I am a bad poker player and an aspiring rationalist. I’m very good for a bad poker player though, and I have quite high aspirations as a rationalist. My game of choice is micro to mid limit large multi-table turbos or even faster tournaments. My favorite tournament is the PokerStars $5+rebuy 2x turbo. I eke out a solid enough $15/hour at a type of game that I find to be decidedly more fun than ring games.
Such games are filled with bad players making bad decisions. This is because in a game that moves so extraordinarily quickly, you are often forced to make bad decisions, and beyond a certain point the game quickly degenerates to the point where the only proper move is all-in or fold. The playable hand range in a turbo or 2x turbo is huge. Often, A9 is a monster but just as often it is nothing.
I consistently find myself internally rewarding myself for making the correct play based on what cards come; not based on what was the highest EV decision I could have made. If I reshove with A9 and my opponent turns over AJ and I won, it’s hard not to congratulate myself instead of reminding myself that I just got lucky and should stop making bad decisions. Similarly, during the early rebuy period when people are shoving with all sorts of hands and taking a coinflip seems like a good decision, I mentally reward myself for making the correct decision when I win, and when I lose I just rebuy with an annoyed grunt at how much the other player sucks at playing poker.
This is the Outcome Bias, which often results in people playing what is known as results based poker. This is the tendency to believe you made the correct decision by playing JJ against QQ, if and only if you win the pot.
What can we do? I work at reminding myself that results based poker is not poker. I try and congratulate myself for making the proper play even when the flop doesn’t go my way. Similarly, I try not to give myself too much mental reward when I win an all-in with the lesser hand. It’s hard to reconcile this with the extreme luck needed to win big turbo tournaments — I’ve never reached the final table in one of these tournaments without winning an all-in showdown with the worst hand.
As I continue to play, I notice myself relinquishing results-oriented thinking more and more. The changes are gradual but with each tournament I play, I’m becoming better able to mentally decouple the things I control from the things I can’t control. I’m able to take more pleasure in making the correct decision in a given situation and let the outcomes do what they may.
Rational Poker 
This post seems to imply that the ‘right’ decision doesn’t depend on the result, but does depend on what cards the other player actually has. Is this intentional? I would have thought the ‘right’ decision depends on the best estimate one can make (based on available information) of what the other player could have, rather than what they actually have?
Forgive if I’m missing something in your comment, but, I think that the correct decision is dependent upon what he actually holds, not what we think he holds.
To take the QQ vs JJ example, if we have JJ and our opponent has QQ and we get all our money into the middle before there is a flop we have 17.343% percent equity of the pot. That’s obviously not good for us. But you might know that already.
Doesn’t the correct play depend on the reality of the situation as opposed to our methodology in deriving it? Getting JJ all in against QQ can often be a very forgivable mistake. But it is still a mistake.
I could be getting a bit mixed up and you might be making a subtler point about frequencies. Since this blog is new, I’m not really sure what everyone’s level of understanding of poker is so I’m trying to avoid poker jargon as shorthand for expressing more advanced concepts. If I’ve missed your point entirely forgive me.
Conchis makes a good point – it would seem that the JJ v QQ scenario is an example of a “mistake” only if there would be good reason for our hero to know the opponent had QQ. If the correct belief distribution has the opponent, say, having < JJ 80% of the time, then the decision to play was a correct one, regardless of what the opponent *happened* to have. To think otherwise would seem to be itself "results based thinking," no?
Of course, it is much harder to determine how correct our play is with regard to what our opponents end up having vs. how correct our play is wrt to what cards end up on the board. In other words, if we just get lucky with the board cards, we *know* we just got lucky, as the probability of any given board is easy to calculate, whereas the odds that our opponent is holding a particular hand is harder to know objectively.