Theories on how online poker is rigged?

[hackmage]

I'm working on a project for my Stats class, and I decided I was going to take a look at those claims of online sites being rigged. Personally I don't believe they are, the casinos have almost nothing to gain by rigging the sites, they win whether you win or lose.

The problem is finding a hypothesis that's fairly testable, so I was looking for the most popular theories the conspiracy theorists have. The other problem is that I can't keep track of the beginning hands because too many will be mucked. Another unknown is how long a player has been playing on the site, which seems to be a leading theory, the cards always fall for the newbs.

So my only thought is to record flops, and see if there is a substantial amount of low cards. That is when compared, to a completely random generation of flops using Minitab. The idea behind this is that beginning players play too many hands, and thus to help them out sites would give them more low cards on the flop, thus increasing their chances to flop a pair. This won't take into account whether they win or not, but just that new players maybe enticed to keep playing those low cards, thus reinforcing bad play. That's the only theory I could think of that would actually be testable. Any input would be greatly appreciated.

Matt

[pshreck]

The only logical theory is that sites 'juice' the preflop hands and flops to encourage more betting and therefore more rake. They could do this in SNG's too, to ensure it ends quicker than it should (they would make more money per hour).

While just about all intelligent online players think this is highly unlikely, I think this would be where you should start with doing some tests.

Check how often multiple players are getting high PP (JJ and up) at the same time preflop, vs how often it should happen mathematically. Think in this manner and you could do some good testing.... however I think it will be a waste of your time and you will just confirm the random number generators are working fine.

[lorinda]

FAQ

The "Rigged" part of this might start you off, there are links to hole cards, past cheats and finishing hands in there.

Lori

[hackmage]

Yea, I'd love to test the PP idea, but once again it comes down to a muck problem, I won't always be able to see starting hands. Seriously, I have NO DOUBT this will be a waste of my time, in fact I'm 99% certain I'll just show the RNGs are working right, but that's the idea of statistics. I want to see just how "right" I am. And thanks both of you for the replies.

Matt

[IggyWH]

Been a little while since I read Lori's FAQ so I don't know if it's in there...

I do remember someone creating a program that checks starting hands and compares them to true odds to compare whether you had a good/bad session. Might want to try something along those lines...

[lorinda]

I think that's Eastbay's own personal program, don't think there is a released version or anything.

Lori

[hackmage]

Thank you much, hadn't noticed this post. The nice thing is I now I have some data to compare to my own (we have to get our own data for this project).

Matt

[IggyWH]

[ QUOTE ]
I think that's Eastbay's own personal program, don't think there is a released version or anything.

Lori

[/ QUOTE ]

I downloaded something on some post but couldn't get it to work because it was missing a DLL. I then said eff it then. I'll dig through my IE history to see if I can find the thread because search hasn't done me any good.

[IggyWH]

[ QUOTE ]
Thank you much, hadn't noticed this post. The nice thing is I now I have some data to compare to my own (we have to get our own data for this project).

Matt

[/ QUOTE ]

I remembered something about the program. He said he used the #'s that were published on pokerroom.com site. Here's a link to them if you want to use their data :

http://www.pokerroom.com/games/cards...ocketcards.zip

[stigmata]

I thought a more interesting hypothesis to test would be the idea of online collusion.

E.g. players working in teams. I think this has a far greater chance of turning up something interesting than checking if the sites are cheating (which they clearly arent).

However this is not as easily achieved....

I guess the way I would do this is:

1. Generate a large pokertracker database. This could be achieved in two ways: A) Ask nicely on this forum and combine peoples databases B) Watch multiple tables on Party, and import Hand Histories to POke Tracker before you close down.

2. Query database to find multiple instances of players playing at the same time. This limits you to players who are either Friends or Colluders (or both).

3. Test playing strategy for two things A) Trapping players between the two hands. E.g. The two (or more) players show significantly more aggression when in a pot togethor than would be expected. B) Best hand plays e.g. The two players only play the best hand between them.

I guess the problem with my suggestions are that it is a far bigger project and is not as statistical (more of a database/programming) exercise. So not much use to you, although it does have a statistical element....