Tilting at windmills and shouting at clouds. Oh what a life the average riggie lives

And that's why they're used. It's a simple situation where you either win or lose, and you have the same chance of either one happening. It's an easy example to use when explaining more difficult concepts. It's only a problem if they're losing more than 25, 20, or 10% of the time. No one so far has shown that that's happening. I disagree, but let's assume you're right: Why has no one shown anything is amiss?
Not to beat a dead horse, but you're wrong. Why don't you go ask in the Probability forum if NewOldGuy is "stupid" and wrong or not? What if you run over EV? Should you email them to let them know that you have to lose more often?

How would we draw an EV line for flipping a coin?

Aside from that, your post is hilarious because the silly little points you're attempting to make about variance "evening out" are actually arguing the opposite.

You also have to keep in mind that it takes quite a few trials just to determine if a coin is fair or not, around 10,000 flips just to be ~95% sure the coin is fair. And that's checking for something with two equally likely outcomes. Meanwhile we have people telling "shills" to "get real" for saying that 10,000 hands aren't enough to check if you lose too many 80/20s.

Lol of course they wouldn't because they are all super cool players who KNOW that they should be crushing mere mortals on the poker sites. They claim that they can win on the penny tables but as soon as they move up a level or 2 then they lose and this is because of the rig NOT because they are terrible players!

It's like joining a tennis club and after 6 months of playing against 12 year olds or retired grandmothers and becoming a mediocre player thinking that you should now be good enough to go on and win Wimbledon and when you don't win ... well its obvious right?

The EV line thing is not intuitive. Correct me if I am wrong but the difference lies in the fact that what will even out towards expectation over time is the number of wins and losses. The EV line deals with dollars won and lost not hands won and lost and there is no reason to expect the dollars to even out over time. If you lose a big hand as an 80% favourite you will always be out those $$ unless by pure chance you win a 20% for the exact same amount but there is no reason to belive that the dollar amounts of your wins and losses will even out.

Have I understood it correctly?

No. You can use any random variable, and the variance scales directly with the sample size, and the standard deviation scales with the square root of the sample size. The confusion comes from the law of large numbers, which tells us that the proportion of the deviation over the total gets smaller as trials grow. But the actual difference gets bigger, and the proportion shrinking is just a result of the denominator (trials) growing faster than the numerator (deviation) because it's a square root relationship. They both grow.

You can look at standard deviation being something like the average distance you expect to be from the mean (approximately true, not exactly). If in 100 hands the SD is 50 (often we use bb/100 in poker), then after 10,000 hundred-hand buckets the SD is 5000, or 100x bigger. And after a million 100-hand buckets it's 50,000 or 1000x bigger.

Two important things to realize. First, past events don't affect the future of random variables. If our result is $100 under EV today (or bb or wins or anything) then our expectation at any future time from now is to still be $100 under EV. Forever. The cards don't know you started out behind. Second, in any player pool the average distance from EV will get bigger as the average sample size gets bigger. Variance does not get smaller with more trials. This is a very simple mathematical concept. Variance is the average squared distance from the mean, obtained by adding up all the squared deltas from every trial. A squared number is always positive. Adding more positive numbers to a quantity will always make it bigger.

I'm not going to give a more detailed math lesson in this thread, in fact I'm going to unsubscribe.

See you tomorrow, Same time, Same channel

And this (if you think about it) explains why the riggies think the difference gets smaller. As the number of samples goes up, so does the dependent variable. The scale increases with the dependent variable but the difference only increases with the root of the dependent variable. And that's where the confusion come in for people who have poor math. And yet it's really very simple.

I wish you wouldn't. I enjoy reading your posts and really appreciate the time you took on the above post as it definitely improved my understanding of the subjects you discuss. I'd also love to see a retort from one of the riggies to what you've said here. It'll probably be something along the line of 'Math? Ha! I don't need no stinking maths!'

lol. Are you that teenager from south park?

Seconded.

No doubt some nonsense about the shills bullying them with book smarts.

I think he's the teenager from Southpark's dumb younger brother that the family are so embarrassed about.

There is a certain argument coming from the shill-turds asserting that online poker rooms would not rig the deal because it would deplete their customer base as each player came to realize he or she was being scammed. I believe this fails to take into account the PT Barnum aspect of human nature. And after a careful cost-benefit analysis the poker rooms have decided that, because of the strong Barnum effect, profits will always surpass desertions.

Exhibit A: http://www.youtube.com/watch?v=r89tSAxAu0c

In fact, so firmly is the Barnum effect engrained in the human genome, I propose you could open an online poker site called the PT Barnum Lightly Regulated Cardroom and Rigged Deal Emporium, and still attract a reasonable size customer base, you know, folks like Exhibit A. True, you’d have to do a little creative marketing: like saying it’s named after PT Barnum because it’s so much fun to play there you’ll feel like you’re at a real circus! And that they care so much about their customers they actually rig the deal IN YOUR FAVOR to enrich your overall carnival experience.

