I’m a little disappointed with a July 26 article from the NYTimes, written by John Markoff, titled “In Poker Match Against a Machine, Humans Are Better Bluffers“. According to the article, on July 25th, Phil Laak (aka the Unabomber – see pic above from NYTimes – go here and see a pic of his professional poker-playing girlfriend Jennifer Tilly, aka the Unabombshell) and Ali Eslami beat Polaris, the Texas Hold’em playing computer built by the computer science team at University of Alberta.
So far so good. Then Markoff writes in the article that “After four rounds of 500 hands each, lasting about four hours, the player with the most money is declared the winner.” But he doesn’t say who won more money, Laak or Eslami. Nor does he say how much money was won.
Rather, he says that the first round was a tie, the second was won by Polaris, the third and fourth won by Laak and Eslami. When I first read this, I assumed that Polaris lost because it won fewer of the rounds. Which is it – rounds or money?
My second gripe with Markoff’s article is with the title. It suggests that in this match humans are better at bluffing than Polaris. This may be true, but nowhere in the article does Markoff make this claim. After all, Laak and/or Eslami may be better than Polaris at betting in poker, given that they beat the computer in this match.
But this does not mean that they are better at bluffing. If they are better at bluffing, then it would have been nice to see some reason to believe that this is true.