NL Bots on Full Tilt

[BluffTHIS!]

nlnut,

Are going to let Brandon go get some sleep now that you're here? He took the night shift for defending your BS story, so you should be able to take the day BS shift all by yourself.

[BrandonJoseph47]

[ QUOTE ]
nlnut,

Are going to let Brandon go get some sleep now that you're here? He took the night shift for defending your BS story, so you should be able to take the day BS shift all by yourself.

[/ QUOTE ]


Lol Bluffthis! I DO, however, like your picture of Doc Holliday.

[Methodz]

so do all of you live together?

or...spend like 6+ hours daily at someones house? or what?

I'd assume you don't have much of a social life then?

[pyedog]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
- I argue that the reason all of the data is so similar is that he is playing on all of them. Think about this: Take 400k hands and split them up into four 100k chunks. Won't the data for each of these 100k hands be very similar to each other?

[/ QUOTE ]

But VPIP, which should obviously be identical across the board, is statistically different between the 4 accounts listed.

[/ QUOTE ]

How about this: At some time t0, he decided he wanted to adjust something in the strategy. He has four accounts with x1, x2, x3, anx x4 hands each. But say they're all different # of hands. The stats for each of those will be different based on how many hands he played at his strategy before t0 and the number of hands he plays with the new strategy after t0.

Example:

Set1: 110
Set2: 111100

Both have average of .67.

Change strategy to play all hands. Add 1 to each

Set1: 110 1
Set2: 111100 1

Average for 1: 0.75
Average for 2: 0.71

Same strategy, different averages, hmm. Maybe theres is a lesson here? [img]/images/graemlins/wink.gif[/img]

[/ QUOTE ]

A valid point.

However, two of the accounts have basically the same number of hands (105k hands and 112k hands), so we can assume they were datamined at the same time. How come their VPIPs are so significantly different then? (13.64% and 14.08%) That difference is over 4 SDs

[/ QUOTE ]

has it been confirmed that the true deviation is over 4SDs? earlier there were like 4 formulas people were trying to use. if true this fact needs much more attention as it has the greatest chance of clearing them

[/ QUOTE ]

I did a few quick calcs...not sure if this is 100% accurate, but I used the correct formula for SD of percentages (it is sqrt( p*(1-p)/n ). The null hypothesis was that they all came from the same distribution, with the mean being the weighted average of all four preflop players (the last line is the sum of all of them). First two columns are the VPiP and number of hands, the third is the individual percentage VPiP. The fourth is the SD using the number of hands of each individual player, and the last is the number of SD away from the mean (13.93% VPiP).

VPiP #Hands pct SD SD away
from mean
14376 | 105366 |0.136438699 | 0.001066697|-2.675179739
15840 | 112514 |0.14078248 | 0.001032258|1.443606827
11683| 82577 |0.141480073 | 0.00120493|1.815679571
5721 | 41414 | 0.138141691 | 0.001701445|-0.676257655

47620 |341871 |0.139292306

So only the first (one4thethumb I believe) is statistically different than 13.92, at the 95% level, but that's not really a result to hang one's hat on. The others are within 2 SD of the mean.

The greater than 4 SD different is adding the two SD from the first two players.

However, add to the SD above slight tweaks to code (assuming they're running bots) or human play that changes things slightly, and I can't reject the null hypothesis that these results weren't generated from the same process (the same mean).

Shane

[/ QUOTE ]

I don't know if you've gone about it correctly.

Also, you're failing to reject the null without specifying any criteria...one of your results differs significantly from the mean (Z of -2.67 and with a 105366 sample size) and you somehow fail to reject. You realize -2.67 has a 0.0038 probability of occurring by chance? That should be very strong evidence for rejecting the null hypothesis)

[/ QUOTE ]

First, I said it fell outside the 95% hypothesis. But I think my results are a bit better than cherrypicking the two most dissimilar results and comparing just those--the issue is with all four of them. And I've dealt with enough tests to know that 2.5SD while according to the 'book' is enough to reject, especially with other issues going on. It reminds me of a quote from a physics prof here (about physics results): "half of all three sigma results are wrong".

What I'm saying is that 'rejecting' a 2.5 SD result while technically correct is a little quick. A slight tweak or human intervention a little bit could cause this difference, and thus just isn't convincing in my mind.

