Unsettling Bot News

[Kevroc]

A friend of mine returning from the WSOP told me that during one of the breaks he sat down with one of the pros who stated that online poker's time is short. He said that because of people that devolop "bots" to play for them, playing optimum strategy and busting out all the humans...

You've all heard this before, now i'm no professional but, I do enjoy playing online and have played over a year on my initial deposit and would like to keep it that way. Can some of the 2+2 crew give me some encouraging words please?!

[BreakEvenPlayer]

All of the posters on 2+2 are bots that are programmed into making everyone believe that online poker is safe. They also try and promote playing styles which are clearly unprofitable.

If I was you I would cash everything out right now, you obviously know little about the bots.

[Kevroc]

Okay, haha very funny, I knew I'd hear some of that from you guys but, I do obviously know little about the bots. I read a post along time ago saying that it can't happen. Its just hearing it again from my friend that bothered me. /images/graemlins/frown.gif

I take the sarcasm as a good thing however and if anyone else has something to add, I would appreciate it!

[Peter Harris]

What is a bot?

A program that plays to an optimal strategy.

What is an optimal strategy?
An impossible utopia.

There is no best playing style for a table, and a bot would have to be very well programmed to adjust to it.

After all, as MANY people here have said before, if i sat down with a list of hard-and-fast rules on what hands i play in what position, how to bet and when with what hands etc. etc., then i am essentially a bot.

There are, at times, 30000 players on Party. If there are 300 individually well programmed bots playing now in the world i would be flabbergasted.

I do not consider them a major factor. Nor would i spot them any further than i spot a tight-aggressive player.

Too much hype and not enough thought has gone in to the bot debate.

Pete H.

[sthief09]

a bot would have to be very very very complex to handle the nuances of the game. you could program a bot to beat a soft game, but once hand reading player-dependent decisions become important, as in tougher games, it would be very difficult for a bot to handle these important decisions.

the routine, robotic, plays, like raising AA, are ones that any idiot can handle. I suppose these bots could hurt everyone's winrate, but it would be hard for a program to perform better than a good fundamental player who has the ability to make decisions based on other players.

ironically, there was a thread a while back from someone who had one of the bots (I'd name the company but I'd hate to give them free advertising), and for some reason it wouldn't raise AA preflop. so it seems like anything decent is probably a little while away.

[jdl22]

Bot - mediocre at best
pattern mapper - somewhat helpful

bot + pattern mapper - unbeatable poker player and endless supply of free money

[Kevroc]

Thanks for the input guys, I'm just gonna continue to play and not worry about it. People always say in Party that when they type "nh" and the opponent doesnt reply... it's a bot. I always laughed at those comments. I'm not going to let it affect my play, I guess I just found a game that I really enjoy and don't want anything to ruin it.

[Kevin]

I have over 10,000 players in my PT database. Of those, I have only 77 Watch Out Players - with optimal preflop and post flop play. I flag someone after 100 hands (at 50 they are auto exported without a flag), but in the end, if all 77 are bots - which I know that at least most aren't because I have seen chat from some/most of them - it is such a small percentage of the total that it is not enough to worry about. A lot of tad loose/tad passive, a lot of weak tights, a ton of rocks, a few lags, and a few fish mixed in, but plenty of money to be made.


P.S. - from your prior message, just because there is no reply to NH doesn't mean a lot. A lot of players play summary without chat because they don't want to "listen" to the table coaches. I have only recently turned on summary because I think that there is more info to be gained from the chat than sanity saved by keeping it off.

Cheers,
Kevin

[tech]

I hardly ever reply to chat. A lot of times I just turn it off altogether.

[cardcounter0]

I would say that being a poker pro at the WSOP gives one absolutely no knowledge of computer programming or progress in AI software development.

[Kevroc]

Hey Kev, am I in your PT database? I hope i'm one of the guys in the "+" column /images/graemlins/grin.gif


I changed computers and never reloaded PT, lazy bum that i am. I miss that program telling me that A7s UTG is costing me dearly....

[CORed]

[ QUOTE ]
People always say in Party that when they type "nh" and the opponent doesnt reply... it's a bot.

[/ QUOTE ]

More likely they don't reply because they're playing 4 tables. Not too long ago, someone on party typed, "Say something so I know you're not a bot".

I responded, "I'm not programmed to chat".

[CORed]

I'm a poker player and a programmer, and I believe that writing a really good poker bot is a formidable, but perhaps not impossible task. For starters, consider how many situations come up at a poker table. You have to code for every one of them. I don't think it would be too tough to write a bot to play ABC poker to beat soft low limit games, and with multiple accounts playing multiple tables 24/7, you could probably make significant money playing. .50/1.00 or 1/2. If you want to be able to beat decent players, or even not-so-good players who have some experience and pay attention to their opponents, it gets a lot more difficult. You would, at a minimum, need to use some game theory to mix up strategies and do some random bluffing and semi-bluffing. If your bot plays the same hand the same way every time, any decent player is likely to figure out how to beat it, even if they don't know it's a bot.

[Kevroc]

to cardcounter.... that was my first thought when told this. I figured the pro just wants the online players to go to the B&M games where HE has the advantage.