The real life examples are already there in Absolute and Ultimate Bet. They say, “Yeah, we had a few bad apples, but we got rid of them and everything is peachy now.” Well, maybe. Allow your gambling commission full access to your software and books, then I’ll consider the possibility you’re telling the truth. But for me, as it stands, this is the poker banter equivalent of Phil Galfond telling you: “Yeah, I was bluffing last time, but this time I really got it!”

I'm guessing there are a few distinct concepts that are getting mixed up here. Which stats is it that we correctly say will approach closer to expectation the bigger the sample?

When spadebidder did his study, he found the results very close to expectation over the large sample.

Thanks for posting that link to you telling us what you think is going on. I had no idea you were so articulate. Impressive!

First off, if it's that easy, I suggest you try it yourself. Getting a successful poker room up and running is no easy task, if it were we'd all be idiots not to each have one ourselves, and giving it an idiotic name like the one you suggest (and note, I'm not calling you an idiot here) would pretty much doom it to failure at birth.
Second, which exact 'careful study' were you referring to in your post? Did you have access to it? Has someone you can swear by told you they've seen it? Or did you simply imagine it up for this example, knowing that blatantlyrude would chime in and announce that everyone knows of its existence?
Third, any true risk-analysis would take into account that the 'rule' you're referring to is actually simply a cliche and that there's far more money to be made without a rigged RNG than with one. Period.
Fourth, the UB and AP examples keep getting brought up. How are they doing these days? Playing there? Cashout recently? Know anyone that has? Any idea who it was who did the vast majority of the work it took to uncover what was being done there? Are you aware that one of those very people posts here occasionally and is not on the 'rig' side? Any ideas why someone willing to work like hell to uncover unseemly behavior by online sites (and thus not likely a shill or a 'site-defender') would not be swayed by the insanely powerful riggie arguments on this thread?
Seriously. You've gone on and on arguing that variance is something it isn't and does things it doesn't do; and when newoldguy posts an irrefutable explanation of what variance (among other things) actually is and how it behaves, you just jump over to some other emotionally-based notion without even acknowledging that you might have been wrong on that one. Come on...

The proportion gets "closer" not the actual difference, and "smaller" is a better word here than "closer".

Here's a simple example:

We flip a coin 100 times and we bet on heads. We expect to have 50 heads and break even. Let's say our actual result is one standard deviation from the mean, which as I mentioned before, is approximately the same as the "average absolute deviation" from the mean, i.e. how far away the average result is ignoring sign. In this case that distance is 5. And 5/50 is 10% off.

Now let's flip 1000 times (starting over). We expect to have 500 heads. If we are off by one standard deviation, that distance is 16. It goes up by the square root of 10, since we increased the sample 10x. And now 16/500 is about 3%.

So for the small sample we were off by 5 heads, and for the larger one we were off by 16 heads. And if we keep going that number keeps growing. And by the way the "variance" is just the square of that number, so it grows a lot faster.

But notice that the proportion to our total went from 10% to 3%. It didn't get closer it got farther, but we are comparing to a much bigger denominator now. And if we concatenate our two results above and have 1100 flips, of course the same properties hold true. But what may not be so obvious (to riggies) is that there is no relationship between the first 100 and the next 1000. If the former is off in one direction, the latter can certainly be off in the same direction. Half the time it will be, because it's random. There is no cancelling out effect or "getting even" effect at all. Variance does not "even out".

Edit: fixed an error in math.

Riggies can't acknowledge they were wrong, that implies weakness. Besides, it also implies that other things they've said could be wrong, and that's just impossible when we're talking about the greatest math and poker minds of our generation.

Thank you! Very helpful!

To make that much of an effort you must be a trained shill trying to deflect attention from the truth. I am going to believe this guy:

He's such a charmer, how could you not?

Bingo:

Step 1: Call B******ed an idiot.
Step 2: Unsubscribe
Step 3: Become a self-loathing xenophobic racist
Step 4: All the real world pussy you can handle.

Welcome to step 2 Newoldguy.

Even after the UB scandal a sizable number of players chose to stay there and be scammed.
.
This is a blurb from a company who develops and distributes poker software as you can see they openly admit the revenue collecting aspect can be altered,
.
No company can guarantee earnings that are dependent upon many factors. The general rule of thumb is, the more money you put toward marketing of your own online casino and multi-player poker room online gaming business, the higher the revenues, the higher the net income. You can expect to keep between 75% and 100% of the gross monthly deposits depending on the turnkey casino, multi-player poker or sports book package you invest in. Stand-alone poker software solution is available.
.
Problem gamblers would read this ignore it and continue to gamble.

So you are admitting that higher-ups at these poker sites do pay attention to what goes on in this forum? Hmmm. That goes against what you shills were saying earlier and your delusions of grandeur argument.

That's certainly one way to see things.

Another way to see it is Stars' head of security posts fairly frequently in this thread.