Are Poker Bots Getting Smarter?

Date: 2014-09-29
Author: Jason Glatzer

Are you smarter than a machine?  Most experienced poker players feel they can outwit and out-battle any artificial intelligence bot on the virtual felts.  While this may be true in the past, whether this remains true in the present and future remains to be seen.

Carnegie Melon University Professor Toumas Sandholm is out to prove that a bot is indeed better than a man.  Sandholm began working on his poker bot, Tartanian7, in 2005 and with the assistance of four Ph. D. students and one to two million hours of super computer core time, came away with a bot that could beat the game's top professionals in Heads-Up No Limit Hold'em.

Sandholm told CardPlayer Magazine, "It's currently unknown whether the best program, which is ours, is better than the top human professionals.  I'd conjecture that it is, but there has not been a controlled man vs. machine match yet."

The professor hopes to put his bot to the test against the game's best players similar to when Phil Laak and Ali Eslami unsuccessfully challenged the University of Alberta's Polaris limit heads-up bot.

Sandholm hopes any such challenge will include enough hands so one side can't claim they got lucky when winning or losing: "The tricky thing for a match is that you need to play a lot of hands.  I would say at minimum 10,000 hands before you can tell who is better."

Sandholm believes his bot's knowledge or Nash equilibrium is superior and that it should be difficult to exploit its leaks: "Nash equilibrium is unbeatable.  We have an approximation of the Nash equilibrium, so our bot is still open to being beaten, but I think it will be very hard to find its leaks."

Tartanium7 also plays a bit unconventionally, which may throw some players off.  Since the bot can perfectly balance its ranges, the bot can open with more betting amounts than a human player typically does.  Also, the bot frequently will open with a limp on the button, which is a raising spot for most experienced human players.

While it is debatable until proven one way or another whether Tartanium7 is better than the competition, the fact that this is even possibly the case shows how advanced poker bots have become in just a few years.  It used to be that many players wouldn't care if there were poker bots since they were so exploitable.  With bots getting more and more difficult to play against, it is important for online poker operators to have the tools to combat this threat.

While this author believes that poker bots that can be used for playing online poker in real-time are bad, he does feel that poker bots have some benefit in the way of training.  For example, sites such as Advanced Poker Training use artificial intelligence to help train you to become a better poker player.

BECOME A MEMBER of PokerSoftware.com today. You can chat with us and ask questions to our poker software experts in the forums and get access to EXCLUSIVE members-only content. Sign up today!


More articles


 

WriteSubmit Your Comment Let us know what you think about this article!





 

Member Comments



 

Josey's Comment

2014-10-07 01:06:23

How is that it is so impossible to get any informed opinions on these Poker forums without the writer getting a backhander $$$$ pushing some companies poker training software trying to convince players that "this is the software you must have" to win big at the tables I have used advanced poker training & one would never in a real game be able to fold all your "Rags" & only play the good hands, also if your an anywhere reasonable player you can go straight to the advanced levels & win their Bots most of the time, but still be along way off making it in the real world of "on-line world of Poker.Still i still enjoyed the game play, its a bit of good fun without either the craziness of the play money tables or the stress of the real money tables.

Like PokerSoftware on Facebook and Get the Latest Software News Plus Weekly Prizes!

recomented-softwares Recommended Software

Tracking Analysis

SNG Tools

Training