[Ulysses]

My general thought on this is that yes, people have written and will write some very good bots. Even a barely profitable bot (which is not hard to write) is very bad for the future of the online game. However, the online sites all know this, so I'm sure they are going to be investing VERY heavily to counter any automated playing programs. This will be an interesting battle and will go on for quite some time IMO.

[Kevin]

Here's what I have (138 hands)

Weak tight and passive on every street decent starters but plays some out of posiiton xxxxvol pf xxx preflop raise. Only goes to showdown 15pct so move him off his hands - especially the non-raises on the flop

[Kevroc]

Wow Kev, whatta weak/tight loser! Hopefully the P&L is okay! P.S. I hope thats a long time ago....

[gunboat]

[ QUOTE ]
My general thought on this is that yes, people have written and will write some very good bots. Even a barely profitable bot (which is not hard to write) is very bad for the future of the online game. However, the online sites all know this, so I'm sure they are going to be investing VERY heavily to counter any automated playing programs. This will be an interesting battle and will go on for quite some time IMO.

[/ QUOTE ]

Actually, it should not be difficult for the sites to make it virtually impossible for 'bots to play without a human to do the clicking. Simply varying the screen layout, buttons, etc once in awhile would do it, tho human players might get annoyed. I am sure there are other methods. Some web sites require the user to interpret a screen picture to prove they are human.

I would be more afraid of a 'bot/human combination. I believe this has much more potential than turning loose a 'bot.

[Kevin]

No offense meant, I just copy/pasted what I had. It was March and you were -15.43 BB/100. You were probably just running into a cold deck and playing accordingly.

Cheers.
Kevin

[stupidsucker]

I think that bots are out there, and some are probably decent programs at the lower levels. As someone mentioned multiple accounts can most likely make a decent amount of money too.

I however dont think that that bots are very damaging, and I dont think that the internet sites will do much about them unless they show some real damage.

Think about this....
Most of the bots are most likely not opperating above the 2-4 and maybe 3-6 level at best. They dont pose much of a threat against any good player, the bad players will be losing their money to someone anyways. The program cant be much better then the programmer.

If the poker sites wanted to eliminate bots they could do so in one small patch. They wont.... why? Because bots are good for them... they can play more raked hands/day then anyone.

If you have a suspision that someone is a bot... why complain? get poker tracker... study the bot and rob him blind.

I just dont forsee this being a problem for a long time.

Eventually someone will make a good bot, and put it on the market. When this happens it could get ugly for a while. At this point online sites will be forced to do somethign about it... until then they simply arent an issue. I dont forsee this happening any time soon... Any one smart enough to make a bot that works is probably smart enough to know they would make more money using it then selling it.

[Webster]

You guys are all so niave. Technology is so slow to develop isn't it. Snails pace.

Are Bots something to worry about near term? NO Are they something to be concerned about - WITHOUT A DOUBT.

Ever here of AI? spftware that learns as it goes? I mean REAL AI. Not code that someone puts into a smart piece of software.

It's out there already, not in any form that Joe Public understands or knows about but I would think in the next 20 years it will be something the poker rooms will have to seriously look into.

HOWEVER - before THOSE type of BOTS take over pokerrooms there will be changes that you and me can't even imagine.

Technology is really on the brink of some truely outstanding breakthroughs, right now it's just making things we have, better. There is stuff out there that will bring on a new era, like computers did. Perhaps free energy, that would change a few things.

Worry about the things we can change and not what we can't.

[CrackerZack]

All your base are belong to us.

[AviD]

Someone set up us the bomb!
...
You have no chance to survive make your time!

[elitegimp]

for great justice, take off every zig.

[Drexel]

I am more concerned with collusion than bots.

[MicroBob]

your friend talking with a guy at the WSOP who has an opinion on bots and the demise of online-poker hardly qualifies as news.
it qualifies as your firend talking to some guy who thinking online poker is doomed.


if i have a friend who returned from Wash DC who talked with someone 'in the government' who says that Saddam Hussein is going to put anthrax in everyone's chocolate-chip cookies would that be news??


(yummmm....cookies....i'm getting hungry!!!)


some poker-players think bots will ruin the game...others don't.
just like some WSOP participants think bots will ruin the game...others don't.

[kem]

[ QUOTE ]
What is a bot?

A program that plays to an optimal strategy.

[/ QUOTE ]

No, a bot is a program that automatically plays *A* strategy. The strategy could actually be quite horrible, but the program playing it is still a "bot".

[ QUOTE ]
What is an optimal strategy?
An impossible utopia.

[/ QUOTE ]

This is true, but then how do chess programs win with no optimal strategy? Perhaps you're confusing "optimal" with "superior".. Who says that a bot can't play superior strategy and beat 99.99% of poker players out there?

[ QUOTE ]
There is no best playing style for a table, and a bot would have to be very well programmed to adjust to it.

[/ QUOTE ]

And why couldn't someone program a bot to adjust gameplay based on table conditions?

[ QUOTE ]
I do not consider them a major factor. Nor would i spot them any further than i spot a tight-aggressive player.

Too much hype and not enough thought has gone in to the bot debate.

[/ QUOTE ]

I don't know if anyone thinks it's a problem now, but I think it is certainly something to think about for the future. There is nothing stopping it from happening other than a bunch of people hand-waving and saying something about it being too hard a problem.

[stupidsucker]

Im telling you.

