Comment by metroholografix | Hacker News Reader

Comment by metroholografix | Hacker News Reader

Parent Full thread

metroholografix

The author makes a lot of assumptions and oversimplifications whilst falling for the mystique of poker as expressed in popular culture. We've all seen the movies where psychological tells ("deception") end up saving (or ruining) the day during a poker game.

Unfortunately, in terms of winning strategies, this couldn't be further from how the game actually plays out today. Back when online poker exploded onto the scene, the old cash game pros quickly discovered that their way of playing the game (described in books like super system and theory of poker) was obsolete.

Seeing veteran poker pros getting crushed in their home turf by teenagers whose playing style was forged in the crucible of online multitabling was the ultimate expose to that old mystique of poker as a psychological battle of deception. One could say that the internet players through their mechanistic play were approximating strategies that were at their core, algorithmic. It's not at all surprising then, to see poker bots today utterly crush certain variants of the game. It won't be long, before poker as a whole succumbs.

This is really fascinating to me as the article at least mischaracterizes chess. It is fascinating that it is also mischaracterizing poker and poker is much more robotic than stories have told me.

That chess is less robotic, is perhaps not obvious. A postmodern skepticism of “even in principle” thinking is part of it: who cares what’s possible in principle, when in practice that is inaccessible to me? The actual narrative is me versus this player, my blunders versus her blunders, and whoever blunders less wins. There is a Platonic Form of perfect play that has no blunders? It is discovered by computers? See, this is something I had to acclimate to in learning chess, because chess engines are available, they will rank your games and tell you that this opening is slightly better than that opening by a tenth of a pawn, and you have to like just throw them in the trash. It's like a standard of beauty that demands you have half a percent body fat and rippling abs and bench 300 easy. And instead you have the desire to be Interesting, rather than to be True.

I play some openings because they are silly, you chase my knight in a massive circle around the board and then I am totally undeveloped but you are questionably developed and it is not clear you are doing better. Gone are the days of memorized 20-deep trees of Sicilian openings where I got a careful half-pawn advantage in my theory and then turned out to blunder my queen and erase it all. That was, if you like, a half-life.

I don't disagree. But the premise is not that getting good at making less blunders than your opponent in chess isn't fun or fulfilling. It is that getting good at making less blunders in chess doesn't transfer to real life decision-making as well as getting good at games like poker.

I just think that they have different things to teach us. Chess gives you transferable skills of thinking through problems backwards, thinking about timing and how to create time, thinking about how you might get trapped later.

By contrast, say, backgammon gives you a transferable skill of luck-creation and the strength of being vulnerable and well, I don't know what exactly cube management is but something to the effect of the wisdom about when to cut out early versus when to stick with something to the bitter end. They are different skills.

Baseball teaches a transferable skill of perseverance, of "hey you are going to miss most of the pitches and in fact if you consistently hit a 25% batting average you can easily ride that into a professional career."

Trying to evaluate any of these in terms of its impact on "real life decision-making" is probably doomed to failure, no?

The way you write is reminiscent of an Adam Curtis documentary and I think it's quite beautiful.

Thanks. I was the author of the post. Are you suggesting it is harder to bluff in online poker? Or that somehow taking the game online means bluffing doesn't work as well?

Here is a good post on bots, if you haven't seen it: https://int8.io/counterfactual-regret-minimization-for-poker...

Von Neumann proved that there exists an unexploitable strategy for any zero sum game, where playing that strategy gives you the best possible outcome if your opponent knows your strategy and can play perfectly against it. Poker players refer to this strategy as "Game Theory Optimal". It is not necessarily "optimal" in the sense of making the most money in real life, EG if your opponent always folds to big bets then you should bet big more often than a GTO strategy would.

Over the last 15 years or so high level poker strategy has increasingly focused on approximating GTO play, and people are moving away from any type of exploitative play, whether that's psychological tells or adjusting to weaknesses like an opponent folding too much. So a GTO player actually will bluff quite frequently, but it's due to the math indicating that it's an advantageous spot to do it, not because of anything he's observed about his opponent.

