by Hillary LaClair, Senior Editor
March 2, 2009
The online poker community has voiced concerns over poker bots, a software developed to play in online poker rooms without the human element. The London Times featured a story that examined online poker bots, which are for sale in many venues on the internet. The bots have had a degree of success for some, but are blamed by some poker players that have not won using them.
According to the Sunday Times, bots have the ability to earn some online poker players over GBP 50,000 each year, appealing to an audience that has not yet mastered the game of poker. While nearly every online poker website strictly prohibits the use of online poker bots, and employ software to detect their usage, the article in the Times suggests that online poker rooms have little defense against customers using them.
The Times quotes a post in the Cardschat forum, “I use software that plays poker for me in six different rooms and I’m raking in the dough. Do you think that’s evil or genius? The software does lose, but I’ve never seen any playing like it. Bottom line: not only does it work, but I’m entertained watching it.”
There are other opinions on how well the poker bots function when in a full table. Heads-up seems to be the type of play in which the bots are more likely to succeed. This is because the number of players significantly decreases the ability for the poker bot to figure the odds for each player, whereas in heads-up, the bot needs only compare its odds with one player.
The first poker bot surfaced in 2005, playing professional online poker player Phil “The Unabomber” Laak, who was able to defeat the bot within three hours. There have been other bots developed by the most skilled in its technology at the University of Alberta. The infamous bot called “Polaris” was matched against a team of the most aggressive and successful poker players who were able to win by a narrow margine.
Phil Laak and Tony Esfandiari were able to defeat the machine with a previous loss of over $955. Because of this, Polaris’ creators added a new technology that would further allow to machine to adapt to and learn from the human strategy. IJay Palansky, one of this year’s human competitors said the machine made very unpredictable moves, unlike anything we’ve seen from live players. Matt Hawrilenko, who survived to the final match was afraid of what Polaris’s win would hold for the future of heads-up poker. “For those of us who make our incomes largely from playing heads-up games, whether computer are ready to beat us or not, they’re certainly ready to beat some of our opponents,” he said.
Polaris II was developed the following year and matched against seven poker coaches. This time, it was the bot that celebrated a victory, in a game of 6 heads-up matches. For each match, two players were put up against Polaris II simultaneously, in separate locations. An identical set of cards was dealt to both teammates’ tables, from opposite ends, to ensure that there was more strategy involved than luck. This proves for online poker advocates that the game of poker does require a great degree of skill. When 500 hands were played, the net score was then tallied to determine the winner.
After the first five matches, the tournament was tied for both the human team and the computer. One match was considered a draw as the rules said that a win of less than 25 bets would be a tie. In the end, however, Polaris took the tiebreaker at 90 small bet wins. This was an historic event for the bots makers, as it was the first time that Polaris was able to best a team of poker professionals.
“It’s possible, given enough computer power, for computers to play perfectly, where over a long enough match, the program cannot lose money,” said the leader of the university’s computer poker research group, Michael Bowling, “Humans will always make some mistakes, meaning the program will have an advantage. We’re still quite far from the necessary computing power for perfect play. However, we’ve been able to take what we learned last year and apply it to improving this year’s program.”
And while Polaris is one of the biggest advancements in Artificial Intelligence technologies, one of the chief architects of the bot assured the community that it will not be the end of online poker. “It’s possible, given enough computer power, for computers to play perfectly, where over a long enough match, the program cannot lose money,” said the leader of the university’s computer poker research group, Michael Bowling, “Humans will always make some mistakes, meaning the program will have an advantage. We’re still quite far from the necessary computing power for perfect play. However, we’ve been able to take what we learned last year and apply it to improving this year’s program.”
IT experts have surmised that 2013, a supercomputer will be built that exceeds the computational capabilities of the human brain and that by 2049 a $1,000 computer will exceed the computational capabilities of the entire human species.
Readers of the article in the Times suggest that the quotes in the paper were spam forum posts and as such as not authoritative. “Such software is hardly as popular or reliable as suggested here,” said one reader. “These non-news items on ‘poker bots’ pop up every 3-4 mos. Or so, it seems. Perhaps a bot is writing them!”
The article reports however that human continue to enjoy the success of poker bots, especially users in the UK. “The UK is enjoying an online poker boom, with almost 1 million Britons expected to take part this year. An increasing number of them will be using bots,” the article reads.
|