Its not going to be a problem unless some idiot decides to market one that plays well.. Even if this does happen, then the various poker sites will take some easy measures to prevent it.

Bring on the bots I say.. If I spot one, I will do what I can to exploit it.

[twang]

h4x0rz!!!!!!

[AviD]

bs haxx0r, FUITF^2

[bdk3clash]

[ QUOTE ]
There is nothing stopping it from happening other than a bunch of people hand-waving and saying something about it being too hard a problem.

[/ QUOTE ]

This is not true. As of a few months ago, if you were running PartyPoker's software and had a browser window with the WinHoldem site loaded, the Party software would tell you to shut the site down. Ditto for running WinHoldem software while playing on Party.

I'm not sure what the situation is now, but the sites themselves can make it pretty hard to run a bot concurrently with their client software.

[M.B.E.]

[ QUOTE ]
Most of the bots are most likely not opperating above the 2-4 and maybe 3-6 level at best. They dont pose much of a threat against any good player, the bad players will be losing their money to someone anyways.

[/ QUOTE ]
A winning bot does hurt a good player because it takes up a seat that could be occupied by a poor-playing human.

If you're a decent player playing 3-6 or whatever, you want the bad players to lose their money to you, not to some bot. It's irrelevant whether you play better than the bot -- the bot cuts into your profit regardless.

[ QUOTE ]
The program cant be much better then the programmer.

[/ QUOTE ]
Yes it can. For example, Deep Blue played chess much better than its programmers could.

[ QUOTE ]
If the poker sites wanted to eliminate bots they could do so in one small patch.

[/ QUOTE ]
No, it isn't that simple. Most software-based methods to thwart bots will create some inconvenience for human players.

[ QUOTE ]
They wont.... why? Because bots are good for them... they can play more raked hands/day then anyone.

[/ QUOTE ]
This is debatable -- I'd love to see a response to this by a representative of one of the big four.

[ QUOTE ]
If you have a suspision that someone is a bot... why complain? get poker tracker... study the bot and rob him blind.

[/ QUOTE ]
If the bot is playing a reasonable strategy in a 10-handed 3-6 game, you just won't get enough opportunities to "rob him [sic] blind". It won't be worth the time you spend studying his play in PokerTracker.

[ScubaDuff]

[ QUOTE ]
Worry about the things we can change and not what we can't.

[/ QUOTE ]

AGREED.

[D.H.]

[ QUOTE ]
This is not true. As of a few months ago, if you were running PartyPoker's software and had a browser window with the WinHoldem site loaded, the Party software would tell you to shut the site down. Ditto for running WinHoldem software while playing on Party.

[/ QUOTE ]

This was a very simple fix by Party. They just checked the window titles of all windows on the screen. If the title contained "WinHoldem" they popped up that warning message. WinHoldem quickly took care of this by making the window title of the program customizable.


[ QUOTE ]
I'm not sure what the situation is now, but the sites themselves can make it pretty hard to run a bot concurrently with their client software.

[/ QUOTE ]

Well, there are other methods they can use, but there are also other methods that the bots can use to avoid being detected.

[kurtcobain]

If you're really worried about bots, play no limit.

[stupidsucker]

Since you took the time to pick apart what I had to say. I will respond.

I do think you or anyone worried about bots has a valid complaint, but I just dont see it being a huge problem in the near future, and I certainly dont see it "ruining online poker for good"

[ QUOTE ]
A winning bot does hurt a good player because it takes up a seat that could be occupied by a poor-playing human.

[/ QUOTE ]

I see this as only semi valid because it does not pretain to BOTS only. If a good human player sits in that seat it will have the same if not worse effect on you. There are lots and lots of tables to sit at. If you feel you are not at a good table because of bot or human player then a smart poker gets up and finds a good one.

[ QUOTE ]
Yes it can. For example, Deep Blue played chess much better than its programmers could.

[/ QUOTE ]

I see your point here, and I agree that the bot itself would have several advantages over its programmer. Calculating perfect pot odds, being able to access strategy within nanoseconds that would be difficult to look up or memorize, but easy to program in. Also the obvious ones like not needing a break. An example of this would be the recent "Lucy Bonanza" I am willing to bet several of those winners were bots. I was able to get in the top 60 both times I tried, but I had to play 4 tables nonstop for 9 hours. Bots have an overwhelming edge here.


[ QUOTE ]
No, it isn't that simple. Most software-based methods to thwart bots will create some inconvenience for human players.

[/ QUOTE ]

I can only think of one way to get the information for a bot, and thats the use of screen scrapes. These rely on exact pixel location. They could easily switch up the cards color or pixel placement. It has been mentioned many times in other posts. What human is going to freak out because the cards are 14 clicks to the right now instead of where they were.. They could randomize this a little , or change the size of the cards. I dont know a whole lot about it, so maybe I am wrong here, but it seems to me they could at least create a pain in the ass for bot programmers this way, and this would eliminate a large sale of bots.

