How not to learn poker

Last night at the MGM Grand there was a 50-ish man who played nearly every hand--80 or 90% easily--and didn't care if there was a raise.

A while into the session, he told a story that pretty much explained it. I will relate it as closely as I can to his own words:

"The first time I played poker, about two years ago, the very first hand I got 7-3 offsuit. I knew from TV shows that this was a bad hand, so I folded. The flop came 3-3-7. I would have made a full house! I said to myself, 'I just learned something important: The hand doesn't really start until the flop. You've only got two cards. You need five before you've got anything.'"

So that's how he played. Predictably, he was clearly a long-term loser. He would frequently end up with hands such as top pair but a weak kicker and lose quite a bit of money, or he'd call down with bottom pair. Of course, once in a while he'd make some sort of monster hand with a weird starting combination that nobody could see coming, but I have zero doubt that those wins are seriously outweighed in his long-term results by all of the smaller losses along the way.

Now, maybe he doesn't care that he's a long-term loser. Maybe he doesn't even keep records and therefore hasn't realized that he is. Or maybe he knows it, chalks it up to bad luck, and keeps playing. Or maybe he knows that he's not very good, but enjoys the game and accepts the losses as his cost of fun--and playing every hand is sure more fun than patiently waiting for the premium ones, all else being equal. I'm certainly not in any hurry to educate him on how to change himself into a long-term winner. He's happy with how things are, and so am I.

But his story caused me to reflect once again on how easy it is to learn the wrong things from playing poker. As many before me have observed, the game messes with our natural method of learning. You often do exactly the right thing and get punished for it, or do the wrong thing and get rewarded for it. The little squirts of dopamine in our brains will tend to engrain the wrong lessons from such experiences.

I remember that once when I was a little boy my mom had left the iron and ironing board out. For some reason, it seemed important to me to know whether the iron was on. Not being sophisticated enough to have figured out any other way of eliciting this vital information, I reached out my right index finger and touched the surface. It was on. Ow! I cried. When mom asked how I burned my finger, I told her I had touched the iron. "Why did you do that?" "To see if it was on." She laughed, which didn't make me feel any better. But I learned from the experience. I never touched the surface of a hot iron again. I figured out that I could put my hand close to it and feel whether it was radiating heat, without getting burned. In this instance, my brain's automatic learning processes did exactly what evolution has programmed them to do.

This educational method doesn't work in poker, at least not very well. You raise with aces, get a couple of callers, bet the flop, bet the turn, bet the river, get raised, call, and find out that somebody caught runner-runner straight on you after having started with 5-9 offsuit from first position. You feel the burn of all the money you lost, and your brain, completely unbidden, starts formulating plans on how to avoid feeling that pain again. This is how people develop the habit of limping with aces; they can't quite bring themselves to just muck them, but they have evolved a strategy of losing less. Others, conversely, put in such massive pre-flop raises that nobody calls them, and they win only the blinds. But both types of players have successfully learned how avoid the pain of a large loss, which has become a more important goal to them than maximizing long-term profit. Their automatic learning circuitry has led them down the wrong path.

Imagine if touching a hot iron sometimes caused a painful burn, but other times--more or less at random--the same action instead triggered some intensely pleasureable, quasi-orgasmic sensation? It would be much harder to learn not to touch it. Poker is kind of like that. It's not too hard to learn to like touching hot irons.

During my first month in Vegas, I was playing at the Golden Nugget when I was dealt J-4 in the big blind, with no raise. The flop was J-J-4. I don't remember whether the pot was large or small, but I won it. That was the first time I had ever flopped a full house. You know how certain scents immediately take your mind to some place and time when that odor signified something wonderful, like homemade bread, or your first girlfriend's perfume? Well, the J-4 is like that for me. It was such a rush, it left such a stamp of pleasure on my hippocampus, that every single time I see J-4 in my hand, I have an involuntary flashback to that flopped full house. No other poker hand triggers a specific situational memory like that for me.

But I don't play it, unless, like that previous situation, I can do so for free. I know it's a dog and causes far more trouble than it's worth. It will lose a lot more money than it will ever win.

So what accounts for the difference between me and the guy I played with last night? Why is he now basically a slave to the bad lesson he learned from his first hand of poker, while I get the visceral reminder of the thrill of flopping a full house but am nevertheless able to resist playing the same hand in an effort to replicate that experience?

I don't know all the anwers, though I think it's a really interesting question. Some of it probably has to do with my having played a lot more hands of poker than he has, and thus seeing more action/consequence pairs. Maybe I'm better at filtering out results that I know are anomalous, and giving them less weight in my later decision-making. Maybe I am better able to put faith in the poker authorities I read, and put less trust in an anecdotal personal experience that runs contrary to the received wisdom.

Whatever it takes to overcome the temptation to learn lessons from single hands of poker, we've got to do it. Those lessons are nearly as likely to be wrong as right, and integrating the experience in the same way that we integrate experiences about the world at large (e.g., it hurts to touch a hot iron) can end up being very, very costly.

But if you run into that guy from the MGM last night, please don't let him in on the secret, OK?