Bots in PartyPoker's 6-max Limit games?

[Robby Box]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Sounds like a good plan, but in point 4 the sites should still not be allowed to buy control of the project, only to be able to use participance as marketing etc., the sites have been given this trust today and have not lived up to it so no reason to easily give them the trust again.

[/ QUOTE ]
The idea isn't really to let them buy control, it's to buy the results. Subscribe to this service, and you can get access to all playing-pattern detectable bots details. The database can also datamine the sites, to ensure that more levels are covered. No need for volunteer data then, though, again, breaks T&C of some sites.

[/ QUOTE ]

Sciolist, but the question remains... IF the poker sites are interested in cracking down on Bots, why wouldn't they have an internal team (that would have access to more info than a dataminer) doing this.

The other problem, is that there isn't complete agreement between all players and the sites on even where the "line" of acceptable vs unacceptable really is, or in fact should be...

Sites allow "Decision Support" software... Players use some automation tools... at the moment, its only when the decision support software is married to (and not just living with) the automation tools, that it violates T&C.

For the benefit of those who didn't understand that... the line that Pokerstars (thx Sciolist for participating in this discussion), for example, has drawn is... computer pushes fold/bet/raise not acceptable... live human pushes fold/bet/raise acceptable, no matter (almost) how that decision was arrived at.

Side comment... I hear there are a few hundred former CS people in India looking for minimal wages + %commission to push buttons... [img]/images/graemlins/wink.gif[/img]


[/ QUOTE ]

I think that this is an excellent point!

If somebody knows they are within the T&C of a site to run the BOT "brain", push the buttons themselves and win 1.5BB/100 at 10/20 on XX tables, they will do so.

I think that the poker sites need to re-examine the allowing of odds calculators.

Whilst recognising that an off the shelf calculator is useless as a "brain", anyone getting hold of a decent profile for one of these can potentially win money pressing the buttons themselves.

Over to you Stars, Party and FTP [img]/images/graemlins/confused.gif[/img]

[Robby Box]

...and if you did decide that odd calculators were no longer acceptable, how would you detect them?

[N 82 50 24]

[ QUOTE ]
Possible solution for sites?

Have the table position change every minute. Move it left/right/up/down a few pixels.

The most recent Party upgrade caused all the software that uses screen-scraping quite a bit of grief.

[/ QUOTE ]
That wouldn't stop anything except really piss off the humans. The upgrade had nothing to do with moving tables.

[teddyFBI]

If I were a poker site, the following seems like a very easy and productive way to hunt down bots.

Assumptions

#1 Bot makers concentrate mostly on LHE, rather than NL games.

#2 Bot-makers will run more than 1 bot (using different accounts)

#3 A single Bot-operator will never seat more than 1 of his bots at the same table.

The above will not ALWAYS hold true, but #3 is potentially the most important. It would be very, very, very easy to run a regression analysis to determine the likelihood of any number of players NEVER sitting together. Even over just a couple weeks, I think the results would converge to a sufficient degree to draw some very, very accurate conclusions. This is made even more true because of how few LHE games of 10/20 and above are left even on the biggest sites...if there are only 20 or so such games going at any one time (assumption), even just a week or two of observation will reveal a pattern of which players NEVER sit with each other.

The preceding is not a FOOL-PROOF strategy, but just because PERFECT regulation is not possible, that doesn't mean that EFFECTIVE regulation is equally unattainable. Catch a few cheaters, confiscate their funds, and send a big message to bot-operators that you're capable of catching them and seizing their funds, and it'll go a long way toward disincentivizing bot use on your site.

The above are only half-baked thoughts, but I think the gist of what I'm trying to say is that a reasonably smart stats guy would have ZERO trouble looking for and finding suspicious patterns simply based on the tendency of players to either sit together (implying possible collusion), or NOT sit together (implying a conscious effort by a bot-operator to seat his bots separately.) Elementary principles of randomness are incredibly reliable indicators. Seems that there just isn't, at this time, sufficient incentive for poker sites to discourage the use of bots...we'll reach a point, though, where it becomes unavoidable. (Stars, are you listening?? Party, I have no hope that anyone in your management is sufficiently literate nor in possession of the opposable thumbs necessary to understand and implement the above, but Stars -- you guys are smart enough to get the ball rolling on this; hire a single Stats PhD for 6 months, and he'll teach you some really important stuff. <insert joke about my mom being a Stars bot here&gt

[OldLearner]

[ QUOTE ]


Quote:
--------------------------------------------------------------------------------

Possible solution for sites?