[ QUOTE ]
This is debatable -- I'd love to see a response to this by a representative of one of the big four

[/ QUOTE ]

Do you really think that they would admit it if it were true? The only actions that I have ever seen any site do is against winholdem(which was a sucky bot anyways)The reason they did take action is because it was a marketed bot. The big four and everyone else would prefer not to touch the subject at all. Right now, I think it benefits them more then it hinders them. They wont worry about it, until their customers show enough concern to stop playing. Instead the online realm of poker continues to grow, but like you said its debatable. I can say with 100% certainty that if asked, they will say they strictly prohibit bots, and bots will not be tolerated. Actions speak louder then words(or typed text). I could care less what they say, I am only intersted in what they have done or do. They have done next to nothing.

[ QUOTE ]
If the bot is playing a reasonable strategy in a 10-handed 3-6 game, you just won't get enough opportunities to "rob him [sic] blind". It won't be worth the time you spend studying his play in PokerTracker

[/ QUOTE ]

I agree 100%. If a bot is playing good strategy with no bugs then its better to just avoid him like a good player, or leave the table. However, I have come across 2 people now suspected of bots. One guy would time out every hand, but he still posted every blind, and sat back in for every hand. It looked like a bot gone wrong. It pissed enough people off that they left the table. a few of us stayed and took turns blinding him to death, laughing all the while.
I do agree that its not that easy to take advantage of even a poor bot anymore then a regular poor player, especially if several people at the table are trying to do the same thing.

I dont want to come off as coy, or anything. I do like debating, even if I dont totaly agree with the subject matter. If I had English skills , then I may come off a little better. I dont :/

All I am really saying is... The world is not crumbling. Bots are there, and some are probably doing very well. I just dont see it destroying online poker as we know it. I play mostly SnGs and multis, so I am not at much risk. When Bots become pesky enough of a problem then the multi milliondollar industry will do something about it.

Also... in order for bots to survive... so must online poker.

[M.B.E.]

[ QUOTE ]
I do think you or anyone worried about bots has a valid complaint, but I just dont see it being a huge problem in the near future, and I certainly dont see it "ruining online poker for good"

[/ QUOTE ]
Yes, I agree with you on this, depending on how we define "near future". Bots are not going to have a huge effect on online poker within one year, but within three or four years? Maybe. In other threads I've seen people comment that when a good backgammon bot became available, online backgammon was ruined. I don't know about that, I don't play backgammon, and maybe there are different considerations relating to poker.

[ QUOTE ]
[ QUOTE ]
A winning bot does hurt a good player because it takes up a seat that could be occupied by a poor-playing human.

[/ QUOTE ]

I see this as only semi valid because it does not pretain to BOTS only. If a good human player sits in that seat it will have the same if not worse effect on you. There are lots and lots of tables to sit at. If you feel you are not at a good table because of bot or human player then a smart poker gets up and finds a good one.

[/ QUOTE ]
Okay, fair enough, except that a player with a decent bot can run it 15 hours a day on each of several sites. Maybe even find a way to have several accounts on one site. So if 200 people design a decent bot in the next year, that's going to have a much greater detrimental effect than if 2000 decent players start playing online in the next year.

[ QUOTE ]
[ QUOTE ]
No, it isn't that simple. Most software-based methods to thwart bots will create some inconvenience for human players.

[/ QUOTE ]

I can only think of one way to get the information for a bot, and thats the use of screen scrapes. These rely on exact pixel location. They could easily switch up the cards color or pixel placement. It has been mentioned many times in other posts. What human is going to freak out because the cards are 14 clicks to the right now instead of where they were.. They could randomize this a little , or change the size of the cards. I dont know a whole lot about it, so maybe I am wrong here, but it seems to me they could at least create a pain in the ass for bot programmers this way, and this would eliminate a large sale of bots.

[/ QUOTE ]
Well, changing up the colours of the cards on every hand would certainly annoy a lot of the human players, and I don't know that it would thwart the screenscrapers. Moving the cards around to different positions in the window might not be annoying to humans, but it would make the software look unprofessional. Also I imagine that the bot programmers would find a way to defeat this countermeasure.

[ QUOTE ]
[ QUOTE ]
This is debatable -- I'd love to see a response to this by a representative of one of the big four

[/ QUOTE ]

Do you really think that they would admit it if it were true?

[/ QUOTE ]
No, but I'm curious to see whether they would expressly deny it, and if so what their explanation is.



[ QUOTE ]
The only actions that I have ever seen any site do is against winholdem(which was a sucky bot anyways)The reason they did take action is because it was a marketed bot.

[/ QUOTE ]
Just because you haven't seen other countermeasures doesn't mean they don't exist. In particular, I suspect that PokerStars and probably other sites have some bot-detection algorithms that they use in the same way as their collusion-detection algorithms. But I don't really know.

[ QUOTE ]
If a bot is playing good strategy with no bugs then its better to just avoid him like a good player, or leave the table.

[/ QUOTE ]
But if you suspect someone of being a bot (with a reasonable strategy), why not email support and ask them to investigate? It can't hurt.


[ QUOTE ]
I dont want to come off as coy, or anything. I do like debating, even if I dont totaly agree with the subject matter. If I had English skills , then I may come off a little better. I dont :/

[/ QUOTE ]
You're advancing your arguments well. I am enjoying this little debate.

One other thing, you said that you are not much at risk since you play mostly tournaments. I'm not sure that's the case: in particular, I believe that designing a bot to beat low-stakes NLHE SNG tournaments would not be much more difficult than designing a bot to beat low-stakes LHE ring games.