That said, there's still room to adjust to exploit an opponent, and the weaker the opponent is the more profitable it can be to deviate from "correct" play.

> but it's due to the math indicating that it's an advantageous spot to do it, not because of anything he's observed about his opponent.

I wonder about this distinction. This seems it would make sense if all the players played the GTO strategy. Suppose one agent were to play the GTO strategy, but an other player was vulnerable to bluffing. It seems reasonable to "tune" the bluffs to that player.

In an extreme example, suppose every time a certain player would fold 100% of the time when another player played a certain way. It seems it would make sense for that other player to play that way when they wanted that player to fold, even under normal circumstances that play would be suboptimal.

That is absolutely correct. If everyone plays 100% GTO, then no one makes any money except the house. In order to actually profit, you need to deviate from the GTO strategy in order to exploit your opponent's mistakes.

Playing GTO against a poor player will make you money. This isn't the case in every game (rock paper scissor is the canonical example), but it is true in poker.

Perhaps, but much less money than exploiting the poor player's many mistakes.

Correct me if I'm wrong, but isn't GTO only valid for 2-player games - i.e. no Nash equlibria for 3 or more player games exist? If so, then "solving" poker by finding good enough approximation of GTO strategy only works for heads-up poker, which is a variant almost no one plays nowadays. Of course, in multiplayer games GTO is extremely useful where there's only two players left in the pot (you can view the remainder of the hand as a two player game and find GTO strategy for that game), but you still need a strategy for the remaining situations.

It's complicated. The short version is that a Nash equilibrium exists for multiplayer games, so there is a GTO strategy for multiplayer poker, but it is not as useful in theory or practice if pots are frequently multiway post flop.

This article explains it better than I could:

http://blog.gtorangebuilder.com/2014/03/gto-poker-outside-of...

It seems there's a lot of deception around how to play poker, so this is the real meta-game of poker: convincing others to play poker badly!

(cue the often useless, recursive aspects of pop-culture game theory)

Actually the entire point of poker is to make your opponent make a mistake or play poorly. A mistake being if you had perfect information a move you would not make.

Clarification: If you played with cards face up no one would ever make a mistake. You could calculate the odds and the math would be perfect. The goal of poker is to make an opponent make a mathematically incorrect play.

This isn't quite correct although that's how poker was taught in the 90s. The goal of the top players these days is to play as close as possible to optimal poker (https://en.wikipedia.org/wiki/Nash_equilibrium) and not worry about the individual hand their opponent might have in a given moment.

This is only true when playing with others attempting to play optimally. If you only play optimal poker it may be difficult to take advantage of sub optimal players without modifying play.

A simple example is a rock paper scissors tournament. The nash equilibrium says you must always simply pick each with 1/3 chance. But if you play this you are almost gauranteed to finish in the middle of the pack. The winner is going to be someone who can take advantage of other players. There used to be a rps AI tournament years ago (UofAlberta I think), and this is why they wouldn’t take submissions that only played nash equilibrium play.

But you are right, in that usually the first step in getting good at poker is to simply play the strength of your cards based on your position. And this becomes more amd more important the higher the stakes since players tend to be better there.

> This is only true when playing with others attempting to play optimally.

It's also true against someone who is playing an exploitative strategy.

I don't think this is true. It's a better strategy to keep bad players in the game by giving them real advice. To ensure they grow and have fun. But not so much advice that they surpass you. That's their own responsibility to seek out.

When those bad players stop playing badly the games die. It is super boring to play live poker with players that don't make mistakes. I know from experience

If people are just folding and folding maybe try playing shorthanded? Then optimal play is much more aggressive and fun.

Nope, even heads-up is boring live with people that play well because you are wasting more time shuffling than playing.

Even on-line heads-up challenges are played on 2 tables in order to not be boring and that is 500-600+ hands per hour

I don't know if the author meant playing poker online or at a competitive level. I play poker with a group of friends, that were all somewhat skilled before we started playing together, and I certainly flex all of the skills he mentioned.

The most interesting part now is the meta. Once you know someone well it's easier to find their tells. Unless they know that and adjust. The final 4 at the table is really fun every week.

Reply on news.ycombinator.com