how much life is left in online poker?

[jeh1978 0]

I know most of the regular posters are online enthusiasts, but I was wondering what future most of you see in it. People began developing chess programs in the late 60's and early 70's and it took tyhen over 25 years to devlop a program superior to most grandmasters. Both computer science and game theory have advanced a great deal since then, and there is a lot more money in online poker than there ever was in chess. How long do you think it will be before someone develops a program capable of crushing the 300-600 games at ub, or the 100-200 games at P-stars. Once that happens, isn't it likely that the programmers of the bot that beats the biggest online games will just start at the top, clean out the biggest games, then the next biggest, and so on until online poker is dead. How many months or years do you people think is left until this happens?

[Smasharoo 0]

How long do you think it will be before someone develops a program capable of crushing the 300-600 games at ub, or the 100-200 games at P-stars.Eternity.Unlike chess, poker's not solvable in a mathmatical sense. People will eventually design bots that can make money at the highest levels, but it's not like if you played perfectly at 100/200 you'd be guaranteed to make money that day, that week, or even nessicarily that month.Do you see why?What is FAR more likely to hapen is that people will design bots that can beat most 1/2 games and online play will die from the bottom up as they become more widely available.Probably at least a decade untill that happens, though.

[GamblinLeaf 0]

 How many months or years do you people think is left until this happens?

Forever. The "biggest game" in the world are the markets (all markets including stocks, commodity futures, currencies) where, literally, trillions of dollars a day are moved around. Many people, including some famous traders, some of whom I know, have spent billions of dollars over the past 25 years or so trying to find a computer program that could consistently extract money from various markets. You name it, they've tried it. They have all, for the most part, failed. The closest thing to a profitable mechanical system are the very simplest systems, the kinds you don't need a computer for, which can have hot streaks but then go cold for years on end. And ALL systems fail to out-earn a savvy, experienced, flexible trader.Poker is the same. There are too many variables - the largest one being the human element - to ever truly develop a one-size-fits all mechanical model that can beat any game. And even if a marginally profitable system is ever developed, the savvy, experienced, flexible player will be able to out-earn that system from now till kingdom come.

[jeh1978 0]

and game theory? The argument that the "human element" could not be overcome was made in the defense of the claim that no computer could beat the best human chess players. In limit poker, there are only 3 possible actions: call, raise or fold. I don't see why a bot could not be programmed to calculate the likelihood of multiple opponents' actions and the optimal response. Human unpredictability will complicate matters, but can be factored into a margin of error. I concede that it is possible that bots will infiltrate poker from the lower limits on up, but I also think that it's possible that someone will bring together the programming skill and poker expertise to create a bot capable of beating the biggest games, and then work their way down. Either way, online poker is dead. Most likely, in a lot less time than ten years. It took 30 years of AI research to develop a computer that could beat the best chess players. The greater financial incentive and more powerful computers should result in highly efective bots in less than a third as much time. Since online poker has been around since the late 90's and people have been programming bots to beat it for nearly as long, I don't think there is much time left.

[rog 0]

Absolutely it can be done. The markets are a much harder problem. There are thousands if not millions of different things that can happen. Everything has an unpredictable effect on the market. Think butterfly effect, and you'll know what I mean. There are virtual infinite decison points, and infinite decisi0ns.Poker has fewer decision points, and fewer choices at each decision point. 4 Betting rounds, capped at 4 bets per round leaves 16 decision points per player. Absolutely a killer bot could be written. A bot to beat most games consistently could probably be written today if you got the right people in a room together for a year. Game theorist, poker expert, AI specialist and a few good programmers.

[JFinn08 0]

and game theory? The argument that the "human element" could not be overcome was made in the defense of the claim that no computer could beat the best human chess players. In limit poker, there are only 3 possible actions: call, raise or fold. I don't see why a bot could not be programmed to calculate the likelihood of multiple opponents' actions and the optimal response. Human unpredictability will complicate matters, but can be factored into a margin of error.  I concede that it is possible that bots will infiltrate poker from the lower limits on up, but I also think that it's possible that someone will bring together the programming skill and poker expertise to create a bot capable of beating the biggest games, and then work their way down. Either way, online poker is dead. Most likely, in a lot less time than ten years. It took 30 years of AI research to develop a computer that could beat the best chess players. The greater financial incentive and more powerful computers should result in highly efective bots in less than a third as much time. Since online poker has been around since the late 90's and people have been programming bots to beat it for nearly as long, I don't think there is much time left.