[stupidsucker]

[ QUOTE ]
You're advancing your arguments well. I am enjoying this little debate.

[/ QUOTE ]

Let me start off with saying that I am glad to hear a friendly debate can happen without anyone getting upset. I love this forum and I am still new, so I dont wish to step on anyones toes, nor Do I wish to make any cyber forum enemies. I am also enjoying this.
I hope others are as well.

[ QUOTE ]
Yes, I agree with you on this, depending on how we define "near future". Bots are not going to have a huge effect on online poker within one year, but within three or four years? Maybe. In other threads I've seen people comment that when a good backgammon bot became available, online backgammon was ruined. I don't know about that, I don't play backgammon, and maybe there are different considerations relating to poker.

[/ QUOTE ]

I dont know much about backgammon myself, so most of what I have to say is assumption, and hearsay. I am to understand what ruined online backgammon was a few programs "snowy" that was created and marketed, it might have even been shareware. Once a mass produced program was available to the public, it did indeed hurt the realm of online backgammon considerably. I have already said that when a good bot is available to the public then it will indeed have a large impact. At this time poker sites will be forced to take counter measures to stop the bots of this nature.

Most of my argument is based on the fact that this is less likely to happen, because of the EV plan of selling a program and constantly having to re-patch it(for free) to keep up with poker sites is far less then just using the good bot for yourself. This is perhaps the biggest hole in my argument. I just have a little faith that the good peeps at the "big four" have a lot more to lose then any one individual player. I trust that it is worth it to invest a million to counterpart this just to keep the millions of $$$ comming in. Hackers and such will always be an issue with anything online. Where this is $$ there is a way.


[ QUOTE ]
Okay, fair enough, except that a player with a decent bot can run it 15 hours a day on each of several sites. Maybe even find a way to have several accounts on one site. So if 200 people design a decent bot in the next year, that's going to have a much greater detrimental effect than if 2000 decent players start playing online in the next year.

[/ QUOTE ]

I see your point here, and I agree that this may already be an issue, but when it gets even close to the point that it may start costing the onlie sites money, they will investigate players that play that much. How easy do you think it is for any online poker site to detect if someone is playing 15 hours a day? If/when this becomes an obvious issue, then support will crack down on this. As easy as it may be to program a bot to click a button, its not so easy to simulate good ol human interaction.


[ QUOTE ]
Well, changing up the colours of the cards on every hand would certainly annoy a lot of the human players, and I don't know that it would thwart the screenscrapers. Moving the cards around to different positions in the window might not be annoying to humans, but it would make the software look unprofessional. Also I imagine that the bot programmers would find a way to defeat this countermeasure.

[/ QUOTE ]

color change is a small price to pay compared to the crumbling of internet poker. No matter what measures are taken both sides will fight each other and constant changes will be made. Banking on the fact that it will take a massive marketed bot to impliment these changes... Who will give up first? Internet poker or the programmer that sold the bots? Ill bet if worst comes to worse, a settlement might solve the problem(or open up a whole new can of worms and make the problem worse)


[ QUOTE ]
Just because you haven't seen other countermeasures doesn't mean they don't exist. In particular, I suspect that PokerStars and probably other sites have some bot-detection algorithms that they use in the same way as their collusion-detection algorithms. But I don't really know.

[/ QUOTE ]

You may be right, but with as bad as online support normally is, I doubt they care as much as they say they do until they see an issue. I still think that the sites welcome the bots. If I were an owner and I felt that the public was worried, then I would at least impliment something that was obvious that I cared and was trying... Even if it was just a showy thing. It makes total sense for them to keep the detection a secret because if the bot programmers know what they are looking for then it would be easier to get away with it.


[ QUOTE ]
But if you suspect someone of being a bot (with a reasonable strategy), why not email support and ask them to investigate? It can't hurt.

[/ QUOTE ]

I will deffenatly give you this one. Just because I dont think its going to be a huge problem is no reason to be complacent about the issue. However, if you constantly email support about this about players here and there, after a while your credibility will fail, so you must be careful about who you report. How many emails a day do you think they get about cheaters, hackers , bots and foul play on their part? Lee Jones(manager of poker stars) commented about this in a recent Card Player article. He also chuckled about it on the phone with me once. When I signed up for the 100-200 game on stars they have to personally call you to confirm some things first. I had the privillage of talking to him for a bit. (I lost my ass in the game, and moved back to 30-60... then moved out of stars period to hit the softer party games)

[ QUOTE ]
One other thing, you said that you are not much at risk since you play mostly tournaments. I'm not sure that's the case: in particular, I believe that designing a bot to beat low-stakes NLHE SNG tournaments would not be much more difficult than designing a bot to beat low-stakes LHE ring games.

[/ QUOTE ]

Maybe I dunno.. I just think that NL tourneys have a bigger chaos effect, and thus limit ring games would be much easier to program for. Certainly not impossible. I practically play like a robot when I play a 20+2 SnG.

I think I have said most of what can argue. We could probably go round for round with it. The truth is.. No one knows the future. All good things come to an end, and online poker might hit its end too. I do think bots will impact online poker, but I just dont see a total crumble. Not now or even 10 years from now.

Thanks for the fun. If you can think of anything else to debate about, post it.. I can take any side. I dont have to believe in it. I dont like the idea of bots either. The game loses some flavor without the human aspect.

