Absolute Soulreading/Rigged thread #3

[krumeluren]

[ QUOTE ]
Results:
for 100 hand spans using 13400 hands the confidence interval (usually you use 90% or 95%, that means in 90% you will be below this win rate over 100 hands) is:
90%: 103.65bb/100
95%: 124.07bb/100
99%: 164.15bb/100
99.999%: 283.99bb/100
99.99999999999%: 480.77bb/100
or 1:10000000000000 or 1:10E14

[/ QUOTE ]


Interesting analysys. Good work. We need more of this.

If i interpret your results correctly, under the assumptions your model is correct, if you have 100.000 streaks of 100 hands each, one of those will show a win rate of 284bb/100 just by chance? How many hands are there in ppls PT-databases (which would be the population) and what is the probability that every once in a while we will find a maniac with those winnings without cheating?

And what if you increase the standard deviation a little? I mean, this kind of maniac could have a SD of 100 right? That would make it even more probable that such good streak could occur right?

your model also assumes that winnings are normally distributed. Do you think this is a good assumption and in what way do you think violations th normality would affect the estimated probabilities?


EDIT: I just saw that the had a SD of 270ptbb/100 and that it was 190 hands. That will give completely different estimates. It would be very interesting if you could present a new analysis with those paramters.

My guessing is that you will find a very high probabbility that such streak would occur in ppls PT databases, more than once.

[darkcore]

i need more popcorn i guess...

[ QUOTE ]
[ QUOTE ]
Results:
for 100 hand spans using 13400 hands the confidence interval (usually you use 90% or 95%, that means in 90% you will be below this win rate over 100 hands) is:
90%: 103.65bb/100
95%: 124.07bb/100
99%: 164.15bb/100
99.999%: 283.99bb/100
99.99999999999%: 480.77bb/100
or 1:10000000000000 or 1:10E14

[/ QUOTE ]


Interesting analysys. Good work. We need more of this.


[/ QUOTE ]

[krumeluren]

[ QUOTE ]
i need more popcorn i guess...

[/ QUOTE ]

Haha maybe you do. Or maybe you just don't understand variance and how it affect the probability of finding extreme outliers in large datasets. No offense m8, but most people don't so your not alone.

[Lyrrad]

[ QUOTE ]
i need more popcorn i guess...

[ QUOTE ]
[ QUOTE ]
Results:
for 100 hand spans using 13400 hands the confidence interval (usually you use 90% or 95%, that means in 90% you will be below this win rate over 100 hands) is:
90%: 103.65bb/100
95%: 124.07bb/100
99%: 164.15bb/100
99.999%: 283.99bb/100
99.99999999999%: 480.77bb/100
or 1:10000000000000 or 1:10E14

[/ QUOTE ]


Interesting analysys. Good work. We need more of this.


[/ QUOTE ]

[/ QUOTE ]

Minor nitpick: These are statistics based on actual good, presumably tight, play. If you based the winrate based on the play of the players in question, I'm sure the average win-rate would be negative, but the STD deviation a lot, lot bigger. It wouldn't change the end result, but instead of 1 in a trillion, it might be 1 in a billion.

[ints]

Even if some individual hands look like that they were "honestly" played, it doesn't disprove the cheating. e.g. if you like to hammer men in their anus but then decide to sleep with a girl, that doesn't prove that you're not gay.

I'm getting the impression that most people here that don't know anything about computers imagine that the cheaters see the other players' hole cards on the table, graphically, next to their name as you are used to. But that's probably not the case. I imagine it something like this: there's a script (written by the bad guys) which parses the data on server -- filters out logs of their table, perhaps reformats it and sends you the information as text, maybe similar to HH, but just constantly updating. Of course you can reformat it or even build GUI (graphical user interface) on it, so it could actually paint very beautiful and colorful cards for you, but that'd be some more work for little gain and the villains might have been in a hurry.

Anyway, I remember that some weird delays were mentioned before the cheaters' preflop action. So, it could just have been that he's playing regular account and has a terminal window open, where the parsed info from the server is shown. So, for example in a donkament, he would have to check 9 players, which names have which hands from that window and then check the table for their positions on him and decide the action. That'd sure make some delays, even if he's not the 1st to act but let's say 5th to act and the first 4 just instafolded.

So, if there were couple of hands that were played like he was not cheating, then what's so weird about it?
* I find it much more bizarre that the cheater would so obivously cheat EVERY LAST goddamn hand. that's so moronic
* he might have just miss some hands, laziness
* he might have decided not to cheat some hands
* he might have misclicked
* he might have misread his 'terminal window' (from my example theory)
* ...

So, I don't see any argument in "he played that hand like he didn't know the opponents' cards".


Anyway, ffs, it's long established that there was some cheating going on. Why are we still debating like it wasn't 100% sure?

Apologies for my bad English.

[Trogdor!]

theres no way a hacker could be getting the hands without some sort of backdoor (otherwise why not just hack a bank or federal reserve or w/e else schemes that make you billions not thousands). Thats why it has to be an inside job or a flaw in AP's program or such

And im just trying to show that DD's winrate =/ (doesnot equal) 90/70 winrate at a highstakes game

if i could show that then there is 0 chance its not cheating as i am not factoring in the river af or suspect HH

[Trogdor!]

if we could compile all of DD's HH (bar chipdumping) into 1 database and get the SD there should be enough as the df would be about 3 or 4 which would be enough.

[Trogdor!]

krum the only reason we cant discount the 190 hands stats is because the sample size is only 2 (if we are counting via 100 hand samples) which is the lowest possible and has a huge rejection region and i dont have a SD but only a couple hundred more would be enough if it was a sustanined winrate (which it is if all of the pt screenshots are compiled together)

this is ofc not counting the river af and HH's which decrease the probability exponatially. The fact that we can even get close to rejecting the stats with 95% confidence with a high sd and very low sample size is ridiculous in itself seeing as 90/70 is not common at HS games which are uncommon themselves

it just keeps compounding and compounding until we get to a ridic prob like 10^20:1 as stated somewhere else

[ints]

[ QUOTE ]
theres no way a hacker could be getting the hands without some sort of backdoor (otherwise why not just hack a bank or federal reserve or w/e else schemes that make you billions not thousands). Thats why it has to be an inside job or a flaw in AP's program or such

[/ QUOTE ]
I didn't say that there wasn't any security flaws. Obviously they have really poor security there if the hand log data is available real time. They all should and I think most good sites do have security systems on par with online banking security.


Anyway, what's with the "only 190" hands crap? There are 4 screens of DD's stats in cliffnotes. My overwhelming maths skills say there are 565 NL hands in total, in case the DB-s aren't overlapped and altered. ~53.7 buyins won, 475 ptBB/100.
(Of course in the eyes of sceptics, this is standard winrate playing 90/70 against NL 10/20 and 15/30 pros. Anyone can go on a lucky 500-hand winning streak, taking 50 buyins.. obv..)

[facialabuse]

im sure this has been mentioned, and it certainly isn't a goundbreaking obseravtion, but isn't it relatviely likely this is far from an isolated incident-it's simply the first time the perpetrators were stupid enough not to balance out their results so as to appear as just dominant/very lucky players rather than obvious cheaters?

if everyone decides that the best reaction is the most thorough, and honest investigation possible, what happens if a comprehensive "zomg it was all rigged" outage goes down?

id play devil's advocate and say I wouldn't want to know, if it would disintegrate online gambling as it is