There are millions of people online right now playing poker. Do you think it is possible for the few people that have the ability to creat these "mega" bots to take over online poker. Im not saying that it is not possible to build such a thing, but I dont think it will be the demise of online poker. Also, remember that the same technology that allows these people to create these bots is available to the poker room operators. There will always be ways to track and prevent this from happening.

[Guest Anonymous]

I know most of the regular posters are online enthusiasts, but I was wondering what future most of you see in it. People began developing chess programs in the late 60's and early 70's and it took tyhen over 25 years to devlop a program superior to most grandmasters. Both computer science and game theory have advanced a great deal since then, and there is a lot more money in online poker than there ever was in chess. How long do you think it will be before someone develops a program capable of crushing the 300-600 games at ub, or the 100-200 games at P-stars. Once that happens, isn't it likely that the programmers of the bot that beats the biggest online games will just start at the top, clean out the biggest games, then the next biggest, and so on until online poker is dead.  How many months or years do you people think is left until this happens?

Despite its complexity, chess is a game of complete information which is why it was eventually solvable. Not all games fit this benchmark. At one end of the scale, you have Othello, which an undergraduate computer science student can write an unbeatable program for in less than an hour. Continuing to use board games as an example, you have Go, the chinese game of stradegy, where even the most complex programs written to date cannot beat an amateur player.Poker falls into the latter category. Sure you could eventually write a bot to kill low stakes limit games, against opponents who are horrendous on average. But it is also ridiculously easy to train an eight year old, or chimpanzee to do the same. In higher stakes games or no limit games, it is conceivable that computer programs will one day be able to rival humans, but we are nowhere near achieving that as of date.

[GamblinLeaf 0]

Absolutely it can be done.  The markets are a much harder problem.  There are thousands if not millions of different things that can happen.  Everything has an unpredictable effect on the market.  Think butterfly effect, and you'll know what I mean.  There are virtual infinite decison points, and infinite decisi0ns.

I am a professional trader. The above statement is wrong. It has been perpetuated by generations of economists (read: people who can't make money trading) who believe in "random walk". The fact is that any market, at any given moment in time, can only do one of three things: Go up, go down, or go sideways (effectively going up and down repeatedly within a confined range).Secondly, with the exception of the truly unforseen event (think 9/11), very few events exist whose effect is unpredictable. We know that such and such company has earnings on such and such date and what the impact of such and such report will have on the stock. We know when the Fed is meeting and what their choices are and how their decision is likely to impact the market (within probabilistic ranges of course, just like poker). All of these things are very similar in very many ways to poker.The fact of the matter is that all +EV/-EV decision-making endeavors, of which trading and poker are two (chess is not, there is generally only one correct move at any given time in chess - there are no pot odds, risk/reward ratios, variance related to drawing with odds, or long term expected value judgements to be made), all come down to the person making the decision. I will grant (as I did in my original post) that someone somewhere might eventually create a bot that can make money playing poker online. But I contend that there will never be a bot that can make MORE money, more consistently, over a longer period of time, than a good, flesh and blood player.

[SOWhatKid 0]

not much more buisness online poker is gonna get from me

[justblaze 0]

and game theory? The argument that the "human element" could not be overcome was made in the defense of the claim that no computer could beat the best human chess players. In limit poker, there are only 3 possible actions: call, raise or fold. I don't see why a bot could not be programmed to calculate the likelihood of multiple opponents' actions and the optimal response. Human unpredictability will complicate matters, but can be factored into a margin of error.  I concede that it is possible that bots will infiltrate poker from the lower limits on up, but I also think that it's possible that someone will bring together the programming skill and poker expertise to create a bot capable of beating the biggest games, and then work their way down. Either way, online poker is dead. Most likely, in a lot less time than ten years. It took 30 years of AI research to develop a computer that could beat the best chess players. The greater financial incentive and more powerful computers should result in highly efective bots in less than a third as much time. Since online poker has been around since the late 90's and people have been programming bots to beat it for nearly as long, I don't think there is much time left.

chess is a game of complete information. poker is not. this whole topic is stupid.