[pzhon]

I'd like to address the history of bots in backgammon since third-hand reports seem to be mentioned in this thread.

There were very weak backgammon bots in the early '80s. The best backgammon bot jumped from the level of a casual player to the level of a top-flight human expert in about 1990 with TD-Gammon, the first neural-net based backgammon program to learn from self play. Instead of imitating human experts, this program figured out how to play by itself. It disagreed with the consensus of human experts in many ways, and in many cases including some of the plays of the opening rolls, TD-Gammon was right and the human experts were wrong.

TD-Gammon was not commercially available except much later on a rare operating system, but later bots were stronger and widely available. They changed the game radically. Human players were able to get objective numerical feedback and had a top-flight sparring partner at any hour of the night. Decisions that were once chalked up as matters of style were recognized as significant. Because of the bots, humans now play much better than they did in the past, and it is much easier to learn to play well.

Bots have not killed money play. However, they have greatly decreased the profitability of propositions, positions that would be played out repeatedly. People used to play the same prop for weeks, but now people are inclined to ask a bot. People trust the bot's opinion or to let the bot play the position out a few thousand times, and they rarely will play the wrong side of a prop according to a bot rollout.

Bots have not killed online money play. There was hardly any online money play before a few years ago. There may be people who have hooked up bots to make every move for them, but that is rare, and easily detected. A more serious concern is that people may consult a bot on a tough decision. This is considered cheating in backgammon (live and online) just as it is in chess and checkers where computer programs are better than at least 99.9% of human players.

I think poker will follow the trajectory of backgammon. As far as I know, there are no expert poker bots. I expect that will change, and then humans will improve by learning from the bots. Many playing styles will be recognized as clearly wrong, and rare plays may become more accepted. Later, using the bots to figure out what to do will be considered cheating because people will be afraid to play against a bot for money.

[M.B.E.]

Thanks for the info, pzhon.

[mat]

isnt turbo texas holdem just like playing against a bot that game is very beatable i think for most decent players even on a tough line up

[D.H.]

[ QUOTE ]
isnt turbo texas holdem just like playing against a bot that game is very beatable i think for most decent players even on a tough line up

[/ QUOTE ]

Sure, that is a kind of bot. But saying that one bot is beatable doesn't really say anything about bots generally. Just 'cause you can beat John Doe who is a "poker player" doesn't mean that you can beat Gus Hansen who also is a "poker player".

[jpym84]

As far as I know, the University of Alberta has the most advanced poker bot out there (see http://www.cs.ualberta.ca/~games/poker/ and http://games.cs.ualberta.ca/webgames/poker/ ). The bot is surprisingly good, but definitely beatable, so I don't think theres anything to worry about just yet.

[tadams]

Programs that play poker well have been researched for some time now and there has been a lot of progress on them.

If you want to see how well a program can play poker go to:
http://www.cs.ualberta.ca/~games/poker/

However, it would be unethical to deploy such programs at a poker site without disclosing the fact that they are programs. In fact, no online poker site would knowingly allow such programs to reside on their site undisclosed. Why? It is simply not good business. If it was found out they would lose all of their human customers. However, sites may allow disclosed bots, but they would most likely be controlled by the site and not playing on real money tables. They would be used as advertising to get people to come play at their site. Online casinos already make plenty of money from the rake and there is no incentive for them to allow undisclosed bots. Therefore, you do not have to worry about online poker being taken over by bots.

[jpym84]

Lets just be glad that Darse Billings is too busy researching Ro Sham Bo (http://exn.ca/video/?video=exn20031028-rps.asx) to improve on that bot. Otherwise the online poker industry could be in some real trouble.

[Kevroc]

True, I must confess that I chose the subject title "Unsettling Bot NEWS" knowing full well that this wasn't news. It was just something I heard and was a bit bothered by it. I went with that subject title to get people to read it and respond...

It worked and I thank all of the posters for their input, I am optimistic that the games will stay good and remain honest. Also, when we say bots it doesnt hafta be a guy setting one up at home and having it play for him. It could be the websites themselves doing this.

Another side note... a player at Pacific Poker told me that the hands are rigged to induce action (such as whenever AA is out KK is or flush flops always pair the board on a later street so the flushes lose to boats etc.) This is probably just some guy that got drawn out on once to often but, hey whaddaya think about it??

Thanks!
KEVROC /images/graemlins/laugh.gif

[BigBaitsim (milo)]

Now I gotta ask, you got notes on BigBaitsim or BigBeitzim?

[M.B.E.]

[ QUOTE ]
Lets just be glad that Darse Billings is too busy researching Ro Sham Bo (http://exn.ca/video/?video=exn20031028-rps.asx) to improve on that bot. Otherwise the online poker industry could be in some real trouble.

[/ QUOTE ]
Does Darse Billings play online? I've always wondered whether Spy328 on Stars is Darse (or someone else in the UofA group).

