Soooo..........the only reason they caught this bot was that there were multiple bots, and the creator(s) didn't alter the parameters at all?

I'm starting to think the secrecy is actually just a cover up for complete incompetence. In other words, if they had to tell us their top secret methods for Bot detection , we would be underwhelmed. Like the Wizard of Oz.

LOL

Although I cannot be sure, I would think a lot of the HU High Stakes players have similar stats, there must be more evidence to convict someone than this.... surely???

You do realize this is bad and unethical, don't you?

Similar is one thing. What if FT digs into all of their stats, and find everything matches? All of the investigated players/bots play KJ os from the SB exactly the same...A5 flopped trips exactly the same...K8 flopped middle pair exactly the same, etc, etc, etc. I have no idea what they actually found, but I could see how you could use stats alone over many thousands of hands to determine that multiple players are either the same person or bots.

Note to FTPSean:

Please make soom room in your PM box! (Search for KeePM in ATF, if you want to keep them all on your own computer, automatically.)

Thanks.

I know what you are saying, but the data we were shown earlier in the thread was NOT identical, only very similar.

That's a very good point. So it took 170K hands for the percentages convergence to become statistically significant? I'd really feel better if I knew some type of bot-detection software triggered the investigation.

No one has claimed responsibility for tipping off FT, and Mr. Gatorade has denied doing so, FWIW.

Every time this happens, it's the same MO. Roughly 50 grand lost, OP comes up with some sob story to convince others of their innocence.

Pretty sure this sillysal is guilty as charged. Think of the hundreds of thousands this player must have made before getting caught.. Don't feel bad for him/her.

Sure FTP handled the situation poorly, but in the end they did what they had to do.

BTW, I am 99 percent certain pokergirl is actually a guy. He might have his girlfriend post here, but the owner of the account is definately male.

How do you know this???

well, if it is a bot that can replicate itself sexually, it could explain the similar but slightly different stats

lol

I realize that, but it's a reality. Many sites allow it and as such it can be exploited.

fyp

FYFYP

FYFYFYP

Guilty.

Still, it's an important distinction, and I thought it should be recognized that whether it's permitted or not, it still happens.

Face it guys, they have no proactive software running looking for bots if they did it did not work worth beans. That is what they are not going to tell you, we got gotts trying to find bots if it does not hit us in the face we can not find it.

So what happened, botfinder kicked out 3 names after years worth of hands and then FT said, hmmm, lets look into this more takes 3 weeks and decides, I don't think so.

I think it is more like someone gives FT a heads up/tip, they look at stats for a few minutes and freeze up the account. Take several weeks to do some analysis and come to a conclusion, if botfinder was real and worked it would have been flagging this account for months showing a situation that needs attention and not have popped out "guys, we have a bot at 170k, should we let it go to a quarter million"

I am a graduate student who rarely has time to play poker. But as a backup in case I don't pass my qualifying exam, I began a data mining operation on Full Tilt's Limit Games about 3 months ago. Like Gehrig, my reasons are selfish. I will use the database to identify the fish, avoid the sharks, and to improve my game play, so that I can emulate the big winners like Freedom25, Fatal Fog, etc. Unlike Gehrig, I am willing to share this data base with anyone who asks. But I do need to find a site that will host it.

Anyone who sifts through the database will see that Gehrig is not being completely honest in his postings. And I feel that for the sake of sportsmanship, the record should be cleared.


In one of his first postings in this thread he writes:

This should trigger warning bells. First he supplies no evidence. But worse beatme1 hasn’t been playing in the past few months! In no small part due to his efforts, she's been banned!

After this post from Ryanski appeared:
Gehrig “flip flops” on Daurgman:

He had no evidence that daurgman is poker girl but he said it anyway. And he's been trying to bring Beatme back into this even though she's been gone for 6 months. Do you really think he's telling the truth now when he says:
I think he does accuse people if he thinks he can get away with it and it benefits him financially to do so.


Another disingenuous tactic is his desire to tie other purported "bots" to Poker girl by posting their statistics. In a manner reminicent of George Bush talking about Saddam Hussein and Osama Bin Ladan in the same breath, hoping an uncritical listener would assume they are linked without ever explicitly saying they're linked, he outs other "bots" by comparing them to Daurgman. Even though he just admitted he's not sure if Daurgman and Poker Girl are linked.

Then, he makes this curious statement:

It is clear after some analysis of FT HU players that this is complete [censored] that he expects to never be called on. (In other words, he’s lying with statistics.)

For example, if he wishes to use this curious stat to out bots, we should take note that of the following winning players who have the same statistic:

Fatal Fog (12.26%) 6137 hands observed (won over $200,000)
Readeren Metal 7508 hands observed (12.12%) (won over $200,000)
Nizot Skizared: 1500 hands observed (14.01%) (won $100,000)
Quaternion (10.89%) 1894 hands observed (won over $28K)

It should also be noted that PokerGirl z and Greggo both have about a 6% check statistic here, which would appear to differentiate them from Daurgman.


Nevertheless if he wants to argue that the check-flop-after-preflop-raise-percentage statistic is a good way to find bots, note that Gehrig’s 3.1% is very similar to FinddaGrind’s ($722,000 in 3 months) 2.54%. Freedom25’s ($449,000 in 3 months) 3.93% and qsdaddy’s 2.44%. Are they all the same??? My database shows that Gerhig has never played a hand with neither FindaGrind nor Freedom25! They must be bots! FTP should seize their accounts and "reimburse the victims”! I bet they could scoop over 2 million this way!

LOL Tetra.

You just happened to be data mining?
You don't play much poker?
You just happened to be browsing 2+2?
You just happened to stumble on this thread?
You decided to read it all?
This is your 1st post?

And I thought you said you were not going to post here anymore?

Hi tetrahedron, I can host your database. I made an ftp account for you to upload it:
ftp://oxnard.dreamhost.com
user: tetrahedron
password: sent by pm

Or, if you prefer, put it on rapidshare etc. and I'll rehost it from there.

If you upload it, I'll make it publicly accessible and post the link in this thread.

My first reaction to the Tetra post is "welcome back Sal".

My second reaction is whilst the person presenting the info might be dubious, the info itself should be able to stand or fall on its own merits.

I have nothing first hand to bring to the table, but i hope Gehrig or someone like that does so. Either way i strongly suspect this is a case of missing the wood for the trees.

Here are tetrahedron's hand histories.

http://littlegroup.org/tetrahedron/FTMine.zip

(27 MB zipped, ~330 MB once unzipped).

Hey, I know datamining is kosher, but isn't sharing the HHs with others against the ToS?

Also, doesn't AP and UB explicitly allow multiple accounts (up to 3 IIRC) per person?

Most players with a bot would probably play without the bot at times. The stats are the not the only piece of the puzzle.