Poker is a whore. But I'm sure you all know this by now. I have run absurdly bad against Gus; I should be up several 100k against him, but instead I'm down several, so I'm going to be playing a little lower for a while and hopefully I can reset my luckbox. It's all pretty frustrating though, PLO is just a really annoying game at times. Lately it's been feeling like one step forward, two steps back. But enough of that, I'll talk about some more abstract stuff today. I'll continue my poker story probably in my next blog post.

There are lots of ideas that I have that I tend to mull over in my mind but don't really have discussions with anybody about, since they tend to be pretty abstract or inconsequential and don't really make for good conversation fodder, so I figure that a blog is probably a pretty good place to inform some of these meditations. So the topic of the day is variance. Obviously, everybody knows what variance is. There are two main topics that I think I want to talk about, and the first one has more of a PSA feel to it, so I think I'll talk about it first since it probably is more relevant to the average blog reader.

To put the thesis simply, variance takes a lot of forms, and people don't really tend to acknowledge variance in every form it takes. The word we essentially use for variance in poker is "running bad" (since nobody ever cares to talk about when they're running good; when we're running good we either tell people we're crushing or don't bother to talk to other people about how we're doing). People say that they're running bad in a number of different situations, the most prevalent of which is when they get their money in good, and their opponent ends up sucking out to win the pot. To almost any poker player, this is what it means to run bad. The other situation when people think they run bad is when they get their money into a spot where they're usually making money but they end up e behind. People think they run bad when they get KK vs AA, they think they run bad when they get a set under a set, flush under flush, etc. And most annoyingly, they think they run bad when a stupid program like PokerEV tells them that they're running under expectation. Now, most of these events are reasonable indicators of running bad, quite obviously. But none of these situations are actually synonymous with running bad. Here's a simple fact that for some reason people refuse to understand: you cannot run bad and still be winning. Period. If you think that you're just so good and awesome that you were running bad today but you still won because you're just so good and awesome, you're wrong. You didn't run bad. You may have gotten more KK vs. AA's then you would on a normal day, or maybe you got oversetted twice and still managed to win which has convinced you that you're God's gift to poker, but this whole way of thinking about variance is just deluded and overly simplistic.

For the hell of it, let's try to break down why exactly we think we're running bad when we get KK vs. QQ and lose. I've thought about this a bit, and it seems that the simplest way to analyze this situation in terms of a basic impulse is simply this: "I found myself in a situation in which I am typically rewarded, but that reward was withheld." Every time we as poker players get dealt KK and get all-in preflop, we of course have a very strong expectation of getting rewarded. After all, when we get KK all-in preflop against QQ, we win a buyin more than 4 out of 5 times, so it's a perfectly rational association that we build between getting KK all-in preflop and getting rewarded. In a (gratuitously cynical) sense, we become like Pavlov's dog - the expectation to win the pot is like salivating when Pavlov rings the bell - the sound of the bell being equivalent to getting dealt KK. And so, getting upset about losing a KK vs. QQ is analogous to Pavlov (poker in our case) ringing the bell, but not bringing us the food we were expecting. Now, I don't know if the emotional palette of dogs extends to feeling frustration, but certainly we humans do when our expectations are so betrayed. (And of course, betrayed is an interesting word to use here - maybe our relationship to poker is a lot closer to a dog's relationship to Pavlov than we're ready to admit, but that's a grimmer discussion for another day.)

So essentially, the emotion of having our expectations betrayed makes us feel that we're running bad. As I said before, these events are generally good indicators of running bad so it's not irrational by any means to assume that you're running bad when you lose KK vs. QQ. But running bad doesn't mean anything unless you define it over a window of time. You can't just run bad in a vacuum, you have to run bad over some period of time, whether it be a hand or a session or a month. Now, let's take a look at running bad over one hand - how should we objectively (rather than emotionally) judge what it means to run bad? Well, running bad means simply running under expectation. So in the course of this hand, we have to look at what our standard expectation is - clearly, we got dealt KK and our opponent got dealt QQ, and so we'll win 4/5 of the time. So we're running under expectation when we lose, because in this situation we should usually win. So then we're running bad? Yep, we sure are. Okay, let's move on to the next window. Say we play a session, and we get dealt AA/KK five times vs QQ/JJ, getting it in good every time, and we win three times and lose two times; those were the major hands, everything else essentially broke even. Are we running bad? Well, again, clearly we should be winning 4/5 of the time when we get in an overpair against a lower pair, and we only won 3/5 of the time. Again, we're running under expectation for this situation. Now, just extend this analogy over a month - say you're in the red with KK and AA but still ended up on the month, blah blah blah, you get the picture, running bad. So, in all three of these situations I agree that the conclusion is that we're running bad - but why exactly are these analyses nonetheless incorrect?

