|
| Author |
Message |
acesupagain Pair
Joined: 28 Jul 2005 Posts: 25
|
Posted: Mon Sep 26, 2005 5:07 pm Post subject: RNG isn't random |
|
|
I'm just throwing this out there and am in no way saying online poker is rigged or not. Thats for you to decide. There is a series on the history channel called breaking vegas that goes into various stories of cheats, card counters, and other stuff. Anyway there is one episode where one of the guys that worked for the gaming commission and actually it was the guy who modernized the gaming commissions got caught cheating at slots and keno. He goes on to explain that a random number generator can never really be random and that it can always be cracked by a smart enough programmer. This guy actually had to write his own program to predict the outcome of the numbers, which he did. I think it took him three times (at least on T.V.) playing the predicted numbers untill he hit the jackpot.
Now this isn't proof that online poker is rigged but I think it qualifies as proof that it can be and probably is in some instances. I'm not a programmer so I can't explain to you how this guy did what he did, but it is a mattrer of public record that he did do it. |
|
| Back to top |
|
 |
zipster987 Four of a Kind
Joined: 18 Aug 2005 Posts: 278
|
Posted: Mon Sep 26, 2005 5:30 pm Post subject: |
|
|
A random number generator is based on time. What a computer does (laymans terms) is calculate the time, including parts of a second to the nth decimal. From there it goes through a series of arithmatic to come up with a your random number. Full Tilts program probably uses something like this, and then has every possible arrangement of 52 cards in an array. thats 52 factorial, which is VERY VERY LARGE. The 'random' number then determines which 'shuffle' to use and then that hand is dealt out. Logisticlly speaking, if I could start every hand at exactl .1145459 seconds past the exact full tick of full tilts server computers, every hand would be the same. But if i screw up.....say i miss by .0000001, and i start it at .1145460 seconds past the full tick, theoreticlly ive got a completly different hand being dealt out.
Thats where 256-bit encryption comes in. It would be very very VERY difficult to 'hack' full tilts computer program, figure out what digits their using for what, and what random numbers have been generated, but it IS possible. just like me and the dealer working together to give me boat after boat after boat, or the guy sitting to your left being my best friend and us sending signals back and forth. its POSSIBLE. its NOT LIKELY, but it is POSSIBLE. Thats where your faith has to come in, and believe that online poker isnt being cheated by a few people who know what their doing, just like you believe that vegas isnt screwing you by using a 51 card deck in blackjack with only 3 aces.
And for those who say this is how online poker is rigged, better rethink that. This 'way in' to cheat the game is possible, but if it happens its one or two or three guys who are playing 1000 dollar games and can tell what people have been dealt, it has nothing to do with someone trying to steal your money on 1-2 NL games. |
|
| Back to top |
|
 |
GripHoldOn Message Board Junkie
Joined: 24 Jun 2005 Posts: 2096 Location: Boston, MA
|
Posted: Mon Sep 26, 2005 6:22 pm Post subject: |
|
|
Here's an experiment that I ran because of this thread. Have at your disposal a TI-83 Plus (not silver edition) graphing calculator that has nothing important saved on it. The following will clear the memory of your calculator, and it will restore factory settings. Follow these instructions:
Press MODE
Press ALPHA
Press LN
Press ENTER
Press ON 3 times
Press Enter
Upon pressing enter, the memory clear will be done and your calculator will return to the home screen. Here is where we will employ the random number generator that this type of calculator features.
Press MATH
Scroll Left to get to PRB (Probability)
Scroll down to 5: randInt(
Press ENTER
Press 1
Press ,
Press 1
Press 0
This tells your calculator to randomly select a number greater than or equal to 1, and less than or equal to 10, and this action will take place as soon as you hit ENTER.
Press ENTER
Your calculator will give a number; Mine gave me ten. I don't know if yours will give you ten, but I do know that if you go and repeat the entire process, starting with the very first step to reclear the memory, you will get the same number you got the first time you did this.
