Two Plus Two Publishing LLC
Two Plus Two Publishing LLC

Go Back   Two Plus Two Poker Forums > >


Probability Discussions of probability theory

Thread Tools Display Modes
Old 03-10-2021, 03:18 AM   #1
Join Date: Jun 2014
Posts: 36
How many flips (25-80%) are needed to prove the water is not clean at a site?

I want to run a stream documenting a flagged to lose account (mine) at a lowlife site, as I have told their support as much, and I intend to keep my word.

Easiest way I realized without losing any substantial amount is 25 cent spins. I'll play to engage in the most flips I can but trying to win not like an idiot.

100k flips? (I'll almost always be ahead, rarely behind, coolers ignored, just all ins filtered through holdem manager because I don't think poker tracker can filter all ins only tourneys).

I mean if you flip 100k coins the ev will match, if at the end of this I am way under ev it has to be screwy right?
rexchex is offline   Reply With Quote
Old 03-24-2021, 10:03 PM   #2
Join Date: Aug 2020
Posts: 115
Re: How many flips (25-80%) are needed to prove the water is not clean at a site?

Calculate a z score for your results. If you are truly doing flips, you expect to win 50%, or 50,000 of them. Subtract the actual number of wins from 50,000 (you can use an absolute value if negative). Divide this difference by the standard deviation, which can be approximated by sqrt (100000 x 0.25) = 158.11. The result is a z score. You can get the probability of a z score this high or higher from tabulated values of the normal distribution, or from the normsdist() function on Excel. Approximately, a z score of >2 has about 5% probability, >3 about 1%. If your getting z scores that high, it’s evidence that the site is non random.

If your win probability is not 50%, the z score calcuation is modified in two ways. First your expected number of wins is not 50000, but 100000 x p, where p is the actual win probability. This is the value you subtract the observed wins from. The standard deviation is also modified to sqrt(100000 x p x (1-p) ), but the basic idea is unchanged. A high score still indicates evidence of non randomness
stremba70 is offline   Reply With Quote


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

All times are GMT -4. The time now is 08:20 PM.

Powered by vBulletin®
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.
Copyright 2008-2020, Two Plus Two Interactive