Printable View
I just read this on 2+2... normally I don't believe in stuff like this but the evidence is too compelling not to. Everyone who plays at FT should be emailing support and voicing their displeasure.
http://forumserver.twoplustwo.com/sh...0&fpart=1&vc=1
Wow, thats nuts...great detective work by that guy.
Very interesting, I didn't read any of the responses...
Wait.
But how GOOD are the bots? Just because A computer is doing the math, doesn't mean that they are unbeatable. It's been proven time and time again that a computer can't handle an aggressive opponent. If you think you've found a bot, kick it in the nuts by being uber aggro every time you're heads up.
Props to that guy for the work.
I'm still not convinced that the accounts are in fact bots, but they are definitely controlled by the same person. I can't believe that FT hasn't taken action yet.
wow i just read that whole thread, took like 15 mins, very interesting... lol they even found out who the guy was
If you read the post, the author did find a "pattern" for them and did exploit them somewhat. Apparently, the owner of the bots may have taken control of them at some point. Whether they are truly bots is probably debatable but there is some type of cheating/collusion going on.Quote:
But how GOOD are the bots? Just because A computer is doing the math, doesn't mean that they are unbeatable. It's been proven time and time again that a computer can't handle an aggressive opponent. If you think you've found a bot, kick it in the nuts by being uber aggro every time you're heads up.
This is scary. The worst part is FullTilt seems to condone the activity displayed by these accounts.
I've played with the three alleged bots a lot. In my mined database, they're all winners, mariojr for 2ptbb/100 over 28k hands, full_tilting for 0.5ptbb/100 over 26k hands and 1forthethumb for 0.75ptbb/100 over 20k hands.
I think the evidence is pretty strong that they're bots, but it could also be explained if they're friends who play the exact same strategy (and always avoid playing at the same tables) or one super-human who 8-tables on 2 or 3 accounts simultaneously. The most amazing part is that three accounts can multi-table and put in enough hours to be 3 of the top 6 volume players in my mined database AND NEVER PLAY IN A POT TOGETHER. The fact that they have identical stats is weird too.
fwiw, I'm down ~$110 over 8k hands against the three of them :( I always thought they were predictably nitty taggs, but I guess I didn't take advantage of it.
QFT, it takes time to develop a read and someone who's just lightly trimming you isn't going to draw attention compared to the guy who's trying to skin you alive.Quote:
Originally Posted by zook
This isn't some uber-bot by the sounds of it. With tools I could download off the internet, I could probably build something like it in a week or two. However, I wouldn't have faith that such a simplistic approach would be +ev.
How would your code interface with the poker program? I'm just trying to understand how this works.
Think of the rakeback on 3 high volume accounts...Quote:
Originally Posted by Fnord
Google WinHold'em. They pretty much solved the problem. I know people who could take that up a notch. The value just isn't there compared to other uses of our time (incluing playing poker straight-up.)Quote:
Originally Posted by litetruck
I still wouldn't have that much faith in the c-bet at 200NL. Very interesting. I probably would waste time trying to balance his flop range better, then dig into 2 barrel ranges.Quote:
Originally Posted by zook
Posted on 2p2 by a mod and friend of one of the "bot" account owners:
This seems plausible to me. He's posted a follow-up that's convincing as well.Quote:
Originally Posted by nation
If you could 8 table on 3 accounts simultaneously, why wouldn't you 24-table on a single account?
And if you could run 3 bots, why not run 30? Or at least 9.
So, something doesn't add up. Sounds like it's some kind of "pseudobot" player assist program, where they are roommates or something and wrote a log-on script that prevents them from joining eachothers table (and also the original 2+2 posters table).
Some kind of sooped up PAHUD.
Finally - if the bot poker algorthim is simplistic, no reason why someone couldn't just play the same way - and actually with the huge number of hands put in by these accounts, their stats would be quite stable.
Full Tilt only lets you play on 8 tables at a time.Quote:
Originally Posted by zenbitz
But you're right, there are other logical explanations, see above.
