A crushing point in gaming and computing came in the 1990s when the IBM computer Deep Blue was able to defeat the world chess champion Garry Kasparov in a much hailed chess match. It turns out Texas hold’em is going to be the next game to be conquered in the rise of the mighty machines. Hold on to your poker chips, fellow gamblers. Science may have just labeled us all as the perpetual sucker at the table.
The specific version of Texas hold’em computers may have now mastered is known as a two player version of heads-up limit Texas hold’em. What makes this computer advance particularly stunning is its difference from other computer solutions to games like chess. In chess, all of the possible scenarios are known to the computer. It is only a matter of looking deep enough to find the right move to eventually checkmate a human. In Texas hold’em, each player has two cards hidden. This means the computer never has all of the information needed to compute all of the possible scenarios. To solve the problem, computers would require more memory and more power.
This perfect play of Texas hold’em is found in an algorithm made by computer scientists at Carnegie Mellon University in Pittsburgh, Pennsylvania. They took an existing algorithm used for Texas hold’em and modified it slightly. The result is a more powerful algorithm that can handle the uncertainty variables in a game of Texas hold’em.
Of course, that such an algorithm for perfect Texas hold’em play exists might be enough to set internet gamblers on edge. How do you know your opponents aren’t cheating? For now, don’t worry. The algorithm requires 262 terabytes of memory. That’s over 250,000 times the amount of memory on the iPhone 6. No one will be cheating with their phone anytime soon.