NOTICE!
Contrary to misinformation in the media, BluffBot is created by an individual developer.
Join the BluffBot.com Newsletter
Use the form above to join BluffBot.com's Heads Up No-Limit Hold'em newsletter.
Discover the winning strategies behind BluffBot 2.0's success, and stay tuned for future BluffBot related news.
You will know WHEN TO...
- Limp and when to raise preflop
- Make a continuation bet on the flop
- Play draws aggressively
- Fire second barrel bluffs on the turn
- Draw against the odds
- Make blocking bets on the river
And learn HOW TO...
- Mix your play the right way
- Control the size of the pot
- Make "delayed continuation bets"
- Use position to your advantage
- Play against over- or under-bettors
- Use game theory to beat all types of opponents
|
|
World Champion Poker Bot
BluffBot is a World Champion "poker bot" created by Teppo Salonen for the purpose of competing in the first two annual AAAI Computer Poker Competitions in 2006 and 2007.
In the 2006 competition (held at the 21st National Conference on Artificial Intelligence) the first version of BluffBot finished in the 2nd place in limit Hold 'em competition beating such opponents as GS2 by Carnegie Mellon University and Monash BPP by Monash University, losing only to University of Alberta's Hyperborean.
In 2007, the new BluffBot 2.0 went undefeated and took a convincing 1st place victory in the no-limit Hold 'em competition beating all other opponents, including Hyperborean07 by University of Alberta, GS3 by Carnegie Mellon University, as well as bots from Gomel State University, Milano Polytechnic, University of Manitoba and University of Minnesota to name a few.
Does It Really Bluff?
Absolutely! Any effective poker strategy requires a balanced use of bluffing and value betting. Contrary to common beliefs, computer programs utilizing game theory are in fact often better bluffers than most humans.
The AI (artificial intelligence) used in BluffBot 2.0 is a non-adaptive expert system designed using the principles of game theory. This approach makes a good practical use of variety of known expert strategies exploiting weaker opponents while at the same time effectively defending itself against exploitation from adaptive opponents.
Can It Beat Human Players?
Both the 2006 limit BluffBot and 2007 no-limit BluffBot 2.0 can give even experienced poker players a run for their money - at least before they start figuring out certain counter strategies the bots are still vulnerable to some degree.
The real question is, however, when will BluffBot 2.0 (or any other no-limit AI for that matter) start consistently beating the top professionals? Personally, I have no doubt in my mind that it's just a matter of time until bots will be able to beat all human poker players. It's already starting to happen in limit hold 'em in the form of University of Alberta's SparBot/Hyperborean/Polaris bots. And it might not be too long until no-limit bots also evolve to the expert level by any human standards.
Can You Beat BluffBot Yourself?
At the moment the answer is "yes" but mostly because certain exploitable patterns that have been identified in BluffBot's current strategies have not been fixed yet. The future versions of BluffBot, however, will be adjusted so that the currently known counter strategies will no longer work.
Top players may still of course be able to discover new strategies to beat BluffBot, but at least the existence of any simple counter strategies should only be temporary.
In fact, I believe that I am always going to be able to create bots that play better poker than I do, due to the simple fact that knowing how to play and actually doing it at the table are two different things for humans. But for computers they are really one and the same. A computer is able to execute it's strategies flawlessly without going on tilt or making other random mistakes. And without physical tells playing "perfect poker" really comes down to knowing and applying the right math for example in a form of game theory.
Where Can I Get The No Limit BluffBot 2.0?
While BluffBot 2.0 for no-limit will not be made available for download, a free web based application allowing you to play it can be found here: BluffBot 2.0
You can also download the original limit version of BluffBot as a plug-in for Poker Academy Pro 2.5 here:
What's The Future Of BluffBot?
Only one thing is for sure. I will not continue developing the limit version due to lack of personal interest in limit poker as well as all the amazing limit Hold 'em research that's already being done by University of Alberta and Carnegie Mellon University.
Assuming I will have time to continue to work on the no-limit version, AAAI 2008 poker competition is one natural future goal, as well as improving the strategies enough to challenge professional human players.
For any future versions some of the new improvements would be better pseudo-optimal strategies as well as adding adaptive strategies to better exploit weaker opponents.
Creating an AI for multiplayer ring games is also a possible future direction.
Acknowledgments
While BluffBot is (contrary to some misinformation that's been around in the media) created by a single person, I want to give credit and recognition to the people and groups of people who have influenced and contributed to the project.
When creating the first BluffBot version that finished 2nd in the 2006 AAAI limit competition, the only "sparring help" I got was BioTools' Poker Academy Pro software and especially their game theory based SparBot. SparBot was already a world class heads-up limit Hold 'em player and strong enough to compete with top human players.
For no-limit however, equally strong sparring bots did not exist and extensive human testing was done instead. Jay Cordes was BluffBot 2.0's main sparring partner spending countless hours and hundreds of dealt poker hands playing BluffBot 2.0 and providing feedback for further optimization of the strategies.
Valuable last minute tips were also provided by professional poker player and PokerZion.com founder John Anhalt (aka "Freakdaddy") prior to entering the 2007 AAAI competition. He took the time to play BluffBot 2.0 and give his professional advice and suggestions despite being busy preparing for the 2007 World Series of Poker.
And last but definitely not least, I owe huge thanks to both Larry Jacobson and Ilkka Huotari for helping me build the website version of BluffBot 2.0. Their expertise in different web technologies exceeded mine far and away and I couldn't have finished the project without their help.
About The BluffBot Author
Teppo Salonen received his MS in applied mathematics from the University of Helsinki, Finland. He was a professional sports bettor from 2000 until moving to United States and Los Angeles in 2005 where he now works as an analyst. Salonen started researching poker as a hobby in 2005 and enjoys playing in the local small stakes Hold 'em games with his wife and friends.

