Warning: your next poker partner may be a bot

The lucrative world of online poker has been put on notice as researchers edge closer to developing software that can outplay even the most skilful humans.

The research poses a significant threat to the multi-billion dollar online poker industry and players, who down the track risk being fleeced by legions of robot-aided scammers.

Computer programs can already beat most humans at checkers, backgammon, scrabble, bridge and connect four. A decade ago, an IBM super computer famously beat the former World Chess Champion, Garry Kasparov, in a six-game duel.

But the huge amount of money changing hands in online poker games - $US60 billion in 2005, according to some estimates - gives hackers a particular incentive to cheat the system.

In an opinion column for The New York Times, Ian Ayres, an economist and lawyer at Yale, said that, in the very near future, "online poker may become a suckers' game that humans won't have a chance to win".

In July this year, two of the world's best poker players only narrowly beat a computer program at Texas Hold 'Em during the first Man-Machine Poker Championship in Vancouver, Canada.

The program, Polaris, was developed over 16 years by the computer poker research group at the University of Alberta. It incorporates a number of fixed strategies but can also adapt based on moves and mistakes made by the opponent.

"We won, not by a significant amount, and the bots are closing in," one of the world champion players, Phil Laak, told The Guardian after the match.

Other poker-playing computer bots, such as Bluffbot and WinHoldem, can already be downloaded from the internet but their accuracy and effectiveness when put up against the world's best players has been questioned.

Professor Jonathan Schaeffer, who co-founded the Alberta university's poker research group, said the team was already working on programs that could take on multiple players in both limit and no-limit games.

His group had already "solved" checkers earlier this year, developing a computer program that knew the best move to make in any possible position.

"All games lead to a draw, unless one side makes a mistake," Dr Schaeffer said.

Polaris is today only strong at two-player Texas Hold 'Em games that have a limit on the amount players can bet.

"Building a program that plays on par with the top human players in this domain is something that will happen fairly soon, perhaps even next year," he said.

The goal of the poker research group, as outlined on its website, is to "produce a poker program that is stronger than all human players".

But despite the rapid progress, Dr Schaeffer stressed it could be a while before computer programs could seriously challenge the top humans in no-limit games and games with three or more people at a table.

He denied the university's research posed a significant threat to online poker sites, saying the sites would "do everything in their power" to block players from using software programs to play for them.

"All these sites have programs that look for evidence that a computer is playing instead of a human," Dr Schaeffer.

But Dr Ayres said computer bots such as Polaris were "quite-scaleable", and it would be virtually impossible to prohibit computer-assisted playing. He said bots could randomise their moves to make them harder to detect.

Australian laws prohibit companies from offering online casino-style gambling services here, but the laws have done nothing to stop swarms of Australians from betting through sites based overseas.

This year, an employee of one site, AbsolutePoker.com, was caught cheating.

According to reports, he exploited a security breach in the site that allowed him to see other players' cards and reaped hundreds of thousands of dollars.

The site has since patched up the security hole and, like all online poker sites, is in a constant battle with hackers. The main sites have so far been reasonably successful at blocking third-party bot software, however, this will become increasingly difficult as the bots become smarter and more random.

Sydney Morning Herald

News Added: 13 November, 2007

Number of views : 184