Well, the simple answer is that they're setting constants that aren't really constant. Let me explain what I mean. Look back at the hand where we get dealt KK - when we analyzed that hand, we were looking at the equity of KK against QQ. But if we want to isolate how we're running - our expectation alone, then we have to recognize that it's arbitrary how much we decide to fix what's constant in the hand. In the situation we looked at, we decided to fix the ranges after getting all-in preflop, and that's where we analyzed the equity and then decided who was running bad (and this precedent of course is established by using things like Pokerstove and hand calculators and PokerEV and whatnot). But that setup is arbitrary: why don't we freeze the hand after the flop is dealt (say the flop is Q 7 5)? Clearly in all hands where we get all-in preflop with KK vs QQ and the flop comes out Q 7 5, we're usually going to lose, so in that situation we're not running bad at all. That sounds a little absurd, but why is that analysis invalid? Well, you could say, "the flop comes out differently, that's just one of the many possible flops, and since you don't have any control over what flop comes out when you get all-in preflop, you should freeze the hand at that point." Well, that's the only counterargument I think that could get anybody out of the Q 7 5 argument, but that argument actually ends up collapsing in on itself too. Here's why: we have established with this argument that it's okay to unfreeze a hand and move the moment of constancy (I'll use this term to refer to the point at which the hand is frozen) earlier in the hand. The reason that we provided was because we had no control over how the flop came out. But once we start to talk about control in poker, we set ourselves upon a slippery slope. So by this line of reasoning, if we want to move the moment of constancy up a little higher, we just have to isolate a point after which we have no control over the results. Well, let's take a look at when our opponent flips up his hand - he shows us QQ, but we had no control over what hand our opponent ended up showing up there with. If he had AA, then our equity would not have been 80%, nor if he had AK. So clearly we had no control over him having QQ, so we can move the moment of constancy up one point - to the point of getting KK all-in preflop against his handrange (we could of course, if we so desired, recalculate our equity for getting all-in against his range in general). But we can go even farther. We had no control over whether or not our opponent would jam over us after we 4-bet him - he could have been 3-bet bluffing us and just have folded his hand, so since we had no control over the fact that he happened to have a hand worth getting it in with after we 4-bet him preflop, we move the moment of constancy up even higher; now all we have is that we 4-bet him after he 3-bet us, which of course is lower EV than getting it all-in preflop. But really, if we think about, we didn't have any control over the fact that he 3-bet us in the first place - he just happened to have a hand. All we had control over was our initial raise in the first place. So we move the moment of constancy up again - all we were really entitled to was however much money we win when we get dealt KK and raise preflop (the average BB won per KK is like, what, 4 or something?). But then we take the final leap - we had no control over getting dealt KK. In reality, the only thing we had control over was deciding whether or not to sit out preflop. By deciding not to sit out that hand, we certainly weren't entitled to getting KK all-in against QQ - nor were we entitled to getting all-in with KK against his handrange, nor getting 3-bet, nor getting even dealt KK in the first place. So what are we entitled to? Well, simple - whatever our average winning per hand is.