This doesn't prove a whole lot in terms of the randomness of online shuffle sequences, but it does show us that randomness is not easily attainable, and the primitive random number generators in these calculators are not random. Randomness really isn't attainable by computers, but this doesn't mean that anyone, even full tilt knows what cards are being dealt. Remember, even humans are incapable of true randomness, just ask Rafe Furst. |
|
| Back to top |
|
 |
acesupagain Pair
Joined: 28 Jul 2005 Posts: 25
|
Posted: Mon Sep 26, 2005 7:08 pm Post subject: |
|
|
| I agree with you Grip I was just putting this out there for the people to decide for themselves. I really don't think that the average person could even come close to being able to write a program that could predict cards. I do however believe that there are people out there that can. Obviously this is true someone on the gaming commision did it. Whether or not there are people doing it in online poker I don't know. And to tell you the truth I really don't care because the number would be so insignifigant that I highly doubt it would affect my game. |
|
| Back to top |
|
 |
billytwoguns Pair
Joined: 22 Jun 2005 Posts: 24
|
Posted: Mon Sep 26, 2005 7:30 pm Post subject: Re: |
|
|
| acesupagain wrote: |
| I agree with you Grip I was just putting this out there for the people to decide for themselves. I really don't think that the average person could even come close to being able to write a program that could predict cards. I do however believe that there are people out there that can. Obviously this is true someone on the gaming commision did it. Whether or not there are people doing it in online poker I don't know. And to tell you the truth I really don't care because the number would be so insignifigant that I highly doubt it would affect my game. |
Aces,
There's really no way to get truely random numbers. What both RNG's and encryption hope to do is create a system complex enough that it's too difficult to solve in a reasonable amount of time. For instance, if I encrypt a message and using the fastest computers in the world it would take 50 years to decrypt my message, then I would feel secure that I have a pretty good encryption algorithm. Likewise, if it took 150 trillion hands to see a pattern in the cards being dealt, I would likewise feel pretty secure that I had something approximating randomness for my purposes.
Now, there are poor implementations of both encryption as well as RNGs. A few years ago a security company took PlanetPoker's RNG and found that it had flaws. They created a program to predict the cards coming out of the deck and then notified PlanetPoker of their error.
http://www.developer.com/java/other/article.php/616221
But, really, the comedy in all of this is that a live casino shuffle is anything but random. A skilled card player can watch a dealer shuffle cards and know the position of key cards. The shuffle machines in the local card rooms have a hardware RNG in them which is as exploitable as an online poker's RNG (actually, probably more so since online poker rooms tend to use a triple random method of randomizing the cards).
And all this boils down to saying that I agree with Zipster. If you're skilled enough to crack the RNG systems used by most online casinos, you're not going to waste your time laying bad beats on $1/$2 tables  |
|
| Back to top |
|
 |
Blacksting Flush
Joined: 07 Jul 2005 Posts: 130
|
Posted: Mon Sep 26, 2005 8:13 pm Post subject: |
|
|
The only "random" source so far is in nature. Some government encryption has used the recorded variance of quasar radio emissions for the random seed... which up to this point has not been found to have any pattern. (The down side is that both sides need the encryption key... the upside is there is no way to break it.)
Any other random generation built by man has bias. |
|
| Back to top |
|
 |
jbrennen Straight Flush
Joined: 16 Aug 2005 Posts: 422
|
Posted: Tue Sep 27, 2005 10:38 pm Post subject: |
|
|
There are some well known cryptographically secure pseudo random number generators (CSPRNGs) which are well documented and believed to be "uncrackable".
For instance, see the ISAAC algorithm: http://en.wikipedia.org/wiki/ISAAC
Any programmer hired to develop an online poker server will certainly have access to such information, and will probably come up with an algorithm which mixes a CSPRNG with ongoing injection of other sources of entropy (such as time stamps of incoming messages, etc.). That way, even if somebody could get a snapshot of the current "seed" of the generator, it would become useless within seconds due to the injection of unpredictable stimuli. And even if you could "game the system" and inject the entropy that you want, you still would need to be able to predict the output of the CSPRNG -- something which is computationally impossible without inside information. What it comes down to is that anybody who can reliably track the output of the CSPRNG probably already has sufficient access to the server that they could simply "see" the cards as they are dealt.
There are far more likely things to worry about than players "figuring out" the server's RNG. For instance, collusion between two or more players in a game. That's almost certainly already happening and furthermore, it's very difficult to detect. |
|
| Back to top |
|
 |
liquidgrip Pair
Joined: 12 Aug 2005 Posts: 45
|
Posted: Wed Sep 28, 2005 2:42 am Post subject: |
|
|
Someone else posted this website but I dont know who.
http://www.billrini.com/index.php/2005/09/25/why-online-poker-is-not-rigged/
I found it very educational. I especialy liked seeing Howard Leaderers comments on the RNG on fulltilt.