[Socrates 0]

It won't last much longer. Besides bots, all of ther horrible players who complain about how bad online poker is will go away because they are "better" than that. Eventually we'll only have good players playing other goods players and the sites won't make as much money and then go away. Online will die with live poker, sometime in the next 2 years (at least from what we know it as today)

[SOWhatKid 0]

It won't last much longer. Besides bots, all of ther horrible players who complain about how bad online poker is will go away because they are "better" than that. Eventually we'll only have good players playing other goods players and the sites won't make as much money and then go away. Online will die with live poker, sometime in the next 2 years (at least from what we know it as today)

I 2nd that

[GamblinLeaf 0]

It won't last much longer. Besides bots, all of ther horrible players who complain about how bad online poker is will go away because they are "better" than that. Eventually we'll only have good players playing other goods players and the sites won't make as much money and then go away. Online will die with live poker, sometime in the next 2 years (at least from what we know it as today)

I'll take the other side of that, as a friendly bet with a fellow FCP'er. I believe there's a whole generation of young professional - doctors, lawyers, etc. - who have not yet ventured into the world of online poker, who will begin to play with their disposable income in coming years. These same types used to take "flyers" on penny stocks, or play around with corn futures, but after the Nasdaq blowup I've seen some of these types become pretty gunshy. As long as ESPN continues to broadcast tournaments where "anyone can win", I think we'll continue to see more and more new faces at your favorite online site. As a useless prediction, if during peak hours the Party network sees +/- 70k people online, I expect that number to double in the next 3 years. By 2010, I think that number will be pushing 200k.

[bdams19 0]

not much more buisness online poker is gonna get from me

What a tragic loss

[Airabus 0]

It won't last much longer. Besides bots, all of ther horrible players who complain about how bad online poker is will go away because they are "better" than that. Eventually we'll only have good players playing other goods players and the sites won't make as much money and then go away. Online will die with live poker, sometime in the next 2 years (at least from what we know it as today)

I 2nd that

Yeah just like the DVD made Theaters obsolete. Not to offend but I don't think you guys understand what makes a person sit down and gamble. Do you think the millions of people that go to Vegas every year expect to win? No they are going to have fun and hope to get lucky and bring home the cash. Even if they lose they still keep coming back because it is entertainment. Most of us on this board play for money (at least we try too), but most of "them" out there play for fun. Once you understand this you'll realize online poker is here to stay. Companies that make as much money as the online poker rooms are making don't go away quietly.

[budfox427 0]

there are tons of kids, ages 10-17 that are playing the game consitently and are itching to play online the day they turn 18, or 21, or whatever age it is they can get on and start playing. There will be a legion of new, dead money hitting online sites from now until the day the poker craze dies out, if it does. But I think this thing has longevity. Look what Tiger Woods did to golf. Everyone expected that to go away, and the numbers at golf courses worldwide have been higher ever sense he hit the scene. As long as poker has a discernible face to the mass markets, I think online poker will thrive.

[SOWhatKid 0]

not much more buisness online poker is gonna get from me

What a tragic loss

you on my d!ck, kid?

[rog 0]

chess is a game of complete information. poker is not.

So what? Computers dont need complete information any more than humans do. At the higher levels, people play MORE like computers ...randomizing decisions... pushing ever smaller mathematical edges. Where exactly do you think humans have an inherent and permanent advantage over machines in online poker? I'm a professional programmer by trade. I have graduate level courses in AI. I dont think a lot of you understand what computers are capable of given enough computing power and the right software.Rog

[Airabus 0]

chess is a game of complete information. poker is not.

So what? Computers dont need complete information any more than humans do. At the higher levels, people play MORE like computers ...randomizing decisions... pushing ever smaller mathematical edges. Where exactly do you think humans have an inherent and permanent advantage over machines in online poker? I'm a professional programmer by trade. I have graduate level courses in AI. I dont think a lot of you understand what computers are capable of given enough computing power and the right software.Rog

This comment alone pretty much invalidates everything else you just said. To act like incomplete information is a non-issue is simple ridiculous.I give you a "So what?".

I'm a professional programmer by trade.

So what? So am I, but I bet there are non-professional coders out there that are just as competent. I'm not saying it can't be done (I'll never say never), but it is not something that can be done easily. In my opinion (You know as a professional programmer... :wink: ) it won't be done anytime soon.

[alf13 0]

When Casinos and Card Rooms open up in every city...Online poker WILL die very quickly.

[justblaze 0]

chess is a game of complete information. poker is not.