Boom. And at this point, the argument should make perfect sense. Every time you play 5 hands, your EV over 5 hands is your EV per hand * 5 hands. So every time you win a buyin over 5 hands, no matter what the situation was or how bad your opponent is, you are running way over expectation. Well, you might think that this sort of thinking doesn't allow room for being able to play better or worse affecting your expectation - in reality, the games you're playing in, the opponents you're up against, and the quality of your play at the time all have an effect of your EV per hand. And you'd be right in that regard - but you have to recognize the extent to which we have control over these things. It's important to realize that at any moment you're playing, the set of all strategies that you'd use in response to a any situation is already embedded in your brain - in a way, you don't have control over that. That is, you can't suddenly "decide" to use a strategy that you don't know is a good strategy, or "decide" to not make a mistake in a spot where you're already predisposed to make a mistake. So, for example, if you tend to call too many 3-bets with weak hands, in that moment you have no control over this leak of yours; it's a part of your average EV in that moment. Over time you can change these predispositions and make your game slowly stronger as you gain more and more good habits and break bad ones, and your EV per hand will slowly increase over time. But in any moment, the factors over which you exert genuine control as a poker player are actually surprisingly small. I think they are limited to these three things - one, how hard you choose to table select. Now note, this is not choosing the table - it's meaningless to analyze the EV of yourself in a game with 5 superdonks, because you had no control over such a table existing; you only had control over your table selection standards. So you'll have an average EV per hand for a certain threshold of table selection. The second thing you have control over is your tilt / self-awareness. This is quite obvious - no matter what game you're playing in or how much you're up or down, you always have control over whether to take a short break, whether to get somebody to talk to you to get your head back on straight, whether to quit the game and cool off, or even just take a moment to re-analyze the game and yourself and reset your clarity. This is what decides whether you're thinking through every hand or simply auto-piloting through. And then the third thing you have control over is just when you play, and for how long (which overlaps somewhat with the other two). These three factors are the only things that you as a poker player have genuine control over - everything else is out of your control. Who you play, how bad they are, how long the fish stays, how many KK vs. AA's or AA vs. KK's you get, how many times you soulread the fish, and how many times you bluff off a stack to your table nemesis - these things are only in your control as much as these three factors are in your control. Everything else is simply permutations that were already there, they are simply one of the many possibilities that could happen given the set of your intuitions about poker and the way you react to different scenarios. So, to put the point simply, your EV over any session is simply whatever the EV is of:

• You playing with your game selection standards
  
• You playing in whatever state of mind you're in
  
• You playing when and however long you play

And that's it. Thinking about what your EV was when you got all-in with KK vs. QQ, or what your EV was when you played heads up with that huge fish, or what your EV was when you made that monstrous soulread, the simple fact is that none of these analyses actually tell you your true expected value. And certainly your PokerEV graph doesn't. The reality is that your EV is going to be, on average (with the three aforementioned factors presumed constant, which they aren't) whatever your winrate is. That's it. If you're winning 50 cents a hand, and you played 1000 hands, then your EV was on average $500, give or take based on those three factors. So, if you won $1000 over that session but you lost two stacks with AA, you were still running good, because you were simply fortunate to run better in the scenarios you were given other than the two AA stacks. Same thing if you're up 100k over a big sample and your PokerEV graph shows you're supposed to be up 20k more, there's a pretty good chance that you ran good on the whole to have gotten yourself into enough positive scenarios beyond the ones where you got your money in good and lost.

So ultimately, what's the point of this big long rant? Well, the point is - let's simply define running under expectation as winning less than you ought to, and running bad as actually running significantly less than you ought to. Under these definitions, and referencing the fact that of course variance for everyone has grown tremendously in the last year of poker - if you're winning at all, then you're not running bad. Ask one of the many extremely good players who are genuinely running bad, losing over large samples of hands, over months at a time, but who are still much better than you and have higher genuine edges in the games that they play - they are fucking running bad, and let's reserve the word for those people out of respect for how nice it is just to not lose money. Lots of people don't appreciate winning 1/10th of what they consider to be a good month, and allow me to set those people straight by saying that those people are bitches, and if they actually ran bad (or remembered what it was like when they ran bad), they'd appreciate just not being down and would suck it up.

So, all-in-all, if you're not running bad, then shut the fuck up and be grateful. Poker is a whore, but she's like one of those whores in a movie or something who everybody thinks is stupid and doesn't know anything but actually knows all of the secrets of life if you ask her. Even whores have something to teach if you're willing to learn.

Well, there was another topic I wanted to talk about, but this ended up taking me way longer than I expected. So I guess I'll post that later (lol). Procrastination FTW. Next post I'll try to make the continuation of the previous blog post though. So, until next time.