Stars Sit & Go poll

[BigBiceps]

[ QUOTE ]
[ QUOTE ]
This gives the indication that something is amiss....

[/ QUOTE ]
Downtown's Handy Response Form v.001
------------------------------------
In response to your post, as a poster on this forum I would like to advise you of the following:

[x] Read the FAQ
[x] Use the Search Function
[x] It's Just Variance
[x] Your Sample Size is Too Small
[ ] This is a Bad Beat Post!
[x] You Might Suck
[ ] Party is Rigged
[x] Pwn3d!
[ ] Push
[ ] Fold Preflop
[x] It's Not Close

Note: Please find above what is amiss.

[/ QUOTE ]

You checked too many boxes.

[downtown]

[ QUOTE ]
You checked too many boxes.

[/ QUOTE ]
It's an art, not a science.

[BigBiceps]

[ QUOTE ]
http://poker.wikicities.com/wiki/Sng...2Fsample_sizes

If you want to mail us (PokerStars) with your concerns, let me know your User ID, and I'll do the response.

[/ QUOTE ]

Although the sample size IS significant the method of data collection is not unbiased, so I cannot reliably draw any conclusions, only an indication.

It is inconclusive if it is rigged anyway. Furthermore, it is not illegal to rig the games in the first place, so why would I email support?

I realize the risks I take when playing online and I just try to make the best decisions I can given the information I have.

[Sciolist]

Because if people think our site is rigged, it potentially hurts our business. Poker sites trade on their good names. 30 SNG is not a statistically significant sample. The ~1-1.5k hands are not statistically significant for anything, except perhaps to determine if the suits are equally likely to appear.

[FlFishOn]

"You cannot generalize what sample size is significant, it depends on the data collected. "

Don't bother trying to explain this concept here. There are none but deaf ears.

[splashpot]

[ QUOTE ]
Ok ... I did statistics for clinical trials as my job for over 5 years.

Do you know what a significance test is?

You cannot generalize what sample size is significant, it depends on the data collected.

[/ QUOTE ]
It serioulsy confuses me when I see people who supposedly have statistics background, yet cannot understand why this is such a pitiful sample. I know how to do a significance test, but in this case it is a waste of my time because this example is so blatently obvious.

Did you even read the link I gave you above? In that experiment, there were over 3,000 people out of 41,000 who won money over those 500 games. Yet you think 30 GAMES is enough.

Whatever. You're not going to listen anyways. I'm done argueing.

[FlFishOn]

"30 SNG is not a statistically significant sample. The ~1-1.5k hands are not statistically significant for anything, except perhaps to determine if the suits are equally likely to appear."

Spoken from true ignorance of statistical methodology.

[BigBiceps]

It doesn't matter if it is 30 games or 1 game. That is where you are confused.

Take the example of 1000 people each flip a coin and guess the outcome. Therefore the number of "games" for each person is 1.

If 700 people lose the game and only 300 win ... the game is biased, because this is a significant result.

Do you understand that it is not the number of games that determines significance -- it is the number of people?

[OldLearner]

[ QUOTE ]
I stopped playing after about 50, because I can't make as much at them as I can in live play

[/ QUOTE ]

So you can win in the live (absolute softest) games but contend that online (much tougher) games may be rigged because you make more in the live (absolute softest) games.

And then you criticize others because they won't consider the possibility that online is rigged.

This may negatively affect your popularity in these forums.

[BigBiceps]

Sorry, I am not trying to hurt your business.

I am not affiliated with any poker site, and I play on all the major sites.

I probably just am a victim of paranoia and variance (because I have similar worries at other sites.) However, I cannot trust anyone blindly.