Wednesday, 29 June 2011

Every Other Sentence Is a Lie

. . . but I won't tell whether the first statement is a truth or a falsehood.

On Tuesday of the second week, Louie and Kevin came by to teach us poker. They began by showing us a video of common errors in probability calculations that people might make during Poker. Then, we split into three tables to learn the rules and play a few rounds. We each bought in for five dollars, because several people said that it is hard to take poker seriously without real stakes.

Poker is a very silly game. It is based on the probabilities and likelihoods of certain hands coming up, with the strength of the hand correlating inversely with its likelihood. Except when the creators of the game didn't actually calculate the probabilities, resulting in some oddities. The "flush" (five cards of the same suit) is worth more than the "straight" (five consecutive numbers), even though the straight is less common. I threw a fit when this was first explained to me, demanding that we (rationalists and all) play them in the correct probability order. Everyone else promptly refused, explaining (very reasonably) that such a system would be useless almost everywhere. (Irrational people, they're ALMOST EVERYWHERE!)

Things got strange right away. Kevin went upstairs and returned with a great deal of alcohol. Arrogant as humans will be, I was all, oh, I'm just going to drink a little bit. It won't affect anything! So we drank, and played, and drank, and played, and very soon things weren't really making much sense at all, and everything was swimming about, and even though I was looking at my cards, I wasn't actually seeing them, so I'd have to look at them again a moment later, and things were generally frustrating. People kept saying things like "you should bet between half and twice the pot," or else things like "I'm sorry you lost that hand, but if it makes you feel any better, that probably was the correct way to play it =D," and everything was all very very confusing. I was being a n00b. I later learned not to spend forever and ever thinking over whether to call each time, because every bet seemed just high enough that I didn't want to take it, but just low enough that it might be worth my hand, and I would sit there wondering and wondering. As time wore on, I cached more of my previous decisions and didn't spend as much time deliberating anymore, or else I just got too drunk to do it. And then things were better.

I often hear from people, "it's like paying five dollars to become better calibrated" or something to that effect. Maybe it's true. But it also can't be that each time someone loses, they become better calibrated. I'm certain that people who have played a great deal are actually updating very little, and that their positive or negative experience of the night actually comes from the randomness in the cards.

The statements are no longer either true or false. Nearly everything at this point is a half-truth, including this statement itself.

In my very very limited experience, poker is a very unpleasant game to play. At a table with n people, one loses an average of (n-1)/n of one's hands, and this is at least half of the time. It is like getting little pings of defeat over and over again. Each time one is dealt a hand, one is hopeful, and most of the time, one is disappointed in one's hand. And therefore, even if one wins in the end, one is likely to spend a majority of the game being disappointed or unhappy. Meanwhile, every amount anyone wins is some amount someone else has lost. It is zero-sum. It is so zero-sum. Money is made and lost, yet no value is generated. The first night, I played for an hour and won a dollar. Yeesh, that's an hourly rate that's really not worthwhile at all. But that's all silly, because the expected hourly profit over a whole table is zero, unless you're a significantly better than average player. With rationalists, I generally assume that the people I play against are at least as skilled as myself.

(I'll take on any one of you on any instrument in Rock Band.)

So the value of the game, what is that? Some answers I've heard:
- thinking when there is money at stake
- estimating probabilities
- calibration of value estimates
- making difficult decisions
- the social aspect of manipulating other players

All good things to learn; all with a nontrivial likehood of being useful or necessary at some point in time. But then again, we would hardly be here if we couldn't do necessary but unpleasant things now, would we?

Since then, I've played somewhere between one and three other times, and some people have played many many more. It's always strange, because I'm always trying to guess at what they're thinking, and they're always trying to guess what I'm thinking, and then we base our actions on what our model of the other person is doing, including basing their action on their model of us acting, and it all spirals into a huge tangled DeathNote sort of mess. They don't trust me, nor I them. I'll admit here to having never yet bluffed a hand, but I'm fully open to bluffing hands in the future.

The element of randomness is scary. Knowing that there is a chance of things working out really well, I'm ever tempted to just play this hand and see what happens, and then I have to be all NO that's not decision with positive expected value! And it's scary, and it's tiring, but I trust that I'm learning really useful things about handling scary situations in the meanwhile, but it's not entertainment. It's like that giant game of diplomacy where we played for many hours and ended in a six-way tie, except that itself was pretty fun. And whenever I think "poker," I get a vague impression of that study done with pigeons, where rewards were given randomly and intermittently. My psychology textbook cited that even when rewards were removed, pigeons would continue to peck at a button hundreds of thousands of times, because "hope springs eternal."

Hope springs eternal. What a lovely thought, and what terrible things it makes people do! People lose vast amounts of time and money for their misplaced hopes. I think I'm just ranting on and on at this point, so it must be time to close off. I still owe you guys a completely positive post, so that will be the next thing after this, I promise. Therefore, another clip of song to sleep upon:
Please come with me,
See what I see.
Touch the stars for time will not flee.

Peace and happiness,


  1. "The "flush" (five cards of the same suit) is worth more than the "straight" (five consecutive numbers), even though the straight is less common."

    How so? I didn't bother to check the math, but wikipedia says flush is quite a bit rarer than straight, i.e. 0.20% vs. 0.39%.

  2. Every other sentence is false, ayneon

  3. Yeah, it seems at least a few of them really are false.

    (When I read the title I was actually thinking about how very often we bend the truth this or that way in our conversations/status games...)