Back from Vacation Link Dump
Speculating about AC’s future.
What games can humans still win? Uhh, poker?
I’ve got to wonder what kind of moron would Tase an adult holding a baby
How to get out of plastic handcuffs. I personally carry a key for the real kind.
Messing with surveillance cameras.
Oddly missing is take cover and return fire.
August 20th, 2007 at 9:38 am
From the “Waht do do in case of a shooting” link…
Ummm, srsly?
August 20th, 2007 at 3:16 pm
Poker: The problem with this whole “Poker is a game of skill” pie-in-the-sky, we-want-poker-legalized-everywhere dreaming is that it is, at it’s core, a game of chance. No matter your skill or the quality of the poker program, one has no reasonable way of predicting the next card. The poker programs can be set up with a really solid ruleset to minimize their risk (which is all poker really is: a risk management game layered on top of random card turns) against the worse possible outcomes. The best long term strategy in poker is always tit-for-tat: Opponent is a tight player. You play tight. Opponent is a bluffer. You bluff. When the flop occurs, it’s going to eventually just be the central tendency theorem, no matter the skill.
August 20th, 2007 at 3:29 pm
Then why do the same 12 or so guys always finish in the big money at the major tourneys and such?
And they will lose. Though at a short stack (i.e., small dollars relative to blind/ante structure) a program could do well.
Poker is the opposite of risk management. It is making and calling bets when odds are favorable (i.e., taking risks).
Err, no. If a player is tight, you loosen up. You bet at them everytime. If they play back, fold. You’ll win more small pots and not lose much when you fold. And you do that ’till they figure out what you’re up to.
There is a theoretical mathematically perfect way to play, of course. But it discounts observing opponents. And it also discounts when I intentionally make theoretical mistakes when the betting is small so as to induce you to make one when the betting big.
You’re invited to my game any time.
August 20th, 2007 at 3:49 pm
Minimize risk, maximize return. If you can calculate the (roughly) expected return on any action you can win easily.
If we flip a coin 1,000 times where I pay you $1 for each heads and you pay me $1 for each tails we will come out close to even. If I find someone willing to pay me $1.50 per flip while expecting only $1 in return I will win. Simple.
The “poker is all luck” crew should try explaining how I’ve managed to average 4-6 (and sometimes more) big blinds per 100 hands played for the last 4 years. And I’m not even that good.
August 20th, 2007 at 5:26 pm
Humans have a vulnerability and a capability computers can’t match: they exhibit “tells” and they can detect “tells” in their opponents. So in a tournament between skilled players who all understand the odds about as well as the human brain can, the ones who are best at concealing their own tells and detecting others tells will win in the long run. Of course, a skilled player can win much faster and bigger by playing against not-so-skilled opponents; if one player bets impulsively, doesn’t know the odds, or has drunk too much, the guy that keeps his head has a huge advantage. (Not that the drunken fool can’t go home rich on a lucky fall of the cards – but if he keeps playing, eventually he’ll lose.)
Computers shouldn’t have tells, but offsetting that is their inability to detect tells. So at first glance, it looks like the issue is solely how well the computer can calculate odds, and thus it ought to beat any human. It’s a trivial programming job to write a poker program that can *exactly* calculate the odds that a particular hand will win, or that the next card will turn your two of a kind into three of a kind, fill that flush, etc. Humans can memorize the odds for various hands, but can’t do the calculations necessary to figure out exactly how turning up a particular card affects the odds of drawing another particular card, etc. If calculating the odds was all there was to it, computers would have been beating humans (on the average) for 40 years.
However, there is one more effect a player can use to estimate the odds and adjust his strategy accordingly: the size of bets and number of cards discarded say something about your opponent’s hand, even if all you see are numbers on a screen. They may be a bluff, but even lies convey information once you come to suspect the lie. So in reality, a good human poker player might use this added information about his opponent’s hand to re-figure the odds of winning and come out closer than a computer – and he can also estimate how his own bets, etc., are going to change his opponent’s thinking and try to use that to manipulate the opponent into a losing strategy. I see no reason that a computer couldn’t do these things too, and better than any human – but I very much doubt that even the poker champions understand well enough how they think about these issues to tell someone how to write the program to do it.
Which means that at present a computer should beat most players but lose the best. To move beyond this point, I think the computer will have to be programmed to learn, and then to play the best players through thousands of hands. It’s all do-able, but who is going to pay for that education? Top poker players don’t waste their time on low-stake games, and the computer will be slowly losing for a long time before it learns – and then once it does start winning, no one will be willing to play the computer for more than peanuts.
August 20th, 2007 at 10:29 pm
“You’re invited to my game any time.”
I’ve got my friend Colin interested in your crackpot ideas about poker. If you are ever up here, he’ll let you sit in on his game on the LES. He’ll even waive the $5000 buy-in. You can’t bring a gun. No need. The guys who stand around “keeping an eye on things” have got your back, or, if you get pissy, they’ve got your ass. I serve the drinks and don’t play anymore. Momma doesn’t like gambling.
Anyhow, until you solve the problem of the complete randomness of the draw, it is a game of chance. Also, if you are taking risks, then you are managing risk. You don’t take a risk you don’t have to do you? If you do, you are just gambling.
August 20th, 2007 at 10:51 pm
I make it a practice to avoid that area.
And does he believe it’s just random? If so, i’d think he’d be playing slots. I doubt it given the stakes he plays. And my crackpot ideas pay very nicely, thanks. Isn’t this the guy you said plays like one hand in 50? And I pointed out that at full ring he has to win some pretty large pots to break even?
Never said otherwise. Just said that there is skill to it, which is why the same 12 or so guys always make the big bucks on various poker tours. And why Laak and the other guy beat the computer. I think a computer program may beat average or even good players. But not expert players (of which I am not).
If you’re unnecessarily taking risks then you are inherently not managing them as managing risk is customarily synonymous with avoiding risk or mitigating it.
Sure you do. Why else would you be playing the game? If you’re playing no limit, your entire stack is at risk.
But isn’t poker a game of chance and random? That’s what you keep telling me.
August 21st, 2007 at 3:19 am
Someone knowledgeable should edit that wiki entry so that it mentions returning fire. Also while they are at it edit the part about taking a martial arts class to learn to disarm someone. I have done drills with airsoft guns it’s possible to pull off a disarm, but I wouldn’t ever want to try.