rigged is deliberate, being flawed just means there is a defect in some way.

Yes but on this particular forum the terms are used synonymously because either fault would tend to vex players.

Admittedly an accidental flaw is likely to advantage or disadvantage players equally but if you spotted a genuine flaw it's certainly likely to ruin your concentration.

Ah right, well I guess that distinction makes sense even if I don't agree with it for one reason.

Hardware RNGs are available on the market from Intel (and probably other manufacturers as well). An algorithm to transform the output of 0s and 1s into an unbiased shuffle requires nothing beyond some very basic skills.

What I am getting at is that it it is so *easy* in the current day to generate purely random numbers and use those to construct a purely random shuffle that I cannot imagine any reasonably sized company getting it wrong 'accidentally'.

Quite.

In all honesty, even the use of hardware RNG's is complete overkill.

The simple action of any reasonable software RNG taken together with the fact that these servers deal between hundreds and over a thousand hands a second is more than enough to guarantee adequate randomness.

I am not sure I agree with this but I have not thought much about 'adequate randomness' in terms of playing poker. Here are my initial thoughts. One of these three things is always true:

a) the shuffle is purely random
b) not all of the 52! shuffles are possible.
c) all shuffles are possible but not all are equally likely.

If 'adequate randomness' means something other than 'pure randomness' then either b or c is true.

In case b) if the possible shuffles are known it may be possible to predict that the next card to fall can never be a given card.
In case c) we have a similar situation except we only know that one card is more likely to fall than another. Either of these seems inadequate.

'adequate randomness' may be possible if one of b) or c) is true, but it is somehow not possible to know which shuffles are impossible / less likely / more likely. I am not sure it is possible for this to be the case, though... would have to research more into it.

It seems like needlessly sticky water for the sites to get into instead of using a hardware RNG though

I would think very, very few. I think most people who play play money think it's really annoying that many people just go all-in all the time, or "don't play seriously". I doubt many think they could win at the same rate in a cash game.

OTOH, many people complain saying that the people who just go all-in (or otherwise "so bad") all the time are unbeatable. I know there's a 2+2 book that talks about this... Small Stakes Hold'em, probably.

This all true. But there are a few of us who understand the concepts of selective memory and fooled by randomness--who would still like to see actual evidence that the major sites are 100% clean. Well at least me and Ed Miller. I don't just take it on blind faith that the sites are not manipulated--given the obvious benefits to manipulating the deck, human nature, and how easy it would be to cover your tracks from a simple card distribution analysis. I still would really love to see a real 3rd party code audit, or regression analysis of high equity showdowns with player demographic info. Neither of those is impossible to do.

Am I crazy, or was I not reading a discussion earlier in the day where a poster (don't remember who) put up a couple graphs showing his play on stars and ftp. A discussion started, but I then had to leave for the day. Now I come back and I can't find the graphs or the related discussion anywhere in the last day's posts. The poster had done well at stars and not at full tilt I think.

Now, I'm exhausted and its late here, but were a bunch of posts deleted?

Re: The great "Poker is rigged debate" - Collected threads edition
Quote:
Originally Posted by TeflonDawg View Post
This thread is so lol I can't stop coming back for a laugh!

Anyone else notice that the average rigtard post:

1) Is filled with incorrect grammar and spelling
2) Is poorly structured
3) Contains epic fail in logic and critical thinking skills
4) Always says something like I won at X site, but can't win at Y site?

#4 is especially hilarious b/c I've played at probably 5-10 different sites, and there's always some moron saying "this site rewards bad play f u (insert site name here)" at every single site!

It's amazing that some of you idiots still think UB and AP rigged their RNG when the scandal was player cheating.

Equally amazing is how some of you idiots think the RNG is rigged, but can't specify how exactly it's being rigged! If I play 1000 hands and profit $500 or play 1000 hands and lose $500, the site still makes the same amount of money off me in either scenario!

This all true. But there are a few of us who understand the concepts of selective memory and fooled by randomness--who would still like to see actual evidence that the major sites are 100% clean. Well at least me and Ed Miller. I don't just take it on blind faith that the sites are not manipulated--given the obvious benefits to manipulating the deck, human nature, and how easy it would be to cover your tracks from a simple card distribution analysis. I still would really love to see a real 3rd party code audit, or regression analysis of high equity showdowns with player demographic info. Neither of those is impossible to do.
Last edited by suzzer99; Yesterday at 09:24 PM.

^ THESE

/thread

I haven't a clue what you are trying to say in this post.

When I say 'adequate' I mean that although a deterministic machine can never produce completely random numbers it could be close enough that it could make no measurable difference.

This is particularly true where you have multiple games obtaining random numbers from the same RNG since tghe numbers for each game are not only randomised by the RNG but also by the precise moment that they were picked.

Absolutely. A hardware RNG is not expensive and is more random than a software one so why not save yourself hassle and go for it.

