Hi all

Firstly, I am sorry if botting is a bit of an over talked about subject here. Hopefully I can come at it from a slightly different angle!

A caviet - I have a personal interest in botting, not as a botter (I am not one!) But a general interest in AI, equilibrium game theory and automating human decision processes. I have no interest in running bots at poker sites, but I do enjoy the process of tweaking and running them in PA

So, lets say a site emerged, with all the usual features of a decent site. Rakeback, bonuses, good game selection, decent deposit/withdrawal ect. But with one big difference. They openly allow bots. This would work in this way:

• Certain tables only are labelled as bot friendly.
• On these tables bots are clearly marked
• Humans are welcome to play on these tables, giving them a chance to beat the bots.
• A strong system is put into place to detect bots on all the other tables. If they are found on a none bot table, they are out. No question.
• The account which sign up as bots are never allowed on any other tables.
• Players are allowed to sign up two accounts, one as a bot (again, clearly marked and only allowed on bot friendly tables), the other as a human, allowed on any table as normal.
• Botting is never allowed using any human account, and again, a strong system of detection/deterence is implimented.

Would people be interested in playing at a site like this? What other safeguards, guarantees would they like to see in place? Any other thoughts or comments at all

David

I guess the whole point of running a bot (for the time being at low limit tables) is doing so "incognito". Bad players will never know the difference (and perhaps good ones either) and so you can gain an advantage from having the machine play "solid poker".

Now if everybody knows they're playing against bots, with the current state of the art, I feel bots wouldn't stand a chance.

There you go, this reply will considerably up the % of serious responses in this thread.

I think the argument against this type of setup is that the bot makers would just use it to fine-tune winning bots and then let them loose on other sites.

If you don't already know about it then the UofA Poki servers might be of interest: A human can play using the applet found on this page and a "bot" can be implemented using this protocol to play their too.

Juk

Thanks for your response. Though I cannot see why this question would not recieve serious replies. I don't believe it is a stupid or careless question after all.

Its true that this might not be perfect for the bots, since as you say flying under the radar is likely to be important to their game. I know myself that once I spot a bot at a table I start being much more able to beat them, since I can rely upon them making the same decision over and over again.

What would you prefer to the above? A site which claims it fights bots, but in reality gives them tacit acceptence since it boosts rake and traffic (who is to say this is not happening already?), or a site whch allows bots but does not mark them as such?

Obviously we would prefer no bots at all, but the question is regarding a site which allows them, and the best way to go about that.

Hi Juk

Yep, I know those links. PA bots are a hobby of mine. The ones I make would never beat humans, but I enjoy tweaking bots vs bots in PA, as well as playing against my own creations. Thanks for looking them up all the same though

Yes, it could be argued that bot makers would use such a site for tweaking their bots before sending them to other sites, this is obvously a downside. However, it could also be argued that humans could use such a site for tweaking their play against bots, and so profiting from them at other sites, perhaps levelling the playing field somewhat in the process.

I wonder where IndianaV8 is.

lol @Gunth0807.

Just came to 2+2 to check the state of affairs and you can imagine the delight!

It goes without saying that I will support such site with all the marketing I can

I argued quite a lot that this is a good idea. A site that promotes bots for real money will attract large ecosystem of players trying to make a winning bot - and all these guys will lose money in the process. Once they build a winner, they will try it at higher limits - and again lose money in the process. This is how poker works anyway. The more people donate money - the better for me (and I suppose everyone).

There are two ways how you can implement such site.

1) Implement it with a nice, easy and conviniet APIs. Bot authors will focus on the bot logic. Any move from this environment, to a disguised botting will be a big implementation change, therefore many people that would do this for a hobby, will not make the change.

2) Implement it with no APIs, just ask authors to code and scrape against the poker interface. This option has more interesting implications:
a) More work for bot authors
b) Authors will be able to move their bot to other tables and try to claim they aren't a bot
c) Site will be able to train their detection systems better

Obviously, both options will allow someone to claim he's a bot - but play as a human. That may have interesting implications too.

There's my buddy!!

Hi

I just want to say that I don't want this thread hijacked by a discussion of the rights and wrongs of online poker botting at sites where it is not allowed (i.e. all of them!). This has been discussed elsewhere at length. I think we can agree that the majority believe this to be, without question, wrong.

