Official Full Tilt Poker Response to Bot Thread

[ClubChamp04]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I don't think the numbers are that sensational.


[/ QUOTE ]

Can only people that understand statistics try to make claims about whether the numbers comparison is meaningful or not? It's embarrassing.

[/ QUOTE ]

It doesn't matter if I say how much statistics I've studied in my life (none or some or much), you will still say the same thing.

I don't think it's sensational for several people to play a simple strategy and get that close numbers.

Adde

[/ QUOTE ]

How do you have any idea what you're talking about if you don't undertsand statistics? Seems like an ignorant thing to say they aren't that sensational, when you might not have the slightest idea what sensational means in terms of the data. Have you ever seen 4 players stats this closely related over a huge sample like this? How about looking through your database and trying to find 4 players at any level you have ever played with stats that look like this and then report back to us? Back up what you're saying if these numbers aren't that "sensational".

[teddyFBI]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I don't think the numbers are that sensational.


[/ QUOTE ]

Can only people that understand statistics try to make claims about whether the numbers comparison is meaningful or not? It's embarrassing.

[/ QUOTE ]

It doesn't matter if I say how much statistics I've studied in my life (none or some or much), you will still say the same thing.

I don't think it's sensational for several people to play a simple strategy and get that close numbers.

Adde

[/ QUOTE ]

It absolutely does matter, because we can test for statistically significant difference based on the data given and sample size. It's not like I just looked at the numbers and said "oh boy those look close enough, I guess they must be bots!" Apparently your judgment, however, was doing just that, looking at the numbers and coming to a conclusion that has absolutely no support behind it.

The chances 4 people could play on their own and get numbers that close, over that large of a sample is close to impossible.

[/ QUOTE ]

just a silly though that came to me -- IF, in the sweatshop there was NOT 1 account uniquely assigned to each individual sweatshopper...e.g. sometimes player #1 played on account #1, but not always and sometimes he played on #2 and #3, and all 3 sweatshoppers did this (let's say just hypothetically it was random who played on which account so that at the end of the day, after 100K hands played on each account, each of the 3 users was responsible for approx 1/3 of that), then might that not be one way that we would expect to see alarming conformity between the 3 accounts' statistics? e.g. it would theoretically be possible for the 3 sweatshoppers to use "the system" in markedly different ways (yes, making different decisions given the same scenario), but as long as they were randomly assigned an account to play on, over the long run, the statistics would approach uniformity.
(This relies on the assumption that the same individual never deviated from his OWN use of "The System"...see >> if each of the 3 people applied "The System" slightly differently, but never deviated from HIS OWN application thereof, then statistics for the 3 accounts would gradually approach uniformity)

p.s. i still think something shadier was going on here...just thinking about hypothetical explanations.

[neverforgetlol]

A person is just not going to play the exact same way as another person over that many hands, as a system does not account for all factors, especially in NL due to the practically infinite bet sizings.

[Bobo Fett]

[ QUOTE ]
A person is just not going to play the exact same way as another person over that many hands, as a system does not account for all factors, especially in NL due to the practically infinite bet sizings.

[/ QUOTE ]
With Teddy's theory, which someone else actually came up with as well several hundred posts ago, all that would be required would be for each player to play consistently themselves...oh, and for each account to be divided up fairly evenly between the three of them. Far-fetched, perhaps, but possible.

[neverforgetlol]

[ QUOTE ]
[ QUOTE ]
A person is just not going to play the exact same way as another person over that many hands, as a system does not account for all factors, especially in NL due to the practically infinite bet sizings.

[/ QUOTE ]
With Teddy's theory, which someone else actually came up with as well several hundred posts ago, all that would be required would be for each player to play consistently themselves...oh, and for each account to be divided up fairly evenly between the three of them. Far-fetched, perhaps, but possible.

[/ QUOTE ]

I really don't think so, they could split up but I don't see why that would even matter. They are still going to have different stats, because even if you play a certain system, that system will not account for every possible situation in the game. It is those situations where they will deviate.

[cardcounter0]

[ QUOTE ]
[ QUOTE ]
A person is just not going to play the exact same way as another person over that many hands, as a system does not account for all factors, especially in NL due to the practically infinite bet sizings.

[/ QUOTE ]
With Teddy's theory, which someone else actually came up with as well several hundred posts ago, all that would be required would be for each player to play consistently themselves...oh, and for each account to be divided up fairly evenly between the three of them. Far-fetched, perhaps, but possible.

[/ QUOTE ]
Uhhh... NO. I think it has been shown that even if you break up an individual player's stats into four 100,000 hand chunks -- they won't match this closely.
So these 3 or 4 individuals were able to play more consistently than even a single person could do.

[Bobo Fett]

Meh...like I said, far-fetched. I'm almost beyond caring...FT has done what they're going to (or not do), everyone's fairly entrenched in their positions...this is a dying thread IMO.

[Adde]

This has turned into a hate feast just because people disagree with each other. No need for hate.

I haven't seen any extensive calculations on the stats, only rather loose formulas and judgments by people who may or may not be skilled enough in statistics, supporting both sides. I myself am not skilled enough to do this today, but I've got enough statistic classes in my masters exam to know that what looks easy at hand can be very complicated to prove when dealing with statistics.

I've introduced a couple of friends to poker over the years. They've read the same books as I have, and without having a shared "system", we all are 14/9-ish players. I can't imagine it would be any difficulties for us to follow a simple strategy and get very similar post flop stats over a substantial amount of hands.

Add to that the many circumstances that, because the sheer amount of them, lead people to cry wolf. Problem is that some of these are plain wrong, but they start to live their own life after being repeated again and again.

I don't know nlnut at all, but from his previous script requests to me and others in the Software forum earlier this year, there is no way possible that he is capable of developing a bot. Still, people only see the words "mouse clicker" and "script" and are all too soon to judge this as bot evidence.

People with academic experience should not be that sloppy with source control.

I could be wrong, but until proven otherwise, I see more signs pointing to fair play than foul play.

Adde

[RunDownHouse]

teddy,

As someone said earlier, one would think its virtually impossible for just one person to play with that sort of consistency over 100k+ hands. If you accept that, you would surely accept its even more unlikely for the play of three people to converge in a similar fashion, right?

[neverforgetlol]

[ QUOTE ]


I haven't seen any extensive calculations on the stats, only rather loose formulas and judgments by people who may or may not be skilled enough in statistics, supporting both sides. I myself am not skilled enough to do this today, but I've got enough statistic classes in my masters exam to know that what looks easy at hand can be very complicated to prove when dealing with statistics.

[/ QUOTE ]

I have not seen anyone with decent stat credentials show evidence concluding that these are not bots.

[ QUOTE ]
I've introduced a couple of friends to poker over the years. They've read the same books as I have, and without having a shared "system", we all are 14/9-ish players. I can't imagine it would be any difficulties for us to follow a simple strategy and get very similar post flop stats over a substantial amount of hands.

[/ QUOTE ]

The problem is you do not understand statistical significance. The more hands you play, the smaller of a difference statistical tests can detect between your stats. That means with a large sample as presented in this thread, the numbers given would have to be increasingly close together to conclude that they are using bots. And that is exactly what we found.