You can get it increased to 12-tables and probably as many as you want by emailing them.Quote:
Originally Posted by zook
This could get ugly.
Bottom line is that, if done right, FTP doesn't know what the fuck is going on.
I guess it's somehwat plausible, but it sounds so contrived to me.Quote:
Now you know why they play so similarly; they play exactly the same. Chuck is a good guy, and to be honest, doesn't even know how to configure his virus protection; I had to do it for him. They play next to each other for motivation and chop up profits to reduce variance. Coming in every day, grinding, and sharing in the highs and lows makes it an enjoyable experience for them and is why they're able to grind for so long every day. It's a couple friends just shooting the shtt and putting a jihad on the 200nl tables.
The amount of discipline required to play exactly the same way over thousands and thousands of hands is extraordinary. To expect that not just one person, but mutiple people are doing this and they all happen to be playing from the same IP address is really stretching it IMO.
Are these accounts ever not playing simultaneously? I mean I've played over at a friend's place before and vice versa but this guy seems to be suggesting that the only time they're playing is when they're together. If I was that into poker that I was putting in thousands of hands on a regular basis I think I'd put in some hands from the comfort of my own home on occassion, not just when I was with my friends. If these are indeed bots though it makes sense that they're always playing from the same IP address. If you had mutiple bot accounts why run just one when you can have several playing. Does none of them ever get sick or have other obligations that they might not make it to play that day?
He acknowledges that they're playing every day together and for very long hours, but he tries it to make it sound like just a group of friends "shooting the shit" and playing poker. I mean who really does that? Is my bullshit meter just overly sensitive or does anyone else not see multi-tabling over tons of hours every day as a leisurely activity?
Who here has multiple friends that play online poker on the same site you do, at the same time and for long hours like you do, and also at the same stakes you do but you have never played a single hand with them? Anyone? I know I'll sit on the same table of someone I know just to fuck with them or we'll play occasionally sit at a low limit table and screw around. You know you're not colluding so you don't even think twice about it. If you were using bots though I can see why you'd make a point never to have them play at the same table to help prevent scrutiny. If these truly are a group of friends just having a good time together I'd expect them to act more like friends and maybe donk it up on a micro table or play HU on occassion. If you were a bunch of bots though, what's the point?
Also, who shares profits like that? Swapping out pieces of yourself with others in a tourney or staking someone is one thing, but for a group of people to grind it out in cash games every day and distribute the wins and losses sounds like more BS. That excuse does cover the bases though if it was just a single person swapping money around multiple accounts.
The whole tone of that guy's statement seems like he's trying to justify his "friend's" behavior. He didn't just say "Hey, I know this guy. He's my friend and I can vouch that he isn't a cheater." He instead specifically makes a point of justifying certain behaviors and mentions several times how they all comply with FullTilt's terms and conditions.
Hi, my name is JJProdigy. What's that? Multiple accounts? Ohh no, that was just my elderly grandmother that I'm teaching poker over at my house playing a large buy-in MTT on an account I created for her. . .
Also, they're not splitting up a very big pie with that win rate. Their time would be better spent flipping burgers. Then again, the same could be said about most drug dealers.Quote:
Originally Posted by DaNutsInYoEye
They might have just done this just well enough to put FTP in a difficult spot.
Edit: Really FTP best and only viable play here is to ask the players in question to leave and not come back (players get to cash out), then show some commitment to removing bots to the community.
Very true, but if these accounts were getting rakeback then it suddenly becomes very lucrative. $1/2 NL isn't micro stakes and even if they were breaking even, the amount of money that 3+ high volume accounts could make would be very respectable in most people's eyes.Quote:
Originally Posted by Fnord
It depends on how FTP calculates the player's rake contribution. Playing that tight will generate less rake over time than play from stronger players.Quote:
Originally Posted by DaNutsInYoEye
I agree with dnuts almost entirely, I can't buy what nation is saying... The stats are almost 100% identical..