The intention of this thread is to discuss whether it would be a good idea to have a site which openly allows bots. Along with the best way this could work, or could fail for that matter!

Indiana, I know you are pro botting, and I respect your right to that opinion (as I respect everyone's right to an opinion, whether I agree or not!). I also know that your knowledge and opinions would be a valuable addition to this thread.

So, forgive my ignorance here, as I said above my experience of botting is observing and occasionally outplaying bots on a few sites, along with playing around in PA. So the connection methods are a bit alien to me. From what I can gather having googled the subject, bots use software which reads the screen (like a human looking at the screen), and click accordingly? Is this correct? If so this is suggestion 2 on your list.

Suggestion 1 would be having some sort of interface which communicated directly with the poker site at dll level? If so this defeats the need for screen reading. Correct?

As for your points, I think it would be vital that any such site be able to guarantee that bots could not operate on the non-botting tables. Not sure how that could be done as yet, but as a thought exercise lets assume that this is 100% doable.

As for people pretending to be bots, I think it works both ways. Bots must be identified, and must be bots, humans must be identified, and must be humans. This seems to be the only fair way to operate.

bots r bad so thats fine to me

they do this on some chess servers. you can run your chess bots as long as you register them as bots and people know what they're playing against. i agree with OP there there is a place for this kind of stuff.

Totally agree with op..its stupid

Yes - just a DLL for direct programatic communciation. It could also be an open protocol (e.g. via XML) that will allow people to develop client side wrappers for any language - Java, C, etc. The will be able to do it over C dll too.

Detection of humans if they say they are bots will not neccessarily be good (nor easy). This idea may open a whole new market with interesting implications, e.g. a semi automated console, which is just a righ client-side GUI implementation where some hands or situations are played automatically, and others are played with human assistance.

People will also certainly start to sell software which make confugurable amount of trivial decisions automatic, and ask the human to play only the tough spots.

Ok, why would any botter want their bot to play on bot tables?
The incentive is about zero percent as far as I can see.

Thats a lot different...
There is no money involved.
Would I set up a chess computer for money, Im pretty sure the computer would win 100times more money without the label (C) on it.

First, cause no newbie would play it and also cause the human would know hes playing a bot and can adapt to it.

Here Here!

No bot or ai can handle a good poker player.

i dont really understand your point other than the part about chess not involving wajoring money. (although i would say that although chess isn't for money, on some of the more secure online chess servers, ratings points aren't completely without value. they have some effect on marketing software, selling text, etc.) of course chessbots get much less actions from humans when they are labeled as such. that is why they have that restriction, and why they are allowed to play with that and other restrictions, in their designated areas/ways. chessbots primarily get action from other bots.

there was a pokercast a while back about some promintent 2p2 guys helping with the development of some limit superbot that can aparently hold its own pretty well vs. the best in the world. most people don't consider this to be evil, because its interesting, and it isn't being used to cheat unknowing people out of money.

I disagree.

Most fish and other marginally +EV players would refuse to play on the bot tables.

I'd personally lke to play against a bot at a FR NLHE table because I (stupidly?) don't think a bot can beat a good human player there.

Ok, lets say we set up special bot tables.
Do you think the botters would play there vs each others (With minimal or no edge.) and pay 5-20$ rake for every table hour?
Of course not!

What you dont understand?
No programmer would set up a pokerbot this way cause no fish would play it so they wouldnt win money (Read steal.)

why's that? i'm sure there are some people that would pay just to have a controlled enviroment where they could run bots for play chips. some people just have an interest in that sort of stuff because they have an interest in AI and poker theory. also, most human players have a losing expectation in the games they play in, and a very small percentage are more than marginal winners. would people just throw away money for no good reason? of course not!

the main point is though, that i just don't think it would be as destructive a thing as a lot of people seem to think. i really don't care whether or not its good business to offer something like that.

there are already a lot of programmers that have spent significant time/effort/money developing bots that can only interact locally and serve no practical purpose other than research value.

This is not true. Given certain conditions, equilibrium can be reached. And, who knows maybe some day we can solve the game more, checkers was solved recently.

well, computers basically just brute force checkers/chess. that's why they got so much better in the last decade as processing power went up exponentially. poker bots are much more limited to the skills of their programmers.