Poker Moves A Step Closer To Becoming A Solved Game

While many board games such as checkers and tic-tac-toe have been solved, a complicated games such a chess has only been partially solved as the amount of work needed to complete the process would require a supercomputer beyond our current technology. An incomplete information game like poker would be even more difficult to solve, and that is why the poker industry was surprised to learn a group of computer scientists from the University of Alberta in Canada have managed to do just that.
However, before anyone gets too carried away, the computer algorithm known as Cepheus may be able to beat its human opponents at fixed-limit heads-up Texas Hold’em (HULHE), but it is likely to be a long time, if ever, that the same success could be applied to more complicated poker variants involving a greater number of players, and no limit betting.
Nevertheless, the team have become the first people to solve a game of incomplete information, and commenting on the remarkable achievement Cepheus Poker Project manager Michael Bowling explained:
“We define a game to be essentially solved if a lifetime of play is unable to statistically differentiate it from being solved at 95% confidence.. Even if you played 60 million hands of poker for 70 years, 12 hours a day, and never made any mistakes, you still wouldn’t be able to say with statistical confidence you were better than this program”
Always hungry for further data, the University of Alberta has even put its algorithm online for anyone to play against, and for those interested in giving Cepheus a go can do so here. However, it good to know a little something about your opponent before playing them, and in this case Cepheus is apparently willing to play a wider range of hands, and bet more of its weaker holdings than considered prudent by most pros, so be warned.