Someone in the thread is also saying how the vpip/pfr isn't 100% identical and would be if it were a bot. This guy has a good response to it:
Quote:
As to your first point, there are 2 explanations. First, even over 100k hands, hand quality distribution would not be even. I haven't run the numbers, but from my experience this is a very tight grouping, not a loose one. Can someone do a monte carlo sim or a regression analysis to show this? And second, in was clearly stated in the very first post of this thread that the bot owner will take manual control from time to time, and this also is enough to explain the discrepancies.
This is exactly what I was thinking. I have a system, but my starting hands vary over 10K samples all the time. I can only anticipate that they would continue to vary over more hands. I also evolve my game as time goes on. I guess these guys have no evolution to their game. Same gameplan everytime. Like DaNuts said, my bullshit meter is sending off all kinds of alarms.Quote:
Originally Posted by DaNutsInYoEye
If you play a set strategy over many 1000's of hands, then your stats will be very stable.
If you are playing 8 tables, it's much easier to play every hand mechanically rather than think about it. The delay in actions is explained by a real person 8 tabling.
This is kind of a reverse Turing test. A "Turing test" for artificial intelligence is that if you ask questions blindly to an AI program and a person, and you can't tell the difference then you have achieved "intellegence".
Finally - if you could write a program to play like this (and probably you could) then you certainly could PLAY like this.
Question is, what difference does it make? If they are telling the truth, then there is clearly nothing wrong with this approach (other than dying of boredom). If they are lying and it is a bot - so what! It's been shown to be equivalent to 3 guys in their undershorts playing unimaginatively.
Didn't read all the replies, but wanted to reply to this:The bots are taking money from the fish. We can take money from the fish faster than from the bots. Hence the bots reduce our winrate. (if we are able to exploit the fish more easily than we can exploit the bot, which is probably true since they are so tight pre).Quote:
Originally Posted by zenbitz
It's dealt not contributed at FTP, so nittiness doesn't matter that much. When I was 6-tabling 1/2 NL FR at FT I was making $8-$10/hour from rb. These guys 12-table and probably play at least 30 hours a week, so they're making $2k/month a minimum from rakeback, imo. Possibly as much as $3k. A 1ptbb/100 winrate on 100k hands is $4k more. Not too bad.Quote:
Originally Posted by Fnord
I'm with dnuts though, my BS alarm goes off at almost every post by nation, nlnut or brandonjoseph over at 2p2. I think these accounts are some combination of botting and supervised botting.
some of you guys seem pretty naive to me.
there's no way i could sit down with my two other siamese brothers and play the EXACT SAME WAY over thousands of hands. ...much less two "buddies" that think completely differently than i about a given situation.
i have an autistic son, and he cant sit at a poker table and play the same cards the same way...EVERY FRIGGIN TIME. the human brain simply does not have that kind of memorization power to remember a "blackjack-type strategy" on steroids.
they are definitely bots, and FTP is giving it their seal of approval.
for once, the guys in the chats yelling, "this #@%!! site is #@%!! rigged" may be on to something.
I googled winholdem like fnord suggested and holy cow. Bot programs, scripts, how-tos, etc are just sitting out there for people. There is a forum for the bot makers on how to set them up and code them.
If these guys are smart enough to code these things to actually win, surely they are smart enough to be good players. I guess 20 bots @ 70% of some people is better than 100% of one.
Is botting online poker counted as fraud in the same way as credit card fraud is?
The short answer is no.Quote:
Originally Posted by Ash256
Credit Card fraud will probably get you sent to jail. Botting will cause you to lose your account (and any money in it) and at worst sued.
Quote:
Originally Posted by givememyleg
i lol'ed so hard.
Looks like estrop finally made it as a balla.
The 97% cbet stat is giving me pause for thought. Do we all accept that this isn't optimal, even for a nitty strategy, at $200NL? However, if you were programming a bot to play as simply as possible while maintaining expected value, you would, I guess, have it cbet a large amount.