In slightly other words, yes, 2.5 SD looks good, but I believe the SD is being (slightly) underestimated for various reasons. Had the results been 4+SD, then we're talking. Otherwise, I don't see convincing evidence from VPiP

Oh, and I think you have an extra 0 in there? 3SD is 99%, so shouldn't that be 0.03, not 0.003?

[/ QUOTE ]
How exactly are you guys taking into account the random actions of other players at the table? This will affect the VPIP and PFR of these 4 suspected bot accounts. I don't see how the random behaviour of the other players at a table can be incorporated into your standard deviation formulas.

For example, assume the bot or robotic human player has the following actions with KQ in middle position.

if (hand = KQ and posn = MP)
{
if ( opponents_action = raise )
fold;
else if ( opponents_action = reraise )
fold;
else if ( opponents_action = call )
call;
else if ( opponents_action = all_fold )
raise;
}

Now there would be a similar case for every hand and position. If one player followed these actions at a LAG table with average stats of 25% VPIP and 15% PFR for 100K hands then his personal VPIP and PFR stats would be drastically different than someone who played with these actions at a rocky table with average stats of 15% VPIP and 4% PFR.

I don't understand how your STD formulas could be incorporating the effect of other players' random actions on the VPIP and PFR stats of the suspected bots. Don't you think that these variations would affect these guys VPIP stats by at least fractions of a percent even if they were following a rigid system?

I agree that if the bots had programmed actions regardless of their opponents actions then you could expect the stats to be nearly identical after 100K hands though. But if this was the case then the bot would be a huge losing player, so this can't be the case. But isn't this the assumption that you're making with these STD calculations?

[KotOD]

[ QUOTE ]
[ QUOTE ]
I played live.

[/ QUOTE ]

1. They have live poker in Pittsburgh?
2. Why would you play live instead of a different online site? That seems a bit weird to me.

[/ QUOTE ]

These people aren't from Pittsburgh. They are from a really awful place called Johnstown, PA.

[nlnut]

[ QUOTE ]
I'll say it again...

In what form is this "playbook" is it on a computer or something like a printout? Simple question that doesn't involve releasing any "strategy" info.

[/ QUOTE ] neither. its all memorized. However, we did first construct it on paper.

[KotOD]

[ QUOTE ]
[ QUOTE ]
I'll say it again...

In what form is this "playbook" is it on a computer or something like a printout? Simple question that doesn't involve releasing any "strategy" info.

[/ QUOTE ] neither. its all memorized. However, we did first construct it on paper.

[/ QUOTE ]

http://www.rofl.name/roflcopter/

[KurtSF]

Caught up after the overnight. Three areas I want to respond to.


(A) FTP response.

[ QUOTE ]
Why do you ppl want a response from FT? They exonerated these guys. What do you expect them to say besides "after 1 month of research, we found the players to be innocent of all botting charges"

[/ QUOTE ]

Yes, that's what I expect, but they haven't even said that.

[ QUOTE ]
It would lay some suspicions to rest. I personally don't see why they haven't done it already, but maybe FTPDough/Sean are talking with the appropriate department before they make a definitive statement.

[/ QUOTE ]

FTPSean is a key member of the security team. He IS the right department.

[ QUOTE ]
This IS a HUGE ISSUE and Full Tilt has to handle it very delicately.

[/ QUOTE ]

Yeah, and I don't mind if they take their time to say "Hey 2+2, we're positive they're not bots but can't release reasons." but I expect they pop in and say "Yeah we know about this thread and these accounts and will make a comment. Try to hold back the wild speculation until then."

But no, they avoid it entirely. Since they are posting in this forum while this thread has been going, I expect they are in full blown panic mode realizing how badly they've effed this up. Just MHO. Again, I will wait for their comment.


(B) The numbers

I find DWarrior's analysis that the numbers are not the same to be suspect. I don't trust his numbers, but I can't do better. There are however, a few glaring issues.

1. His analysis of VPIP assumes that each hand is an independent event. This is not true. Each "player" is in fact an amalgamation of 9 different algorithms, one for each position. See Trebek's post that contained their positional stats. The "script" clearly has different formulas for each position. I would like to see the Standard Deviation numbers run on these positional stats.