There is alot of good links at the bottom of the page. I also read up on how Paradise pokers RNG was cracked and believe alot of site learned from it and are implementing better RNG's today. |
|
| Back to top |
|
 |
Salt-N-Pekker High Card
Joined: 24 Sep 2005 Posts: 7
|
Posted: Wed Sep 28, 2005 7:03 am Post subject: |
|
|
| RNG is not random and programmers can design the software in the house's favor in a heart beat. Also the number of house players, and there are many, make on line poker sites even more rigged. |
|
| Back to top |
|
 |
KGBlovesOreos Moderator
Joined: 28 Jun 2005 Posts: 5318 Location: VA
|
Posted: Wed Sep 28, 2005 9:45 am Post subject: |
|
|
| i saw the same episode too, but i still don't think its possible to crack the codes for the RNG or whatever... if somebody did, i think full tilt would know about it immediately, and they would probably find the hacker... |
|
| Back to top |
|
 |
SatchDork Full House
Joined: 13 Sep 2005 Posts: 245
|
Posted: Thu Sep 29, 2005 3:06 am Post subject: Re: |
|
|
| Salt-N-Pekker wrote: |
| RNG is not random and programmers can design the software in the house's favor in a heart beat. Also the number of house players, and there are many, make on line poker sites even more rigged. |
;_( I can't win, it must be rigged. |
|
| Back to top |
|
 |
SatchDork Full House
Joined: 13 Sep 2005 Posts: 245
|
Posted: Thu Sep 29, 2005 3:29 am Post subject: |
|
|
Saw that episode of Breaking Vegas and, if I'm not mistaken, the guy had access to the Keno program's source code and that is how he was able to write a program that predicted the upcoming numbers. Remember that his program worked based on inputting actual results. The sequence went something like this:
1-"The Accomplice" sat in the Keno bar and saw the results for an actual Keno draw
2-The Accomplice relayed those results back to "The Guy"
3-The Guy put those actual results into his program and the program spit out the most likely result of the next draw
4-The Guy relays those predicted results back to The Accomplice, who bets on those results
5-The predicted results do NOT come up, so The Accomplice relays the actual results to The Guy again
6-The Guy inputs these new actual results into his program, which accounts for them and all previous actual results that were input, then again calculates the most likely result of the next draw
7-The Guy relays the new prediction to The Accomplice and the process repeats over and over again until the predicted results are actually correct
Like you said, it only took about 3 or 4 tries for them to hit it in the episode. It never says how many tries they actually had to go through. Anyway, the real point is that the guy had access to the source code in the first place. Therefore, he was able to design a program to figure out what "seed" (to use a term someone else used in this thread) the Keno program was currently running along.
To apply this to online poker, you would need the source code of the RNG and a program like The Guy's that could take into account your seat, your hand (since, regardless of how good a hacker you are, are still the only cards you can see at the beginning of a hand), the flop cards, the turn card, and the river card. Your ingenius program could then combine the cards you've seen with the RNG to predict the entire deal (to all players). Naturally, players coming and going from games (not just yours but every other on the site) leaving you in various states of short-handed and full-tabled are going to give your program fits. The other option would be to record millions upon millions of hands and THEN design a program to derive the algorithm for the RNG from those actual hand histories and THEN combine this with the prediction program. Of course, if you're a crappy player like Salt-n-Whatshisname, it's going to be very expensive to do this R & D. You're better off just working on becoming a decent player. |
|
| Back to top |
|
 |
SatchDork Full House
Joined: 13 Sep 2005 Posts: 245
|
Posted: Thu Sep 29, 2005 3:53 am Post subject: |
|
|
Another point just came to mind. What the "online poker is rigged" sect is really suggesting is not so much that the RNG isn't truly random so much as that it's rigged to create more "action hands" than should actually statistically occur. Their claim is that hands like this the following get dealt more often than they should be (gonna' use a 6-max table for simplicity's sake):
Player 1: AK
Player 2: 88
Player 3: 99
Player 4: 76
Player 5: JT
Player 6: AA
Then the flop comes K89 with a J on the turn and 2 on the river.
It doesn't take a very experienced poker player to see the action this deal is going to create. The heavy action creates a big pot which creates a big rake. The conspiracy theory asserts that online poker sites have rigged their RNG to create these types of deals more frequently than they should statistically occur. Normally these theorists are Player 1 or 2 who think that they should win 100% of the time that they flop top pair with AK or a set with a pocket pair. Sometimes they're even Player 6 who thinks his Aces should hold up every time they're dealt to him. The difference between the conspiracy theorists and actual good players is that they've seen this sort of thing before and know that, while rare, is does (and, in fact, WILL) happen if enough hands are played. Therefore, the good player dealt AK in this situation recognizes the situation he's in and, while he might still go to the showdown and lose, sacrifices a lot less money (by not getting involved in this particular raising war) than the conspiracy theorist.
