Apart from collusion, another common fear that many players have when they play online, is that they might be playing against poker robots (autonomous poker-playing computer programs) rather than humans. The concern is that with modern programming techniques, these robots (controlled either by the cardroom or independently by other players) could possibly play poker at such a high level that they can beat the game consistently. Whereas losing money at poker to another human is unpleasant, most players would regard losing money to a computer as being cheated. There are essentially three issues to address with regard to these robots (’bots’):
•    Is it feasible that computers could play better poker than most humans?
•    Does such software already exist?
•    Are bots necessarily a bad thing?

The Feasibility of Online Poker Bots
Until recently, the most powerful commercially available poker programs were those produced by Wilson Software. Although these programs are quite sophisticated, with extremely good instructional modules, the best human players can defeat these programs consistently. Why? Because poker is not just a game of probability and statistics, it is also a game of reading people. Whereas sophisticated computer programs are excellent at the former, it is in the latter area that they fall down, which is why poker programs have not yet been able to match the success of ‘Jellyfish’ in backgammon and ‘Deep Blue’ in chess. How do you program a computer to sense when other players are on tilt, playing too passively or aggressively, or bluffing too much? That is the key challenge that the programmers face if bots are ever to be able to defeat the world’s best poker players.

Does such Software already exist?
Although it is not entirely inconceivable that online bots are already being used to win consistently against very good players, this would appear to be unlikely at the present time. In the past couple of years there have been several attempts to market bots which play online. Most notably the developers of the Winholdem program have been involved in a long-running cat-and-mouse battle with the major online cardrooms, trying to enable its bot to operate undetected. According to PartyPoker’s general manager Vikrant Bhargava in a recent interview: ‘There are a few commercially available programs which people have tried to use. We make a change and these guys again try to beat the system. With our last update, I believe we have rendered the commercially available bots useless.’ Be warned - the major sites may well confiscate your online funds if you are detected running an online bot.

The University of Alberta poker research team has been working on its ‘Poki’ program for many years. Although it is quite a sophisticated program, the researchers have claimed that ‘Poki plays a reasonably good game of poker, but there remains considerable research to be done to play at a world-class level.’ In fact, ‘Poki’ has now been developed into a commercially available software program, Poker Academy Pro, which functions well both as entertainment and as a learning tool.

Having said that, the potential financial rewards from developing an online bot that is capable of winning consistently at $5/$10 or above are immense, so it is far from impossible that it could happen one day, perhaps in heads-up play, where there is only one opponent to read and poker is reduced to its most basic components.

Are Bots necessarily a Bad Thing?
Many people automatically assume that poker bots are a negative development in the game. However, this is not necessarily the case. At UltimateBet bots were used to fill up the play money tables for many years, so that anyone could practise their play, day or night, sure in the knowledge that there would be opponents to play against. Indeed, Mike Caro has gone on record as saying that, although there are no bots at the site he used to endorse (Planet Poker) he is not necessarily averse to using them to start games, so long as they are clearly identified and guaranteed not to win (or if they were to win, then the money would be returned to the players through promotions). However, if bots were one day able to defeat the best human players regularly, this would naturally be very detrimental to online cardrooms, since no-one would bother playing unless they had access to an equally powerful bot themselves. The online game would eventually be reduced to just computer against computer.

Indeed, in July 2005 the concept of ‘poker machine vs. poker machine’ was put to the test at the first World Poker Robot Championship in Las Vegas, which carried a $100,000 first prize. The winning program was ‘PokerProbot’, designed by Hilton Givens, a car salesman and software developer from Lafayette. PokerProbot defeated five other computers and went on to challenge WPT winner Phil Laak in a heads-up match. Thankfully for mankind, the human emerged victorious after three hour’s play .