Party Poker Player Bias

[MyNameIsMud]

[ QUOTE ]
of an event like "hit a set or better with a pocket pair" should be roughly equal to the square root of the expected number of hits.

Based on your numbers, you should have had 197 hits with your pocket pair (calculation below), you got just 168, short by 29. sqrt(197) = 14, so you are 2 standard deviations short. Unlucky, yes (about 2% chance of doing this badly). Shows Party is out to get you? No.

(Also some other posters have suggested why big pairs might flop sets less often.)

Math:
If you have a pocket pair, what is the probability of hitting one or two of your cards? (This isn't quite the same as odds of a set or better, since it ignores trips on the board which may or may not be a good hand for you.)

probability of missing on 3 card flop = 48/50 * 47/49 * 46/48 = 48*47/50/49.

prob of hit = 1- prob miss = 11.8%

[/ QUOTE ]


This is exactly what I was looing for, thank you.

A 2% chance of geting this unlucky, (actually less since there are some valid points about big pairs)

Seems unikely, but possible. I will continue to play at Party for now, now I have a little better understanding of the math, and I want to see more hands.

I am also impressed by the low number of infantile responses.

[crazy canuck]

This is how you can calculate the probability of having your success rate for flopping sets:

I copied this from excel help manual:

BINOMDIST(number_s,trials,probability_s,cumulative )

Number_s is the number of successes in trials

i.e. the number of times you flopped sets

Trials is the number of independent trials.

i.e. the number of pocket pairs you saw the flop with

Probability_s is the probability of success on each trial.

i.e. the probability of flopping a set=11.8% (please somebody correct me if i'm wrong )

Cumulative is a logical value that determines the form of the function. If cumulative is TRUE, then BINOMDIST returns the cumulative distribution function, which is the probability that there are at most number_s successes;

i.e. Yes we want cumulative. This gives us the probaility of having such a bad or a worse succes rate. Note that calculating the exact probaility is meaningless.

So in the end:

BINOMDIST(161,1571,0.118,TRUE)=the probability of flopping 161 or less sets in 1571 trials = 3%.

Note that I removed the number of aces becasue you're often against another ace (can be true for kings but you can do the calculation).

Well, does this mean that party is rigged? Not really becasue you isolated events from the past. If you want to prove it is rigged based on the number of times you hit sets, you will have to make an assumption before you start the experiment.

[Michael Davis]

Romanians can't read. When they can, they move to France and write absurd dramas.

-Michael

[Thythe]

moondogg, I believe you are correct. It definitely does affect it. An easy way to see this is to assume that everyone at your table folds to a raise preflop and you always raise AA preflop. Here then, you will never get a set of aces which will make your stats look weird, even though its perfectly normal. The other poster is just being ignorant and the only thoughts he can come up with are "you are wrong"

[Flawed]

[ QUOTE ]
moondogg, I believe you are correct. It definitely does affect it. An easy way to see this is to assume that everyone at your table folds to a raise preflop and you always raise AA preflop. Here then, you will never get a set of aces which will make your stats look weird, even though its perfectly normal. The other poster is just being ignorant and the only thoughts he can come up with are "you are wrong"

[/ QUOTE ]

Thats not exactly what moon is talking about because if you see no flop the hand will not be counted. Your stats would be 0 flops and 0 sets

moon explained it pretty well, If you always raise AA preflop on average you will have less outs to hit your set

[MicroBob]

both ideas apply to AA.

if you raise it and induce everyone to fold then you are going to have some hands where you never saw the flop in the first place. these are also more likely to be the hands where you would have flopped an A....because the fact that everyone folded indicates a slightly higher possibility that nobody was holding one.


with the 22 hand.....the more callers you have, the better your chances of spiking a deuce on the flop.
each caller or raiser indicates a slightly greater chance that your coveted deuce is remaining in the pack. most hands that contain a deuce find the muck.

if you call in EP with 22....and get no callers and only a check from the BB, then i submit that your chances of hitting trips are slightly reduced because the series of folds from the rest of the table is indicative of an ever-so slightly greater chance that they were holding one of the other two deuces.


i don't believe that these small changes are significant enough where they should change one's approach to the game though.
i would be interested to see if those with large p-tracker databases have flopped trips with their low-pairs more often as the field of callers increases.

