 500 hand session to be played, the issue of whether human poker players are better than bots has yet to be decided.

Thus far, three of the four scheduled sessions have been played, with a technical tie called for the first session played Monday afternoon, Polaris winning the second session Monday and Phil Laak and Ali Eslami scoring a win during the Tuesday afternoon session.

The unusual contest is a University of Alberta project to pit man against its Polaris poker-playing research robot. Polaris is described by its university research team creators as the world's most advanced poker-playing computer program, replete with artificial intelligence (AI) and capable of 'learning' or adapting to the play of its opponents as a game develops.

The scientists have chosen a high profile time and venue - the annual conference of the Association for the Advancement of Artificial Intelligence (AAAI) - for the test now underway in Vancouver, Canada. In order to motivate the human players, a $50 000 prize is reserved for the overall winner.

The set-up consists of four 500-hand duplicate matches of Limit Texas Hold'em with blinds at $10/$20, with Eslami in one room and Laak in another. In each match the same series of cards will be dealt, with teammates playing the opposite hands in each game. So whatever cards Eslami gets in one hand will be the same the computer gets against Laak, and vice-versa. Community cards will be the same for both, and no communication will be allowed.

At the end of each session, the combined bankroll of the human team will be compared to the combined bankroll of the bot team to determine the winner.

If the human team wins by more than 25 small bets in a session, they'll take $5 000. If it's 25 or less (a statistical tie), the players will get $2 500 per session.

The four separate sessions will be played over two days, allowing both teams to learn more about their opponent and adjust their strategy accordingly.

Monday evening's session saw Polaris switch to a super-aggressive style of play, resulting in a win for the robot, but the human poker pros came back with a new strategy of fluctuating conservative and aggressive tactics for the afternoon match Tuesday which gave them a third session win.

Laak is complimentary about Polaris, saying it is a strong opponent and the end result could be a draw, although it is not yet capable of out-skilling humans. "Right now I think Polaris is a phenomenal player," Laak said, adding that in his opinion it is not yet superior to skilled human opponents.

Chess and checkers programmed computers have been able to beat talented human players in these games in the past, but poker represents a more formidable challenge for programmers due to the higher degrees of uncertainty and flexibility inherent in the game.

"These elements make poker an interesting research challenge and are also prevalent in real-world problems for which [Artificial Intelligence] techniques are being sought," a spokesman for the University of Alberta team that created Polaris said.

Keep up to date here:
You do not have permission to view link Log in or register now.
 
This is indicative of a problem the whole "skill games" industry faces. "They" have also just developed a draughts computer that can't be beaten. It's only a matter of time before public software is available that can be programmed to input moves and stats and figure out the best next move for most skill games. Unless a game has a "house edge", an element of chance or a bias towards one player, there will come a time when effectively, no-one can win if they are allowed to play unmonitored.
 
Update

Looks like the humans have won the last game, giving them the series.

Game 1 - draw

Game 2 - Polaris wins

Game 3 - Humans win

Game 4 - Humans win
 
Currently bots are beatable by a good player but I doubt this will always be the case.

I dont think these test should be allowed for the simple fact that the more time and money is invested, the better the software will become. This could be the start of the end for online gaming. Online poker will become another roulette. I am sceptical as to the reasoning behind such tests.

Perfect poker is easy to beat but if a bot is programmed to randomly generate a percentage of imperfections that makes it slightly less predictable I dont see how it can be beaten in the long term by the best of the best human players.

Its a little mind bogging trying to work out the variables but im sure a machine is better equipt to deal with it than any human.

2000 hands is nothing though as poker has so many fluctuations
 

Users who are viewing this thread

Meister Ratings

Back
Top