Have the table position change every minute. Move it left/right/up/down a few pixels.

The most recent Party upgrade caused all the software that uses screen-scraping quite a bit of grief.


--------------------------------------------------------------------------------


That wouldn't stop anything except really piss off the humans. The upgrade had nothing to do with moving tables.


[/ QUOTE ]

Moving the area that the screen scrapers "expect" to find what they are scraping for effectively disables them, does it not? (I may be just naive about this).

Moving the cards/text on the screen a few pixels or subtle resizing will not "piss-off" the humans. They probably wouldn't even notice. And if it is noticeable, knowing that it is to combat bots would make it acceptable to me.

[Arnfinn Madsen]

Good suggestions, Teddy, there are about a 100 different similar things they can do too but don't feel like posting details here since it will just make it easier for botters to stay ahead. A friend of mine says that nobody can run a backgammon bot for more than one week without him noticing simply because he has developed a few tricks to identify them that the bot makers are not yet able to beat (it stems from analyzing the logic of the player, in certain situations the computer simply does not manage to appear as a human), if somebody would put the effort into the same in poker they would manage. This is a war to be won if the effort is put forward, therefore everybody should keep on pushing the sites until they step up and do what they have to do.

[maurile]

[ QUOTE ]
Assumptions

#1 Bot makers concentrate mostly on LHE, rather than NL games.

[/ QUOTE ]
I'm not so sure this is true. A winning short-stacked NLHE strategy would be pretty easy to program. A bot should just buy in for the minimum, look to get all in (or at least half its stack in) before the flop with its strongest hands, and leave the table if it doubles up.

As long as sites allow very small buy-ins, I think a short-stacked NLHE bot would be easier to program than a LHE bot.

[N 82 50 24]

[ QUOTE ]
Possible solution for sites?

Have the table position change every minute. Move it left/right/up/down a few pixels.

The most recent Party upgrade caused all the software that uses screen-scraping quite a bit of grief.
[ QUOTE ]
That wouldn't stop anything except really piss off the humans. The upgrade had nothing to do with moving tables.


[/ QUOTE ]
Moving the area that the screen scrapers "expect" to find what they are scraping for effectively disables them, does it not? (I may be just naive about this).

Moving the cards/text on the screen a few pixels or subtle resizing will not "piss-off" the humans. They probably wouldn't even notice. And if it is noticeable, knowing that it is to combat bots would make it acceptable to me.

[/ QUOTE ]
Well, the bot wouldn't be clicking in a spot on the screen as a whole, it would be clicking within the table itself. So moving the table does nothing. I don't know how else to say this except that this idea won't work so there's no point in discussing it further.

I like some of teddy's points and I like the idea of outside analysis to provide a check on the security teams at the major sites.

[michaelantoi]

[ QUOTE ]
1. Getting site's tentative aggrement (say, PS, Party, FT)

[/ QUOTE ]

Shouldn't someone draft an email and get in touch with the poker sites to atleast determine there stance on the subject.

It would atleast get the ball rolling. I just hate the fact that we're sitting around talking about it while the bots are out there stealing everyone's money.

[Sniper]

[ QUOTE ]
Assumptions
#1 Bot makers concentrate mostly on LHE, rather than NL games.
#2 Bot-makers will run more than 1 bot (using different accounts)
#3 A single Bot-operator will never seat more than 1 of his bots at the same table.

[/ QUOTE ]

Your assumptions are flawed... there is not only 1 type of botter

[ QUOTE ]
based on the tendency of players to either sit together (implying possible collusion), or NOT sit together (implying a conscious effort by a bot-operator to seat his bots separately.)

[/ QUOTE ]

Isn't that Table selection... oops

[ QUOTE ]
Seems that there just isn't, at this time, sufficient incentive for poker sites to discourage the use of bots...we'll reach a point, though, where it becomes unavoidable.

[/ QUOTE ]

I don't understand what you mean by "unavoidable"... and actually the point of this sentence altogether... since, Poker Sites do already "discourage the use of bots"...