So what? Computers dont need complete information any more than humans do. At the higher levels, people play MORE like computers ...randomizing decisions... pushing ever smaller mathematical edges. Where exactly do you think humans have an inherent and permanent advantage over machines in online poker? I'm a professional programmer by trade. I have graduate level courses in AI. I dont think a lot of you understand what computers are capable of given enough computing power and the right software.Rog

OK, i dont think you really understand whats involved in solving games. Chess is a solvable game - computers can play chess perfectly (or very very near perfectly). connect four is a solvable game. YOU CANT SOLVE POKER. Computers will never be able to play poker perfectly, because the concept of perfect poker does not exist. This is due to the fact that poker is a game of incomplete information. Therefore, while computers may one day be able to play poker at a world class level, they will never be unbeatable, like they are in connect four.

[rog 0]

Absolutely it can be done.  The markets are a much harder problem.  There are thousands if not millions of different things that can happen.  Everything has an unpredictable effect on the market.  Think butterfly effect, and you'll know what I mean.  There are virtual infinite decison points, and infinite decisi0ns.

I am a professional trader. The above statement is wrong. It has been perpetuated by generations of economists (read: people who can't make money trading) who believe in "random walk". The fact is that any market, at any given moment in time, can only do one of three things: Go up, go down, or go sideways (effectively going up and down repeatedly within a confined range).

When I talk about thousands if not millions of things, I'm talking about causal events, but even if I were talking about the market effects, there's more than just "up, down, sideways". Some sectors go up...some go down. Some currencies go up...some down. Some companies go up...some down. Unless the only thing anyone is interested in are the indices, there's more than 3 outcomes. The makeup of an overall trend is important too.There are too many things which have an impact on the markets, and the exact nature of their impact is not known. If it were, you'd be out of a job. The markets would be completely useless. If it's a solved problem, then EVERYONE knows the future of the markets down to the last penny. Since the markets are zero-sum, if you're right then nobody can make money off the markets anymore because nobody has an advantage. Economists, mathematicians and physicists are working for all the world's major financial institutions right now trying to model global markets right now, and you're telling me the problem is already solved?So, what are the possible outcomes of the arms talks with North Korea? What what will be the exact effect of each of those outcomes on the world's major markets? Interest rates are such a small piece of the overall picture.

[krup24 0]

OK, i dont think you really understand whats involved in solving games. Chess is a solvable game - computers can play chess perfectly (or very very near perfectly). connect four is a solvable game. YOU CANT SOLVE POKER. Computers will never be able to play poker perfectly, because the concept of perfect poker does not exist. This is due to the fact that poker is a game of incomplete information. Therefore, while computers may one day be able to play poker at a world class level, they will never be unbeatable, like they are in connect four.

You are so philosophical when ur high........ Connect four was a bomb azz game

[BetItAll33 0]

chess is a game of complete information. poker is not.

So what? Computers dont need complete information any more than humans do. At the higher levels, people play MORE like computers ...randomizing decisions... pushing ever smaller mathematical edges. Where exactly do you think humans have an inherent and permanent advantage over machines in online poker? I'm a professional programmer by trade. I have graduate level courses in AI. I dont think a lot of you understand what computers are capable of given enough computing power and the right software.Rog

OK, i dont think you really understand whats involved in solving games. Chess is a solvable game - computers can play chess perfectly (or very very near perfectly). connect four is a solvable game. YOU CANT SOLVE POKER. Computers will never be able to play poker perfectly, because the concept of perfect poker does not exist. This is due to the fact that poker is a game of incomplete information. Therefore, while computers may one day be able to play poker at a world class level, they will never be unbeatable, like they are in connect four.

You don't have to "solve" poker, you just have to beat the game. Rog isn't stating that he can "solve" the game.Imagine playing a limit game against 9 bots where none of them ever make a mistake with regards to pot odds and continuing draws. How do you think that would affect your win rate?

[PrtyPSux 0]

from Paul Phillip's blog.

World Series of Poker Robots: I have said many times (search rgp archives) that bots will inevitably and very soon be better than humans. They will be widespread online and completely impossible to spot without taking measures that will annoy the crap out of humans, like popping up occasional windows with multiple-choice questions that a computer couldn't answer. Better get used to it. As for this supposed bot contest, I know nothing of it. At the moment the "public" bot world is primarily composed of scam artists like this guy. The bots you'd want to use are not available to you. Yet.