[BusterStacks]

Eh, ok some things I'd like to reply to. First of all, to any doubters, a bot is most certainly possile and very lucrative. The only real downside is the amount of integration and calculation required to make the bot a winning player. To the guy who said it would only be possible with "screen scrapes", um, you are retarded, turn on full chat and post a window of your game. You will see that the cards that are delt are in easy to understand text format, as well as all moves by all players. I mean, it's cool that you don't know about programming, but seriously, let's not give too much credit to the makers of poker software. It's horrible and absolutely buggy. I'm no expert in visual basic, but if I were to make a simple bot, here are the FACTS of what it could do:

a) run in real-time and undetected

b) incorporate specific fields from external programs (pokertracker) and utilize the data as decision making influence.

c) Keep a small database to interpret table texture and playing styles as well as recent trends.

d) make an educated move after considering hundreds of times more factors than any human could.

That being said, I think an actual well-programmed poker bot would beat just about anyone at any game. If you think you can trick a bot, you are fooling yourself. You may win incidental victories, but let's think about this for a second: Poker is not an abstract art, it is first and formost a game of technical strategy. The conclusions you draw about a player or playing style are based on static events that could easily be calculated by a machine. At each intersection requiring input, you have an amazingly small number of options: 3 (fold, call, raise). Now given that a computer can remember every single move it has ever seen you make just as well as the previous hand, do you actually think you could hope to beat it? I don't think so. A computer would would walk all over unless you changed gears literally every hand, at absolute random, at which point you would be playing incorrect poker in an attempt to throw the bot off your scent. I know we would all like to think that a person has some intrinsic ability to do something that a computer cannot, but my friends, it ain't poker.

[Richard Berg]

I agree with BusterStacks. Some games like Go have incredibly expansive search trees even after applying the best-known pruning strategies (which typically derive only from expert contributors' gut feelings). Poker is the complete opposite, even NL. With such a limited state space, its primary heuristics lie in backhistory. In other words, establishing the basics is trivial, and past that point the player with the most information wins. That's not a task on which you want to wager against a computer.

[brianmarc]

Nope. winholdem already has a fix for the identification problem (view/hide tille bar)

[lint]

I don't think they block holdemkiller. And it's possibly a bigger factor because it's free.

[Ed Miller]

It is very difficult to create a bot that plays "optimum strategy." Unfortunately, it doesn't have to play optimally to create havoc in the low limits.

The problem with bots is not that they play better than the best people can (they don't, and they won't for a very long time)... the problem is that they can play tirelessly in countless games simultaneously. Even if a bot can pull only $5 per table per hour out of a $3-$6 game, one person with that software can play countless $3-$6 games simultaneously, 24/7, and pull thousands per day out of the games.

If enough people run this racket, the suckers will go broke far faster than they do now, and they will stop playing. Only the better players (those that can beat the bots) will be able to play for any sustained period of time.

Furthermore, I have NO DOUBT that the bots are coming. They may be mostly rumor and speculation now, but eventually they won't be. And yes, they will ruin low limit poker. And there's not a damn thing anyone can do about it, as far as I can tell.

[Randy Burgess]

I had read in that NYT Science Section article on poker & AI that poki in particular was very strong ... still beatable by the best humans, though.

Maybe the online poker sites will develop algorithms to detect bot-enhanced play and disqualify it, ha ha.

But if bots do come to infest online play, I'm not entirely persuaded that's a bad thing. Not to get overly philosophical, but the world is going to hell in a handbasket partly because we have too damn much virtual reality and are prone to overook other planes. At any rate I can't get all worked up about it, even though I'd be one of those affected.

[J_V]

The sites will be able to stop the bots. However, it will greatly inconvenience people. They could have a random pop up word come up and the players would have to type it to prevent being logged off or something.

[MicroBob]

[ QUOTE ]
Another side note... a player at Pacific Poker told me that the hands are rigged to induce action (such as whenever AA is out KK is or flush flops always pair the board on a later street so the flushes lose to boats etc.) This is probably just some guy that got drawn out on once to often but, hey whaddaya think about it??

[/ QUOTE ]

do you just believe anything that anyone happens to tell you??


i didn't see this response the first time around.

this question is so silly that i wonder about its legitimacy.

however, i suspect it is real and that the poster really is asking for opinions on what some guy at Pacific (of all sites) told him.

FYI - it is popularly rumoured that the sites set up 'action flops' to produce bigger pots and bring home more rake.
somebody comes on these forums once a week or so claiming this is OBVIOUSLY happening.


it is also commonly believed by most on this site that these people are pathetic losers making excuses for their own poker inadequacies. those that make these claims typically use adecdotal evidence about how they saw 'way more high-card flops then they ever see in a B&M poker-room'. etc etc.


in short - some people just aren't terribly smart and refuse to identify themselves as the reason why they keep losing at poker.


also - if pacific really wanted to increase their take....why don't they just take steps to speed up their tables AND/OR allow multi-table play on their site?


the overall premise of action-flops is just goofy.


[ QUOTE ]
True, I must confess that I chose the subject title "Unsettling Bot NEWS" knowing full well that this wasn't news.

[/ QUOTE ]


i should have also stated that some guy who went to the WSOP telling you about these evil-bots not only doesn't qualify as 'news' but also doesn't qualify as 'unsettling'.


i am more 'unsettled' by the number of people who actually think there are rigged games, action-flops, and killer-bots on all of their tables.

i also find it 'unsettling' that one would think some 'pro' at the WSOP or some guy on pacific talking about action-flops would be deserving of any credibilitiy at all.

