bots in the FT mid-high HU games

[admiralfluff]

Ughh. I browsed through this thread, and have a few things to add. First of all, I will not play HU v. an unknown. I believe the most advanced poker AI's out there are for HU limit Hold'em. This is the simplest form of the game, and is what U Alberta, and various revs of the poki intelligence focused on back in the day. Though the game is still far too complicated to solve optimally with current technologies, it is possible to creat a simplified game-space that is solvable. U Alberta took this approach in the 90s and did pretty well with it, and they did not have extensive poker expertise. Using one of these faux-optimal solutions, and combining it with some exploitive strategy based on opponent modelling would yield a pretty sick bot. Also telling, I remember reading one of the UA papers where they quoted one of the pros who played the bot saying how it was extremely surprising how often the bot went to SD with A,K, and Q high hands. If you've read The Mathematics of Poker you should remember that as pots get smaller optimal bluffing frequency goes up, and with it, the calling range threshold goes down. These SD numbers may look bad to many good players, maybe even some of the winningest pros, but if they are bots they may be optimal.

The most suspicious nugget is that these specific players never play against each other, but will freely play against other pros of equal caliber and results.

[ktulu22]

You people need psychiatric help

[MrGatorade]

Bryce,

As we have talked our heads up live game is still open for you to accept. I will accomidate hotel and food. =) 10 or 20k freezeout if you like. And the 3 that you listed are 100% bots and I challange you to play them heads up for 3000 hands as they will gladly do that because there BOTS and dont tire while I watch and if they beat you, you pay me what they took from you... if you beat them I pay you what you took from them. PM or post if interested. Official bot challange, but I want to pick the bot that I have the toughest time agasint. I am putting my money where my mouth is and I know you to be a better heads up online player then myself so i am sure you can beat them right? =)

[DrewOnTilt]

[ QUOTE ]
Bryce,

As we have talked our heads up live game is still open for you to accept. I will accomidate hotel and food. =) 10 or 20k freezeout if you like. And the 3 that you listed are 100% bots and I challange you to play them heads up for 3000 hands as they will gladly do that because there BOTS and dont tire while I watch and if they beat you, you pay me what they took from you... if you beat them I pay you what you took from them. PM or post if interested. Official bot challange, but I want to pick the bot that I have the toughest time agasint. I am putting my money where my mouth is and I know you to be a better heads up online player then myself so i am sure you can beat them right? =)

[/ QUOTE ]

MrGatorade, I have no reason to doubt the existence of bots. I do believe that bots are out there and that they are problematic for various reasons. But this challenge that you have issued to TheBryce is pointless, and would not prove anything, other than one of the following:

1) TheBryce ran hot over 3000 hands, or the suspected bot ran cold
2) TheBryce ran cold over 3000 hands, or the suspected bot ran hot

[MrGatorade]

I want him to see what i see and realize that he is playing a bot. Thats all and I know he is smart enought to see it. He will recognize that no human can do the exact timings that the bot does. he should also sit in and out once in a while and the bot will pick up right where it left off like nothing happened.

-Crazy Mike

[The Bryce]

[ QUOTE ]
can u elaborate on how u think they get tired/emotional? all these players bloat every pot and never fold, their standard play can pretty easily look like tilt

[/ QUOTE ]

This isn't something I'm really going to be able to quantify, since it's happening on an intuitive level. If I'm saying these people react / think / deteriorate like human beings I can't really show a value for that on paper, I can just tell you that that's the way it is (it is possible, of course, that my observations do not reflect the actuality of the matter).

That being said, what's the criteria being used to identify a bot here? A solid, winning player who doesn't play against other players of similar or greater skill, can play long hours and does not appear to tilt? Allow me to posit that, if indeed this is the criteria being used, it could, hypothetically, be the result of the same mentality the "my results are poor, therefor I am unlucky" thought process. Now, I don't disagree that a bot would show the characteristics that were just listed, but the point is that anyone who is playing well should also be displaying the same characteristics.

Now, as it happens, there is a guy here (me) who happens to have played more against most of these people than anyone else, and for various intuitive and somewhat more blatant reasons he believes that these players are not bots. So I'm just saying that, based on the information available, if you had to make a critical decision about whether these players were either bots or good players, it might be sensible to lean towards "good player" unless you had some additional information.

And if we take it a step further and say that, in theory, these are perfectly beatable players / bots who's abilities are fairly representative of a normal, winning player for the limits they frequent, how will that impact the way the rest of the community plays? Certainly, if they are indeed bots, it would be prudent to badger the sites in question into doing what they can to tighten up on security. Aside from that, however, what will change? Will you boycott the bots? Boycotting a bot will only have an impact on the bot if you were expecting to lose money to that bot. As an individual, I think if you chose to play in a spot where you're going to show an expected loss it doesn't make a fiscal difference whether your opponent is a bot or not.

[The Bryce]

[ QUOTE ]
Bryce,

As we have talked our heads up live game is still open for you to accept. I will accomidate hotel and food. =) 10 or 20k freezeout if you like. And the 3 that you listed are 100% bots and I challange you to play them heads up for 3000 hands as they will gladly do that because there BOTS and dont tire while I watch and if they beat you, you pay me what they took from you... if you beat them I pay you what you took from them. PM or post if interested. Official bot challange, but I want to pick the bot that I have the toughest time agasint. I am putting my money where my mouth is and I know you to be a better heads up online player then myself so i am sure you can beat them right? =)

[/ QUOTE ]

Are we talking 3000 hands in one sitting? I've played plenty of crazy 7 hour sessions, though I guess if they were willing to stick it out for 3000 hands that would substantiate your point. I'll have to consider whether I'd be willing to do that against Bono or BeatMe, but I can tell you that, off the top of my head, I'd be pretty happy to accept against any of the other players. What happens if they quit me?

[gehrig]

[ QUOTE ]

That being said, what's the criteria being used to identify a bot here? A solid, winning player who doesn't play against other players of similar or greater skill, can play long hours and does not appear to tilt?

[/ QUOTE ]
if flopbenice datamines and knows that bobaloca isn't really worth his time, then why is he giving u action? its not like u play a bizarre style that would trick people into thinking ur a fish

the three most active 30/60 hu players in my db are flopbenice, bobaloca and lip2lip. those three have never played one hand against each other that i've observed. also in the top ten in hands played are you, me, et tu brute and inna smart, all of which have a bigger winrate in my db than those three, and all of which don't seem to have much trouble getting action from the players accused of being bots.

that by itself isnt exactly daming, but also consider:

- flopbenice, bobaloca and lip2lip all have identical stats
- bobaloca and lip2lip (i dunno either way about flopbenice) play with a bizarre tempo. why exactly would it take 4 seconds to call every single preflop cap after having 3bet? it's not lag either, they check the flop after defending relatively quickly.

[Victor]

[ QUOTE ]
Posting blind,

FolpbeNice, Boba, and Bono are definitely not bots. Further, only one player on this list is good enough that, if he or she were a bot, that they would be a cause for concern.

[/ QUOTE ]

see bryce this is why i hate you and think you are a complete prick. just bc most of the alleged bots are beatable by you does not mean that they are not bad for the game and "cause for concern."

try to look beyond yourself to realize why.

[The Bryce]

Gehrig, that's much more substantial. Now, saying that you had someone who you thought could very likely be a bot, what would the plan be?