Dear Full Tilt Poker,

why do you punish good players and reward the bad players on your System so much? I think this is really not fair and and really annoying! I know that we all live from bad players' money and they have to be kept. I agree that they have to be motivated to cash in more money after losing everything and therefore they cannot be exploited too fast. They should go broke while having fun. BUT honestly I think you are giving them a little bit too many suckouts lately. Of course this is good for you because by keeping an equilibrium between good players and bad ones you make the most money.
How can it be that I lose like every situation where I am 95% favorit on the turn and lose to any 2-outer? Why do I lose more than 50% of the situations where I am a HUGE favorit preflop (80%+?). Why does Ace-rag almost always beat the better Ace and why always on the river? All over the internet people are reporting the same.
I mean I really don't mind if your System is skewed a little bit in favor of the bad players if your profits and mine are somehow equally. The last couple of months you are going a little bit too far with the skewness.
Please don't put the doomswitch button on my account now.


Thank you for my upcoming upswing in advance.

Yours faithfully,
rake-paying customer

you clueless american. site will make same amount of money off you if you will lose 500$ or you'll win 500$ (they profit from rake you pay), but if you lose 500$ and go busto you most likely will leave and no more rake from you my friend. if you win 500$ you might cash it all out and leave or most likely you'll stick around to win more 500$ and generate more rake, don't ya? isn't that human nature?

so basically they don't gain same amount of profit from you if you win 500$ or lose 500$ (in short run they do), but in long run they will maximize their profit by making good conditions for you and others to stick around as much as you can. and business is all about long run.

My all-in luck in Pokerstars SnGs:

last 2600 allins:



last 560 allins were a nightmare:




Is it selective memory?
The graphs don't reflect the fact that my average equity dropped from 51% to 49% because almost every time my pocket pairs KK and lower were dominated.

No doubt PS doomswitched me. I hate this scam site and won't play it ever.

And I'm surprised how many idiots beleive online poker is not rigged. People are so stupid in general.

lol in before merge

if you put this much effort into improving maybe you wouldnt lick balls

I don't have to improve because I am almost always a huge favourite when I get my chips in (at least 70% favourite). Like I said before losing against an all-in at the turn with 95% chance of winning happens to me way to often lately. And I don't think it's because of the variance. I am playing online poker for many years now and have been through many downswings. What happens lately is not normal in my opinion.

Btw. I only like basketballs.

well said

I'm sure you're not interested but you can come up with a very good estimate for how unlucky you have been in all-ins if you know the total amount of tournament chips that were at stake in these all-ins, and your average equity in the all-ins (would need a few decimal places).

If you post these numbers I will tell you how many s.d away from the mean your results are.

yeah what was i thinking?

You clueless European. You miss the obvious distinction between the site owners who make fantastic money and the employees who make peanuts. Is security vigilant enough to stop innovative employees from rigging the game? The UB rigging proves security is lax. Management is happy if employees make some extra on the side, they would then have to shell out less wages, No? Your argument has been repeated many times and is complete bunk.

Maybe the new UB owners were paying off debts by letting those former owner?/employee/consultants have a run at the tables? No?

The actual RNG software or hardware would not have to be changed but only bypassed.

go kiss a crocodile

How can your average equity be 49% and your green line still be going up? Are there a lot of three - way allins in this data?

If it's possible, filter out the three way all-ins and tell me what your average equity was in those pots for the first graph.

That graph doesn't even look that bad.

I might be wrong though. lol donkaments, you know?

Ok. I played around a bit with the first graph and am pretty sure your average equity was not 51%. If 51% had been correct then the observed data is insanely close to the expected data. If anything you would be posting because it was interesting you ran so close to the expected, rather than because you were unlucky.

Here is my math, perhaps you can correct me.

Average equity 51.0%
Expected number of tournament chips won 640000
1% equity is therefore 640000
64000000 tournament chips were at stake in total.

We will estimate each hand as an independent identically distributed binomial variable. This ought to be always a big *underestimate* on the variance, (since variance ~ squares of the data, and squares are minimized over data with a constant sum if the data is identical). by underestimating the variance we overestimate how unlucky you have been.

64000000 / 2612 = 24502 tournament chips at stake each hand.
Each hand has mean M = 0.51*12256 - 0.49* 12256 = 245.12
(check - 245.12* 2612 hands = 640253 EV - close enough).

Each hand has variance

V(H) = (12256-245.12)^2 * 0.51 + (-12256-245.12)^2 * 0.49
= 73573231 + 76576220
= 150149451

v(2612 hands) = 392190367617 (since variance is additive)

s.d. (2612 hands) = sqrt(variance(2612hands)) = 626251

Your observed result of being 80k below e.v. is approx 0.13 s.ds below the mean, so you could expect to have done better than you did 55.17% of the time. In other words, 45% of the time you would have run worse than you did.

- interesting edit. If your equity was 52.0% then this reduces the total number of tournament chips by a factor of 2, the number involved per hand by 2, the variance by (roughly) a factor of 4, the variance over 2612 hands by a factor of 4, and the s.d by a factor of 2. This isn't quite right.
- but for numbers up to 60% it's probably not a disastrous estimate. If your equity was 51+x% where x is from 0 to 9, then the s.d of those 2612 hands is roughly 626251/x. Even if you were running at 58% equity (and thus the variance was a lot smaller), your observed result would be only 1 s.d. from the mean.