2. His dismissal of the hypothesis ignores that one of the pieces of this hypothesis is that the players do take control from time to time, probably for coverplay. While this doesn't help the case that they are bots, it makes it very hard to dismiss the case with such a simple numerical analysis.

3. NLnut has himself admitted that they changed the "playbook" before, and might again. These 100k hands sample contain several algorithms, which will throw off the SD if you treat them as one.

And this ignores the post-flop numbers, specifically the river numbers. To any poker player those numbers are insane. I can think of no other explanation other than a computer "script" is generating such congruent numbers. Maybe another script is inputting them into the FTP client, making it a traditional bot, or maybe there are people inputting the actions in. But I am convinced that some computer program is determining the actions. Many others are as well.

Perhaps these player stats should be posted in the Probability forum. There are some smart folks in there always willing to help and their opinion, I would expect, would be enlightening. I'm at work so I doubt I'll get the chance, but I'll do it if I can.


(C) Response from nlnut and Brandon, et al

Lies. Changing stories. I think the collective BS meter has gone off the charts. We can all see the smoke and are being told there is no fire. I think we need to see the firepit before we decide this is just fog. nlnut and Brandon, please consider being VERY forthcoming. I know from previous experience that, while embarrassing, 2+2 and the zoo can admit when they made a mistake and were victims of groupthink. If you provide proof, I assure you a whole flock of 2+2ers who haven't posted yet will come out of the woodwork and join your chants calling all those of us who have participated idiots and sheep. Sounds fun, right? Help us out.

1. It won't prove anything, but it would reassure me (and probably others) greatly seeing a photo of the room the borg play in. Not proof, but seeing a livingroom with 8 monitors and 4 guys sitting around will ease concerns that this is just 16 CPUs in your closet. The sooner the better.

2. Why did your team not play together yesterday when its existence was made public? When did you find out about this thread on 2+2? Who found it first... I doubt it was Brandon who never even knew about 2+2 before yesterday, more likely nlnut or more likely still nation. Every other day you all play at the same time, concurrently, but yesterday you took turns and played sequentially. This reeks of trying to cover your tracks once you were outed. Why, for the last 3 months, do you play together every damn day and then separately on the day you're exposed?

3. How in God's name are those river numbers so similar? I know you don't want to share your playbook, but pardon my annoyance seeing as that is all we do here: share our strategies! It's 1BB/100... you didn't discover the holy grail of poker. You won't be revealing anything new here. Even if you don't divulge the exact algorithms which created those stats, at least share the CRITERIA on which the decision was made. What did you look at when you wrote the wrote the "book" on the river?

4. If you won't share the contents on the "book", please share its form. It it written in a notebook that everyone was made to memorize? Is it on giant posters on the walls where everyone can see it while they are playing? Is it in another computer or window that can be viewed while playing? Pics?

FOR TERBEK: Do you play in the evenings? nlnut has shared that he takes a dinner break at around 5 or 6 pm Eastern time. If you frequently (or semi-frequently) sign in around these times that would account for the appearance of the borg avoiding you.

[Backspin20]

[ QUOTE ]
Anyone not care if there are bots playing?

I play the game sucessfully on-line with my bare hands.

I find it funny that the OP is using a robot(pokertracker software) to mine and organize huge volumes of data but he and others are in panic mode about a what they perceive to be a more automated bot that another player is using.
I have tried the 30 day free trial of pokertracker, so i know what it does. I do not use it.

So if a site allows pokertracker, in my opinion, it should allow pokerbots.

With regards to poker tracking software, how would you feel if a site, like fulltilt, fully incorporated poker tracker into their software?

[/ QUOTE ]

FT would never put PT in its software... Why enforce that your a losing player by showing it to you.

[Adde]

[ QUOTE ]
[ QUOTE ]

You guys do remember that one of the ring members (nlnut I think) has posts here on 2+2 looking for mouseclicking scripts that work on Stars right?

Mark

[/ QUOTE ]

This is really a problem with some of the software-threads here on 2+2, they help people a lot wrt developing a bot interface.

[/ QUOTE ]

Arnfinn,

The gap between content of software threads at 2+2 and bot development is too large. Bot developers don't benefit from this. People don't read the software forum and then go develop a bot. It's harder than that.

Also, bring Thor back! [img]/images/graemlins/smile.gif[/img]

Adde