= 27d567bc2d48d9f40e9ccc20ea370b15)
nazgulnarsil comments on Verifying Rationality via RationalPoker.com - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (154)
you may want to warn people that they need to play a large amount of hands for variance to go down to acceptable levels.
You may want to warn people that "a large amount of hands" means in the order of hundred thousand hands and more.
And to be more exact, variance only goes down relative to the expected winnings. The standard deviation of a sample increases as a square root to the number of hands. Whereas the expected winnings increases linearly. In Limit Hold'em, a 1,5BB/100 hands expected winrate just barely covers two standard deviations from the mean over 100,000 hands. Experienced player can perhaps play 4-6 tables simultaneously, which means that he can accumulate approximately 500 hands per hour. So 100,000 hands would take around 200 hours of play.
The real challenge of poker is dealing with the inherent variance of the game. The immense variance is the reason why poker is so profitable, but even the most experienced players are unable to cope with the most extreme swings of negative luck. The brain constantly tries to pattern-match the immediate results and however much you reason that it's just bad luck (when it really is bad luck!) it will make you sick psychologically.
Note that we assumed we know the expected winrate of a given player. However, conditions change, profitability of the games fluctuate, etc, so it's practically impossible to quantify any given player's current profitability. This makes it vastly more difficult to know whether bad past results are because of variance or because of sub-optimal play.
200 hours is 1 month of 50 hour weeks, or 2 months of 25 hour weeks. Is it really that big a deal for your results to only matter month to month rather than day to day? I mean, yeah, it can be frustrating during a bad week, but it's not like the long run takes years.
Depends on what your luck attribute is. I'm well above average, both in life and the poker hands I get dealt.
(Yes, median Less Wrong user, this comment is a joke, mostly. You need to play a large amount of hands for your variance to go down to acceptable levels, especially with no limit poker).
The last time someone told me "Good luck", I replied, "I don't believe in an ontologically fundamental tendency toward positive outcomes."
I've always been fond of the Penn Jillette line, "Luck is statistics taken personally"
I remember a line from the book "Blindspots" by Sorensen that goes something like, "random selection is biased in favor of lucky people".
Statistical luck is definitely real but ontological luck isn't. Most poker players don't know the difference. I am nearly positive that I have had above average luck since I started playing regularly again in December (and I have the data recorded for me to actually calculate it, but doesn't seem worth it to figure it out if my software doesn't do it automatically).
As an example, since you replied to my comment, I rode a <1/1000 wave of luck to 5th place in a poker tournament, for a $540 win.
Wanna bet? ;P
Thereby guaranteeing that this would be the last time that anybody said that to you. (^_^)
Eh, I think I'd do it again.
No chance that http://www.richardwiseman.com/resources/The_Luck_Factor.pdf holds any truth?
So, a while ago I succeeded in phasing "good luck" out of my vocabulary, and I replaced it with "enjoy," which has the great virtue of being only 2 syllables. But reading that has inspired me to seek another replacement.
At the moment, what comes to mind are "make your own luck," "think positively," and "prepare well." All of those are longer, though, and it's not clear they're better. Thoughts / suggestions?
How about just one syllable -- "win". Maybe this should be the standard well-wishing utterance among the Less Wrong Conspiracy.
Between members, perhaps, but most of the people I interact with are not part of the Less Wrong Conspiracy, and it's not clear to me what that would mean to them, whereas something like "choose well" seems less ambiguous.
Every Conspiracy needs a secret handshake.
I've used "have fun" for the past several years. "Choose well" occurred to me within the last week or so, I've been signing my emails with it. Both are two syllables, "choose well" works for rationalists and sounds like what you're looking for.
Oooh, I like that. I'll give it a try.
Be lucky! It sounds very similar to good luck, and is clearly a substitute; it's just a bit more active. It does have three syllables, however.
I guess you could maybe get away with that reply if the correct decision theoretic generalization of anthropic selection (in a sufficiently big universe, ain't gotta be quantum) isn't technically ontologically fundamental... but alas, I'd bet the infinitely reflective meta-contrarian stack returns 'true' for Luck. (Just not Luck for people who aren't you (given some actually coherent definition of 'you', which of course might not look much like it does at the moment).) It's almost as if the Universe likes to laugh at prodigies of refutation, or something. (Maybe we should reify Irony, too?)
Edited to add: This Wiki article on the fallacious side of reification is mildly informative. LW talks a lot about map-territory confusion but it seems as if reification is a particularly dangerous special case. Also, kinda relatedly, I've started to notice how common is synecdoche, which is sort of worrying since in practice synecdoche seems to mostly be an accidental confusion of meta levels...
what
As someone in the midst of a 500 big blind downswing despite mostly getting it in with the best hand, I can confirm this :/. (And that's not even that bad of a downswing compared to what some people suffer).