[Flawed]

[ QUOTE ]
both ideas apply to AA.

if you raise it and induce everyone to fold then you are going to have some hands where you never saw the flop in the first place. these are also more likely to be the hands where you would have flopped an A....because the fact that everyone folded indicates a slightly higher possibility that nobody was holding one.

[/ QUOTE ]

I wouldnt consider this to be 2 seperate ideas

[ QUOTE ]

with the 22 hand.....the more callers you have, the better your chances of spiking a deuce on the flop.
each caller or raiser indicates a slightly greater chance that your coveted deuce is remaining in the pack. most hands that contain a deuce find the muck.

if you call in EP with 22....and get no callers and only a check from the BB, then i submit that your chances of hitting trips are slightly reduced because the series of folds from the rest of the table is indicative of an ever-so slightly greater chance that they were holding one of the other two deuces.


i don't believe that these small changes are significant enough where they should change one's approach to the game though.
i would be interested to see if those with large p-tracker databases have flopped trips with their low-pairs more often as the field of callers increases.

[/ QUOTE ]

This is true but it wouldnt have an affect on your long term stats of determining whether you're flopping the correct number of sets.

[MicroBob]

[ QUOTE ]
This is true but it wouldnt have an affect on your long term stats of determining whether you're flopping the correct number of sets.

[/ QUOTE ]

i agree that the effect might be barely more than negligible at the outside.

if someone had 1-million hands in various databases....that MIGHT be enough to start to see an effect.

1-mil divided by 220 = 4545
meaning you should get 22 or 33 or 44 about 4500x each after 1-mil hands (and this is assuming you get to play ALL of your low PP's which is extremely unrealistic).

so you should hit your set about 400x or so if you actually are crazy enough to play each PP EVERY TIME....and, even then, missing this total by +/- 50 is probably not too many SD's off i suspect.

[Flawed]

[ QUOTE ]
[ QUOTE ]
This is true but it wouldnt have an affect on your long term stats of determining whether you're flopping the correct number of sets.

[/ QUOTE ]

i agree that the effect might be barely more than negligible at the outside.

if someone had 1-million hands in various databases....that MIGHT be enough to start to see an effect.

1-mil divided by 220 = 4545
meaning you should get 22 or 33 or 44 about 4500x each after 1-mil hands (and this is assuming you get to play ALL of your low PP's which is extremely unrealistic).

so you should hit your set about 400x or so if you actually are crazy enough to play each PP EVERY TIME....and, even then, missing this total by +/- 50 is probably not too many SD's off i suspect.

[/ QUOTE ]

he's counting hands whether he sees the flop or not as long as there is a flop. so if he folds his 2's or calls because theres a lot of people in the pot wont have any effect/affect on his stats he cant control how many people limp in. Now if he always raises with 22 than he should hit his set more often than normal.

[MyNameIsMud]

Well maybe party is not rigged, then I must have worse luck than over 99% of party poker players.

I will keep posting blocks of 5K hands until I resolve this. I play 3 or 4 tables at Party (200-350 hands/hr) so It doesnt take too long when I set my mind to it.

I know no one believes its remotely likely that Party would bias the deck against certain players, and I am in agreement, but I can not continue if this trend doesn't end eventually, I am barely making wages. I truly hope I come out of this looking like an ass.


I have been playing at party for a long time, and for the first time I have deep suspicions.

The next 5K which I began when I started analyzing the first 33K

4991 hands:


AK pair or better on flop
8/39 20.5%

AQ pair or better on flop
12/53 22.6%

KK, at least 1 A on flop
5/12 41.6%

Big pair sets:

AA - 1/8
KK - 1/12
QQ - 2/17
JJ - 2/17

6/52

TT - 0/21
99 - 2/21
88 - 2/15
77 - 2/26
66 - 2/13
55 - 3/17
44 - 1/16
33 - 1/23
22 - 0/17
---------
13/169 = 7.7%

Total 19/221 = 8.59%

-MyNameIsMud