Sounds like bots to me. And I am totally anti-conspiracy theory in general.
Wow, I can't believe how long that thread is on 2p2...
Great detective work by the OP - bots for sure. I'm avoiding Full Tilt like the plague.
lol, guess the only network i dont have an account on...
That thread is pretty entertaining. That guy is definitely is up to something. I'm sure he plays occasionally to not arouse suspicion and manually takes control but he has some program set up to play for him for sure.
He couldn't even get his story straight in the thread at times.
From the 2p2 Forum:
LOLQuote:
a sweatshop nit farm.
Seriously though, i hope FTP does something about this. I really like their RB system and am not keen on taking my "business" anywhere else.
It's 3 "bots" playing 8 tables each.Quote:
Originally Posted by HalvSame
There is no evidence they do this (at least presented). The evidence is that:Quote:
Originally Posted by Chopper
1) they play 14/7.
2) they always c-bet the same amount
3) they always raise 3x
How hard would it be to make a list of starting hands and raise/limp/fold/re-raise them? Isn't this how we all learned to play LHE?
I am not saying they are NOT bots, but it's not like there's PROOF.
Why would you program a bot to play as simply as possible? How would you get 97%, anyway? Always c-bet unless flop is monotone and we don't have one of the suit? Or is it checking 97% randomly?Quote:
Originally Posted by biondino
97% seems so far from optimal, that it's probably evidence AGAINST a bot, and FOR a person playing 8 tables who can't make super deep decisions in 30 seconds.
I mean - if there's one thing a bot can do that a person can't it's make straightforward decisions incredibly fast.
Funniest shit I've seen all week. Wow never laughed at loud so hard on this forum like this lmfao.Quote:
Originally Posted by givememyleg
BTW no way in hell these aren't bots.
There is more evidence that than. Making and sticking to a post-flop strategy and playing it perfectly over several 100k hands is pretty unlikely to say the least.Quote:
Originally Posted by zenbitz
Here are all their stats:
http://img137.imagevenue.com/aAfkjfp..._122_352lo.JPG
The bot is programmed to c-bet 100% of the time I guess, remember the owner takes control manually at times so he could help lowering it to 97% by not c-betting.Quote:
Originally Posted by zenbitz
UPDATE: Official Full Tilt Poker Response to Bot Thread
http://forumserver.twoplustwo.com/sh...0#Post10324465
OMG this is soooo LAME
Not that I would before, but NOW, I'll never play FTP again, EVER.
One thing I like about FTR is that even on the most popular threads, you don't need to lose three hours of your life reading just one of them.
from 2+2:
"But guys, think of all the rake they make off these guys, they play all day and don't even need breaks to take a piss!!! "
QFT LMFAO
That thread is now just insane. 1842 replies in less than 10 days, and still not a good response from full tilt. It took ~5 mins for the whole thing to load!
for a month or so about a month back on Ultimate Bet the .05/.1 LHE 6max games were full of bots. they weren't the best, but were there working off bonuses or whatever. my friend and one other guy, it seemed, were the only 'real' players who knew about the bots. at first, my friend was losing to them, but when he realized they were bots he started playing every single hand dealt agaisnt them, and ended up absolutely raping them for a very high winrate. he wasn't using PT so he doesn't know how much exactly, but he was making like $100 a week off them. keep in mind this was .05/.1 LHE. extremely low stakes.
once UB eliminated the particular bonus structure all the bots disappeared, and the UB .05/.1 LHE 6max games reduced in number dramatically.
these bots were probably not the 'best', however, they were beating him until he realized they were bots and began to capitalize on their rigid, flawed programming.
i would gladly play at a table full of bots, and if i was i wouldn't tell anybody about them since all i'd need to do to annihilate them is figure out when they're weak and when they're strong, and they will not change.
