Yesterday I went to a seminar on academic research on poker by John Wooders. While I was more impressed with Wooders’ work than with Steven Levitt’s paper on skill in poker, I continue to be struck by how far the academics are behind the actual poker players. The paper Wooders discussed involves experienced poker players and novices playing a simplified poker game. When I suggested that many skilled online players would have enough familiarity with game theory to find the Nash equilibrium in the experimental set-up, he scoffed that most undergraduate students couldn’t do it. Well…they probably wouldn’t fare very well at online poker then.
Something funny happened during the talk that isn’t especially related to the research, but is instructive for would-be poker players. By way of showing what a hand history is, Wooders showed a history of a hand he himself played online. In a limit hold’em game, it was folded to the small blind, who raised, and Wooders re-raised with QJo from the big blind. He explained with some pride that he had been observing this player for awhile and concluded that his raise from the small blind was a steal, so he went for the re-steal with QJ.
In the context of a blind battle, QJ is quite a strong hand, certainly not a bluff, but that’s more or less rote knowledge, no reason to know it unless you’re a serious player. What’s more interesting and to me more surprising is that Wooders made a conceptual mistake typical of weaker players. He had observed this guy raising a lot in steal positions and correctly concluded that he would be capable of raising with weak hands. But loose players get dealt strong hands just as often as tight players. Presumably this opponent would always raise his strong hands too. He can have a strong hand or a weak hand, there’s not enough information to say for sure. You often see weaker players substitute “I knew he was bluffing” for “I knew he could be bluffing.” Professionals account for uncertainty in their decision-making process and play to maximize against their opponent’s range of possible hands. Amateurs guess their opponent’s hand and act on their guess. I would have thought Wooders, whose job it is to study mixed strategies and uncertainty, would steer clear of this pitfall. I guess it just shows how unintuitive poker can be, even for intelligent people with relevant training.