[Kevroc]

Thanks Bob, these were the type of comments I was hoping to get. A touch hostile but, the ones I wanted anyways. There's no such thing as a stupid question, we approach these forums for reassurance as much as we do for answers to questions about difficult poker situations. You got my Irish up for a second there but, ultimately I walk away feeling better about the whole thing.

Also, ya gotta trust that I am no fool and do not take the so-called pro's comments and the Pacific guy's comments to heart. This thread provoked alot of conversation and thats what it was intended to do. I say bring on the bots! We can take em!

I can see it now...

Party 2/4 9 handed UTG is LAG, MP and MP2 are BOT.. /images/graemlins/tongue.gif

[adanthar]

[the following is greatly simplified programming-wise, not least because I don't know *how to* do any of this myself]

Most multiplayer online games have all had similar problems with bots and 'duping' (duplicating virtual items/gold) and have come up with solutions (not perfect solutions, but pretty good ones) after some growing pains. All of these are applicable to poker.

The thing about bots is that they, much like any other similar application, need access to the memory space where the card graphics are shown on the screen. Encryption won't work for various reasons (the client must show a card on the screen, and if you can see that, a bot can be written to see it as well), *but*, in theory, a poker client application can always be written to see any other application attempting to access that memory space.

Here's the kicker: every time a poker site requires users to download a new patch, every user of every bot is essentially risking that the update contains a new, superior bot detection feature (which would make Party alone millions, since they don't have to pay any accounts they freeze for bot usage.) The eventual result will be something along the lines of the DirectTV hacker cases- ten thousand people with smart cards that got zapped remotely, then sued for $7500 apiece. Except that, instead of being sued, all these people already have a ton of money on Party's own virtual servers. Oops.

So, basically, bots will proliferate for a while, until a patch wipes 95% of the account holders out within a couple of hours. Have fun suing a corporation operating out of a Canadian Indian reservation, folks; I hear the jurisdiction issues are a bitch.

[soah]

[ QUOTE ]
The sites will be able to stop the bots. However, it will greatly inconvenience people. They could have a random pop up word come up and the players would have to type it to prevent being logged off or something.

[/ QUOTE ]

This would accomplish nothing. You can download software for free that is capable of cracking letters/numbers in image format. I played an online game (free) which had that security feature, and I more frequenty heard a PLAYER complaining of being unable to read the image than I heard of a program having trouble -- and I had plenty of connections with the "cheaters" of the game. =) Successfully getting away with cheating in that game was a constant evolution to avoid the latest countermeasures. Unfortunately (for us, at least!), the threat of account deletion was constant. Unfortunately for the general public, innoncent bystanders were frequently hit as well, especially as the cheating improved and became harder to detect.

As was mentioned in a pot above, freezing accounts of bots that are detected would be a deterrent, but I think the sites will always be one step behind, and mistakenly labeling a few innoncent players as cheaters could generate a lot of bad PR.

[adanthar]

Being one step behind is not a problem with a situation like this. You only need to tie once to break every bot player on your site logged in that day within an hour. Moreover, the site has as much monetary incentive to take effective countermeasures as the bot user does to write (or use) a bot; they save themselves thousands for every account they close.

Mistakes aren't really an issue once you get into countermeasures involving more than just 'detecting the word 'winholdem' on the computer screen.' Once the program starts actively detecting memory sniffers, or looks for specific files and time stamps on a hard drive coupled with particular betting patterns, things get a lot more interesting. And like I said, they only need to catch up *once*.

[astroglide]

he's referring to the kind of letters on the bottom of this (http://registernet.passport.com/reg.srf?id=2&lc=1033) page. you're aware of tools that are capable of interpreting them?

[Cry Me A River]

FYI: There's currently a very similar thread over in the Psychology forum on bots:

http://forumserver.twoplustwo.com/showflat.php?Cat=&Number=777069

[nykenny]

[ QUOTE ]
Can some of the 2+2 crew give me some encouraging words please?!

[/ QUOTE ]

us 2+2ers beat the bots day in / day out /images/graemlins/grin.gif no worries.

[astroglide]

if the bots can beat the non-2+2ers, there are plenty of worries

[largos]

I think that bots are a greater threat to limit than no limit. So start playing NL and you `ll be OK.

[BarronVangorToth]

[ QUOTE ]
All your base are belong to us.

[/ QUOTE ]

I can't even tell you how many times I saw the above on a T-shirt this past weekend at Origins... Hysterical.

Barron Vangor Toth
www.BarronVangorToth.com (http://www.BarronVangorToth.com)
"The Contest Is Over"

[Oman]

Bot, Collusion, whatever.

If your at a table and someone is kicking the crap out of you online, move tables.

Problem solved!

[Mayhap]

Whoa!
For a minute there, I thought you were talking about that famous periodical "Bot News". Anyway, for those of you that missed it, here's today's headline:
111001101010100111010101010101010100101011111000
10100101010101001011100100101000000011101101001010
11110101010010110!
Now, you are no longer in the dark.
/M

[Stefan Prodan]

Native Canadian? Weird, for some reason I thought that the site was run from Ecuador.

[Kevroc]

LOL! nice one, I can't stop picturing you typing in all those zeroes and ones. LOL

[Yeknom58]

like 1/2 the time I can't read them. I don't know how a computer could.