Which this guy did - learned how to exploit the bots - as part of the process of determining if they were bots..Quote:
Originally Posted by wufwugy
However - bots are against the T&C of poker sites and thankfully the guy int he article was good enough to report it so they could (hopefully :p) stop the bots exploiting other people - not to mention the rake back systems etc..
no way im reading that entire thread, but from the little i did read it seems a little more likely that it's not bots, but one guy with multiple accounts. some guys can 24 table just fine. a few of the 'bot tells' are not bot tells, but actually the opposite, and if a mod is vouching for one of the accounts yet leaving some info out it may be because the account holder is using multiple ones, which is desired to be kept secret.
yea that's not a bot. good player massive multitabling more like it.Quote:
Originally Posted by 2p2 OP
the idea that the 'bot' was tweaked in such a way in such a short period of time strikes me as ludicrous.
and yet again i am reminded why most of the posters on 2p2 (or anywhere, for that matter) are morons. nation's explanation is substantially more plausible than any others.
Im not convinced either way, but some of the stats look a bit bot like to me. The beast has a 14/3 split in early and mid positon (exact same stats in each seat), but when it's on the HJ it jumps to something like 16/12, and is the same on the CO and button. Also, 97% continuation bet!.
You would think one person playing this number of hands would have more balanced stats. IMO this is either a sweat-shop, or a bot-shop.
I think only a fool would think nothing is up with this, Wanting to believe this is just some guy or a couple guys playing the same or together is foolish and ignorant. Stop closing your eyes, this is not impossible and is probably being used and tested in more places than you can imagine. Even the AI on Video games could probably beat $50NL at most tables, if you program them weak tight and put in some stop gap measures so you don't lose your shirt.
uh, some of what the OP points out for these being bots actually suggests the opposite.
recognizing the evidence for what it is isn't eye shutting. however, i haven't been over all of it so i wont state anything emphatically.
WTF? you come on and say this is absolutely, in no way a bot and people are idiots for thinking so. But you don't have the time to state any reasons why?Quote:
Originally Posted by wufwugy
Posts like yours are why the 2+2 thread is so long. There are lots of people making contrary posts without people offering any basis for their thinking.
no i didn'tQuote:
Originally Posted by swiggidy
people are idiots for jumping to conclusions without adequate evidence. standard human practice, unfortunately.Quote:
and people are idiots for thinking so.
i did state a few. did you not read?Quote:
But you don't have the time to state any reasons why?
not quite. ducy?Quote:
Posts like yours are why the 2+2 thread is so long.
lol i love ducy
true. there is something interesting and unresovled, no doubt, as well. but jumping to conclusions is, well, for a lack of a better expression, ZOMG FRIST KILLED POKER TIME TO GET OUT BEFORE WE ALL LOSE OUR MOBNIES TO SHARKS AND ROT IN POUND-ME-IN-THE-ASS FEDERAL PRISONQuote:
There are lots of people making contrary posts without people offering any basis for their thinking.
http://aarontodd.casinocitytimes.com...les/34625.html
The head guy for this sweat-shop/bot-group whatever you want to call them is interviewed and gives his side of the story.
Almost makes me feel sorry for the guy (almost).
"We make more than we would make without a college education, and I don't have a college education. I finished high school and that's as far as I went. Any job I've ever held, I've never made more than $10 an hour, and I can tell you I'm making more than that playing online."
Anyone think he's paying his taxes?
lOLQuote:
Originally Posted by givememyleg
That's one of the most unsettling threads I've read in some time. I can only hope that Stars has some sort of net in place to catch these criminals. LMAO@:
Quote:
I even started doing goofy stuff like bot opens, I minreraise, bot calls. I cbet 1/3 of the pot, bot folds. Next orbit, bot opens, I minreraise, bot folds. I’m going to be rich.
so not botting, and even if it is it's not a problem. there is too much guesswork for a non-AI bot to beat a thinking, experienced player. it will never happen.
I'm not worried about the bot beating me, I worry about all the fish the bot is rat-holing money from.