The real point I wanted to make is this: even IF a poker site is "rigged for rake," it's poor play that causes ANY ONE INDIVIDUAL to lose money. Given enough hands (and given a rigged site), each of us takes a turn at being each Player 1-6 in the above hypothetical (THE "RIG" EFFECTS US ALL EQUALLY). It's how we actually play each hand that makes the difference in our actual earnings (remember, the site is "rigged for rake", not rigged in favor of any one individual). So, learn how to play well and the money you make when you're Player 3 will off-set what you lose when you're any other player. |
|
| Back to top |
|
 |
jbrennen Straight Flush
Joined: 16 Aug 2005 Posts: 422
|
Posted: Fri Sep 30, 2005 12:43 pm Post subject: Re: |
|
|
| SatchDork wrote: |
| The real point I wanted to make is this: even IF a poker site is "rigged for rake," it's poor play that causes ANY ONE INDIVIDUAL to lose money. Given enough hands (and given a rigged site), each of us takes a turn at being each Player 1-6 in the above hypothetical (THE "RIG" EFFECTS US ALL EQUALLY). |
Ah, but that's the problem. It doesn't affect everybody equally. Let's say that they rig the game so that the chance to fill a flush on the river is 50% higher than it should be; in an honest game, there are 9 cards to fill the flush out of 46 cards unseen, so the chance is 20% (rounded off). So let's say they rig the game so that you actually have a 30% chance to fill a flush on the river.
The uneducated fish will continue to call on the turn, going for that flush draw, and they don't care about the odds, they just know that if they hit their flush, they could score a big payoff.
But on the other hand, those who know how to compute pot odds will make incorrect decisions, because they believe that filling the flush is a 20% proposition, so they need to be able to win a total pot at least 5 times the size of their current bet in order to justify going for the draw. But in actuality, since it's a 30% proposition, they could make their bet and go for the draw even if the eventual pot size is only 3 1/3 times their current bet. That's a big difference, and given such a rigged game, the educated player will make the wrong decision quite often.
Looking at it from the other side, if I'm holding a hand like two pair, and my opponent seems to be on a flush draw, I'm going to bet on the turn a bet which denies him the correct pot odds to call me. If I think he's 20% to make his flush, but he's actually 30%, I'm likely to make a too-small bet which actually gives him a positive expectation for calling me.
Of course, this is a totally silly proposition, because if some online poker site did rig the game to fill flushes on the river 30% of the time, that sort of statistic would be obvious over a sampling of several thousand hands. They could probably fool a table full of players for a while without anybody suspecting a rigged game, but over thousands of hands, the deviation from the honest expected rate would rise to the point where it would be obvious and couldn't be explained by random variance.
Let me finally add that, no, I don't think online poker is rigged. As somebody who is fairly well versed in probability, statistics, and combinatorics (I actually know the mathematical definition of variance and what it means), and a professional programmer for 20 years now, I can't imagine how you could rig the game for significantly increased rake without causing significant perturbations in the observed statistical patterns of card distribution. I haven't done any extensive analysis myself, but I have yet to hear of a single long-term analysis of online poker hands which deviates significantly from the expected honest behavior. |
|
| Back to top |
|
 |
Showe101 High Card
Joined: 30 Sep 2005 Posts: 2
|
Posted: Fri Sep 30, 2005 4:35 pm Post subject: |
|
|
All of the previous posts seem to have been very knowledgable and interesting, thank you.
Now my opinon. 8)
Though I do not think Full Tilt rigs games, on many instances, fishy things have occured. Let me bring about a great example.
I joined a $5/$10 NL table, starting out with $500 in chips. Before I entered the table, I noted many of my opponents. No one stood out (acting on tilt or what not). I finally sat, and picked up KK the first hand. Four limpers, I raised to $100 preflop, attempting to avoid ANY outdraws. This fellow on the button calls my $100 raise and the flop is 3 4 6. Perfect flop, I go all in, for $400 more, and he calls. This fellow turns over 7 5 off suit, and took down the pot.
Aside from that, you may also find this story interesting. I heard a rumor on a fairness policy on Pokerstars. The object of the randomness was the following. You needed to keep track of all of your hands, and ASSUME which hands were your "winning" hands. After about 50 hands, I noticed a pattern, for example, 7 losing hands, then 2 winning hands in a row. Sure enough, there was a pattern, slightly changed when someone would enter or leave the table, but there was a definate pattern.
In conclusion, I have intermediate expirience with Visual Basic programming. Getting a "random" number isn't possible, but I'll tell you what is. If I want a random number between the numbers 1 and 52, I can easily achieve this. Therefore, a some what random number can be attainable.
Matt |
|
| Back to top |
|
 |
|
Powered by phpBB © 2001, 2005 phpBB Group
|
|