Poker on the Brain: Memory

A few months ago, I contributed a short piece to CardPlayer.com called "Poker on the Brain." I figured that since I have a professional interest in how the brain works, it might be fun to explore how it works when we play poker. In that first piece, I wrote about one of my favorite parts of the brain, the enigmatic prefrontal cortex. Don't worry if you missed it, but if you happen to be browsing the Card Player website soon, you might take a look. In the meantime, I'd like to write a little about memory in poker. This isn't the kind of column that'll be chock full of ways to improve your game, but it's fun sometimes to think about which parts of your brain are doing what while you're trying to win money.

The hippocampus is a brain structure that sits close to the middle of your brain. Like many brain areas, its functions are poorly understood. However, we do know that it has a lot to do with long-term memory. Your memories aren't stored there, but if you suffer damage to your hippocampus and/or nearby structures, you're liable to have a lot of trouble remembering things. That isn't to say you'll forget who you are or where you were born. All the memories you already have will be intact. But, new stuff will have a hard time getting in, and things you haven't completely internalized yet might not make it. If you saw the movie Memento, that's pretty much what it's like.

What's most interesting about patients with hippocampal amnesia isn't what they can't do, but what they can do. If you introduce a patient to the game of tennis, and play with him for an hour a day for a solid month, he might never admit to remembering you, or to having played tennis before. But he'll get better at tennis. If you poke him with a stick every day, he probably won't like you very much, although he might not have any idea why.

Something sticks, but just not the kind of explicit memory that allows most of us to say, "I remember that."

Memory can be a double-edged sword. We know that short-term results and single events, the kinds of "episodic" memories of the kind that amnesiacs can't form, tend to be misleading anyway. Just because you recently folded 9-2 and the flop came 9-9-2 doesn't mean you should play it this time. Perhaps many players would do a little better with a touch of amnesia. And to the extent that getting better at poker depends on slow learning over a long period of time, the hippocampus might not be that helpful anyway.

Of course, that's just learning by practice. For other modes of learning, like reading or achieving insights by analysis, your hippocampus is certainly critical. And it's even useful while you're sitting at the table. Some of those sudden explicit insights are actually valuable. Some tells you pick up subconsciously, over a period of time, and consider them part of your intuition. But others you notice explicitly, suddenly, and you certainly want to be able to retrieve them when the time comes. You may notice that a particular player is capable of folding to a raise on the river, or that another player always bets if checked to on the button. To the extent that you can make appropriate use of these observations, it's probably best not to go in with any part of your brain turned off. But sorting out which are useful may be the hard part. While it may have taken you only a few moments to realize that folding your 9-2 wasn't really a bad idea, there are many other kinds of episodes at the poker table that are both potentially informative and potentially misleading.

As you get better at poker, you learn how to outthink your opponents – how to make better use of the information available to you. So, the better you play, the more useful your hippocampus is liable to be. A novice player might remember that a particular player bluffed in a similar position an hour ago, and have no idea what to do with the information. A more experienced (or more intuitive) player would be in a better position to know whether that means another bluff is more or less likely for that particular player.

Of course, poker presents many more mundane and recurring uses for memory: boardcards in stud, action from previous rounds, finding your car afterward … if your hippocampus weren't working, you'd remember these things for a few minutes, but not much longer. Is that long enough to make it through a hand? It would be a close call. It definitely wouldn't be long enough to carry you through the bad-beat story later.

In my earlier column on the frontal lobe and poker, I argued that as you improve at poker, your frontal lobe ought to be working less and less hard. This pattern likely holds for your hippocampus as well, which also tends to respond well to anything novel. The more you play, the less things will seem to be truly novel. But in order to maintain a high level of play, your hippocampus will always need to do some work. Even if integrating information about tells and a player's past behavior is second nature to you, you won't be able to hold on to those bits of information without it.