Big question about Paradise

Everyone knows what it is. Is the game legit? Do the cards come honestly? We can all agree that collusion exists and is a drain on any honest player but are we getting a square deal from the house? We will probably never have the full scoop but I'll toss out my most recent investigation and you can make me laugh with your mindless defense of these criminals in Costa Rica.


I've been winning at the $11 tournaments for a few weeks now. I love the action, a great rush. The money is small but I will still stoop to pick up a C note twice a week. All my data comes from these games.


Boiled down, I've suryeved another ~2000 hands and there are too many suited starters dealt. I've suspected this previously and surveyed this same statistic 4 times previously. Only once did I get a number that fell on mathematical expectation, all others were above expectation. The most recent survey had suited starters 2.2 standard deviations above expectation. This should alarm anyone that understands what that means. If you don't understand that please don't argue any numbers here.


To me, I don't care. I'm beating the dink tournaments and will NEVER EVER play a ring game or the big tournaments. If I bust out I'll just wait for another buy-in bonus and play a freeroll.


Those of you with bigger money in play should be very concerned. If Paradise is willing to screw with the distribution of starting hands then all other methods of sheparding the games come up for discussion. Would they also push the odd pot from a winner to a new player or a known pigeon? Would they pump the rake in some fashion? All manor of cheating is now possible if you believe they are buggering the deal. Fox Moulder wants to believe, I believe.


I really want someone to do the data collection. Then we can have a cabal of two.

"To me, I don't care. I'm beating the dink tournaments and will NEVER EVER play a ring game or the big tournaments"


Actually you have just given a pretty strong argument for the integrity of Paradise Poker. It goes something like this.


1. I have been getting better hands than expected over a fair period of time.


2. I am beating the game pretty good.


Well, what did you expect? When you receive a better than average run of cards over a prolonged period of time your results are suppose to be very good, and that is exactly what is happening to you. If on the other hand you were getting this good run of cards and were not doing well then that might be an argument that something is wrong because it would imply that everyone was probably also getting better than expected cards. But that's not what's happening to you.


If the game was not legit, you should lose when you run good. However, in on the square games, when you run good you expect to do well. Thanks for pointing this out to all of us.

Out of 25,210 starting hands that I've been dealt at Paradise that I've kept track of, I've gotten suited cards in the range of 58 times for 72s to 97 times for K3s, with the majority of suited starters being dealt 70-80 times.


Statistically, the results should be:


114 times for each pair

76 times for each suited starter

228 times for each unsuited starter


The number of pairs and unsuited starters I got dealt were all around the numbers above. I'm sure every other person who's kept a database of their hand histories will get these results, assuming their data is large enough.


How large is your sample set?


- Tony

"Well, what did you expect? When you receive a better than average run of cards over a prolonged period of time your results are suppose to be very good, "


This is an argument with little merit in this specific case. If I'm alone in getting dealt more suited starters then we are talking about me playing perhaps 15 additional hands over the course of ~40 tournaments. I don't play 72s outside the BB. The additional value to me would be very small. I'm beating the game (probably 250 tournaments so far) not on starting hand quality but on other factors unique to my ability.


Mason, I know you understand statistics. If I sample these hands again and the result is again a few SDs above expected what then? I already know the answer. Nobody will give it any credability. This should be of at least some minor concern for you as the integrety of any advertiser reflects on the medium in which it advertizes.


"Well, what did you expect? "


That's easy. I expect a sample at some time to come in under 12/51 suited starters and so far that hasn't happened.

Doug,


The starters do NOT have too many suited or anything else too many:


Paradise posted first 5 million and then 25 million hands. The data analysis done by me and others about a year ago shows that the cards are random. Please read the archives, there is a lot of data, done by people with advanced Math and CS degrees (me and others) and then a lot of whiners who cannot believe they should lose who complained about the analysis.


See my post above in a few minutes for a reason other than colluders that Paradise is harder than other places to play.


Mark

Still a traitor and a stooge for the illegal gaming community, I see. Have you looked at Mr.Duke's profile? He's your kind of guy.

Come on! If you can't trust me who can you trust? Even the GCA said that I was honest -- "as pure as ivory soap" was the way they put it.

"Paradise posted first 5 million and then 25 million hands. The data analysis done by me and others about a year ago shows that the cards are random."


So you were able to prove that the hands Paradise gave you to digest were in fact random. BFD. I'd have bet my bankroll they would be random. They would have to be brian dead to provide nonrandom hands for you to inspect.


My hands were all dealt to me and they could be random. If they are then they are 2.2 SDs away from the expected mean. This makes randomness suspect.

My guess is your sample size is too small, you made a mistake, or you just got a wacky distribution due to pure chance.


You mentioned "three out of four". Would you mind,

for each of the 4 samples, posting N_tot, N_s, and N_o, for the total number, suited and unsuited? I would like to dupiclate your 2.2 SD result.


zooey

The formal numbers from surveys 1,2 and 3 are not at hand. They were done a while ago. Survey 4 is as follows: 1886 hands, 483 suited, 1403 unsuited.


You quote a database of 40K hands. Please tell me which limits you play.

This is at .50/1 limit.

I agree, your sample is skewed towards suited hands by over 2 SD's. While that is unlikely, it is not unthinkably improbable. I think you need a larger sample size.


Silly question: I assume you are only counting your hands, and not every hand shown in showdowns? The natural tendancy to stay with draws would skew the sample set that way. (I am not trying to be condescending, just a quick trouble shooting question.)


My limits: (why do you ask?)

.50 -1 : 2k hands

1-2 : 21k

2-4 : 13k

3-6 : 6k

5-10: 3k


Im not in the mood to count up the suitedness for each limit right now, but I will for a given limit if you're curious.


Best wishes,


zooey

I'm iterested in reproducing my collection of data an it all comes from $11 tournaments. Ring game data could very well be different.


I'm counting only the hands dealt to me. In my haste to count them it's possible I overlook an occasional suited hand but most unlikely that I would include one that wasn't there. Is it possible that PP sends 100 +/- hands when you request 100 hands? I didn't count the total in each e-mail.

30 is alarge enough sample size. This is hocus pocus.

I can't believe that you consider 2000 hands even close to a proper sample?


Remember that true ramdomness means that a the same hand may be dealt one million times in a row! (the percentage chance of dealing 72s is the same each time the deck is dealt). When you shuffle manually it is virtually impossible to get a true random deal. The result is therefore likely to be LESS uniform with a computer generated hand until you have a large enough sample (I play about 2000 hands per week so I would suggest stats for at least a year.)


At any rate, I see no logical connection to your stats and whether or not PP is straight or rigged.