Trait Ascription Bias

[shipitfish]

Back in March, I mentioned that I was hoping to write up various cognitive biases and how they relate to poker. I'd first like to cover the trait ascription bias, because I think it drives at the heart of one of the first things poker books teach us.

All introductory poker books tell us to profile players at our table. Generally, there are four categories: tight-weak, tight-aggressive, loose-passive, and loose-aggressive. We are often encouraged to make these assessments quickly.

However, the trait ascription bias indicates that people are more complicated than that. For new players, these four categories are a way to begin to learn how to classify the play of others. But, if you are finding, as you advance, that you still have only six words you use to describe someone's play, you have fallen into the trait ascription bias.

The answer, of course, is right there in the “other half” of the bias' definition. If you monitor your own play carefully, you'll see that you have probably fallen into every single one of those categories from time to time. Of course, your natural tendency is in one specific place. (Schoonmaker's Psychology of Poker helps you fill out nice charts to understand yours and others natural tendencies.) But, you have the ability to move around in tendencies based on your mood, the game conditions, or even how much sleep you've had.

Other people are like this, too. Greg (of River Street) once told me that he doesn't like to take notes on players online, but rather focus on “session reads” to see where the players are at the given moment. There is some serious value in this, because it helps Greg avoid the trait ascription bias by not stereotyping players.

I have done reasonably well avoiding it, but in a different way. My online notes usually say things like this: I've seen this player do X in situation Y. In other words, I keep track of what I've seen, but don't ascribe to it any particular classification. You begin to understand tendencies when you observe similar behavior over long periods of time, but at each given moment, you have to assume that it is not necessarily a predictor of future behavior until it is observable as a pattern.

Another piece of this trait attribution puzzle is that people learn. I mentioned a while back that I recently played W.D. heads up. I played my usual hyper-aggressive heads-up style against him which failed to work. He's learned a lot of poker since the last time we sat down. I fell into the trait ascription bias by assuming that his skill had stayed static, when of course it hadn't. It had been months since we played heads-up, and I had actually been a witness to his improvement, as we talk about poker almost daily! Trait attribution bias can be quite strong if it can cause someone to hold the bias in a situation that they have personally witnessed change.

Comments

[patty-bush.livejournal.com]

I still think that it's a bad idea to not take notes. Some players are bad and predictable and have decided to plop down a crap load of money to play with online. These people are only going to be online for a couple of weeks or so, until they burn through their deposit. You need to learn as much as you can because they'll likely be playing a lot in a short period of time, and will have one set style of play.

Maybe a compromise will be to date your entries. If it's a note you tookd four days ago, it's likely valid, but if you took it four months ago... maybe not. What I do is make notes about what level they're playing. I think that's the most important. If you have a lot of notes at say 5/10 and then you read those same notes at 10/25 they must have gotten better and have moved up, but if you have notes at the same level, they likely haven't changed much...

I don't know, I'm a little drunk right now, so I'm rambling. hey when are you going to be at the WSOP? I want to come watch!

[shipitfish.livejournal.com]

Oh, I definitely agree that you should take notes; I gave the example of Greg as an extreme. I still take notes, and am sure to note as much about the context of when something happened as well as what happened. My notes usually look something like:

When short-stacked in a 1/2 cash game, tends to move in preflop with weak hands of the AJ/88 variety, but seems to be only when it is folded to him or if there are limpers with big stacks. Doesn't reraise much preflop.

I like your idea of dating them, and I'm going to start doing that too.