Libratus Beats Top Pros in NLHE; Scores Milestone Victory for A.I.


Top poker pros were no match for the Carnegie Mellon artificial intelligence unit ‘Libratus’.

Over the past 20 days, four top heads-up No Limit Hold’em players have been battling it out one-by-one against ‘Libratus’ – a Carnegie Mellon University Artificial Intelligence bot. The marathon challenge, which took place at Rivers Casino, in Pittsburgh, came to an end Monday with surprising results.

After 120,000 hands of play, Libratus had $1,766,250 more chips than the four pros, all heads-up specialists: Jason Les, Dong Kim, Jimmy Chou and Daniel McAulay.

Libratus’ developers – Tuomas Sandholm (a professor of computer science) and Noam Brown (a Ph.D. student in computer science) – say that the huge victory was “statistically significant” and “not simply a matter of luck”.

“The best AI’s ability to do strategic reasoning with imperfect information has now surpassed that of the best humans,” Sandholm said.

“This experiment demanded that we assemble some of the world’s best professional poker players who specialize in Heads-up No-Limit Texas Hold’em and that they would play to the best of their abilities throughout the long contest,” Brown said. “These players more than met that description and proved to be a tenacious team of opponents for Libratus, studying and strategizing together throughout the event.”

It wasn’t a complete loss for Les, Kim, Chou and McAulay; the four will split a $200,000 prize for their efforts.

“Usually, you have to lose a lot and pay a lot of money for the experience,” Les said. “Here, at least I’m not losing any money.”

McAuley stated that Libratus was a tougher opponent than he expected, but that “whenever you play a top player at poker, you learn from it.”

The opposite was also true. Before the match began, many were speculative as to how Libratus would improve day to day. It did so the same way players do; by learning and adjusting.

“After play ended each day, a meta-algorithm analyzed what holes the pros had identified and exploited in Libratus’ strategy,” Sandholm said. “It then prioritized the holes and algorithmically patched the top three using the supercomputer each night. This is very different than how learning has been used in the past in poker. Typically researchers develop algorithms that try to exploit the opponent’s weaknesses. In contrast, here the daily improvement is about algorithmically fixing holes in our own strategy.”

The milestone victory has great significance for the future of AI, as Frank Pfenning – head of the Computer Science Department at CMU – explains:

“The computer can’t win at poker if it can’t bluff. Developing an AI that can do that successfully is a tremendous step forward scientifically and has numerous applications. Imagine that your smartphone will someday be able to negotiate the best price on a new car for you. That’s just the beginning.”

Business negotiation, military strategy, cybersecurity and medical treatment planning are just a few of the areas in which Libratus-like AI could be applied in the future.

Sandholm will soon be sharing the secrets of Libratus now that the competition is over, according to