Relevant

The RNG at Full Tilt Poker Keeps A Lot of Players Thinking Late at Night.

JP Chips in again as the Full Tilt RNG situation keeps alive in the forum.

There was a team at a university, I believe I read it was a university team not a company, that found a flaw in a poker site RNG and were able to predict with exact certainty all the cards to come as well as opponents hole cards. They discovered this "glitch" and notified the site. The site got it corrected and then the team published the results of the study and the glitch they were able to exploit if they wanted. To their credit they did not use any of the info and before letting any info get out to the public, the system in question was alerted there was a problem as soon as it was discovered.

RNG's are just programs. They have limitations. There is the "washer" using 3 formats of rng's and combining the results. Then the continuous "shuffle" once streamed. There is also the photon or proton fired at a mirror or through glass (i forget which it was) that deflects different every time at random as a fact of nature and cannot be predicted. This is somehow measured and a value assigned to incorporate into a rng. The atomic clock principle is another method but very predictable along with the one that uses names, birth dates and other fixed data to "randomly" mix and spit back out.

ALL of these still come down to the fact that in order to meet statistical norms if it were "live deal" there ARE fixed combinations granted the numbers are extremely hi, there are still "statistical norms" that are to be met IF the program is to be considered "accurate" and as random as possible to compare to "live". According to an e-mail from FT, they did mention something to the effect that if we tracked it we would find that their deal would meet long term correct EV statistics. (something to that effect). There was also a non-sanctioned comparison done of 3 different poker sites that DID make their data public on hands dealt. Among the 3 sites ALL HOLE CARD STARTING HANDS DEALT were within .0x and .00x of each other across the board. They went live and of course could only do a 1,000 hands and not the millions in the databases. but they transposed exponentially and found that live deal had a bit more variance in the numbers but were VERY close to the computer deals in all but half dozen hands or so. Everything was 1% or less deviance and the few exceptions were up to 3% difference. I think they only compared two game deals in the numbers due to time. They said perhaps over time the deviance may balance out to the same or extreme close similarity as the computer models.

Now then, since the different sites are all so ridiculously close, it can be postulated that "statistical averages" are being dealt and not specific randomness. In order for statistical longer term EV to be any kind of basis for a "normal average" then there must be a measure in place and a protective apparatus to catch anomalies which would occur and skew results.

Technically, if you flip a coin 10 times and 10 times it hits heads, then the odds of it being heads on the 11th flip is still exactly 50%. There can be outside influences which can affect the outcome to a degree to give it a slight skew to favor one side over the other. But theoretically, all things equal, the number is finite and exact to be either x or y which is 50% if we presume it can only be x and y with zero allowance for "edge" (coin standing on edge).

Now then, it's been mentioned that it would be ridiculous for a site to mess with anything to skew numbers. Why would a site rig this or that? Yes, why WOULD a site risk millions of dollars in profits they will make anyway? GREED! To say it would have been uncovered by now if it were occurring is just the ostrich effect. Head in sand, don't see it, it don't exist. The online gaming industry is still relatively "new" on the scene whereas poker roots go back to 600BC and through the millennia the game evolves, cheats occur, get discovered eventually and the game evolves further. Not only is the industry new it is highly secretive with a lot of nepotism and very few controlling bodies at the top levels. Half the sites in the world probably go back to the Kawanakee tribe in some way for example. A LOT of them do, we do now that for sure. As to cheating not possible all we have to do is look at the AB/UB scandal. That is still unfolding and even with the cover ups so much is coming to light that it seems to be going all the way to the top there. Now THAT was complete and blatant stealing and cheating due to corruption and greed.

This is NOT to say that rigging and cheating is being done at Full Tilt or any other site. (Note, there are a LOT of unscrupulous sites out there and some open and close we never hear about)

There are things that could be done and there ARE things that can favor FT or any site for that matter. ARE they being done? Who knows. It does not involved direct stealing or rigging of games for one player over another or game or level over another. Step up the pace enough and increase rake per resource for example. I am not saying this happens, but there IS the conflict of interest in there for it to occur. This is one reason why a number of top sites open so much up to public scrutiny and post the information. To help show they understand our concerns and there IS potential there to exploit the masses. FT on the other hand, still refuses to do so and even refuses in dependant audits. That's like a U.S. Citizen telling the IRS we did our own audit and everything is perfect on our tax form, accept it as fact for the amount we pay you. For the most part this is going to be accurate, but for some it won't be. Hence the market for independent auditors for people who go beyond to prove their own integrity. FT keeps everything secret and in-house. That does not help assure the utmost integrity. AB and UB have been around longer than FT? IF anything is wrong, maybe it's a matter of time until discovery.

While you are completely correct, if they DO follow that format for RNG/Wash/Shuffle/Deal, then it is completely random and cannot be altered or predicted in any way. You are also correct, if they step in and make a change, it could make the outcome predictable and exploitable. HOWEVER, I disagree that it could be exploitable by anyone including math/computer science majors or a compendium of such. The reason you could NOT predict it is because you would not have ALL of the info available.

Because in nature, even totally random actions or events can be "streaky", yet nature does not measure and does not care to make anything "normal" are "statistically averaged". If there are more sun spots in one sector of el sol than the rest, well, that's just it, there will always be a higher number there. The sun won't fire off the other side to balance out. MANKIND on the other hand is different, we cannot accept "nature" and everything must fit into a form, be tangible, measurable, predictable, average and statistically "correct" in order to be "normal".

We know that "cards" can be expected occur in certain numbers over time. It's measured, fixed, proven and with slight variance (which has also been able to be measured and proven mathematically to occur) will continue in infinity. If something begins to occur in a streak, then we look at it as odd. The roller hitting "lucky 7" repeatedly is cheating or there is something wrong with the dice or table, the roulette wheel is rigged or the ball etc... Ball is changed, die is changed, decks are changed, dealer is changed. Everything is done to assure "randomness" according to EV or statistical expectations and protect against cheating.

Now we come to virtual cards and rng's and math, statistics, odds, random, nature and a whole plethora of measurable and IMmeasurable aspects. Since everyone knows the statistical average numbers over millions and millions of hands live dealt in real life these are used as a basis to begin the programming. HOW it occurs is mute. All we need to know it that it is random and over time will meet expected averages.

Slot machines for example ARE rigged and regulated by law to BE rigged because the casino MUST pay out xx% per year back of money taken in via these machines. HOW they do it is up to the casino as long as the audit shows the numbers meet the letter of the law. Now then, back to cards.... Things ARE streaky and that is also a fact of nature. If the software is measuring something out of the norm, then it must have built in safeguards to protect "predictability" and to also make sure it meets "expected statistical averages". When this is picked up then an adjustment or tweak must be applied to bring things back into balance. WHEN these occur we can't possibly know or tell. We can't possibly predict if we will be on the winning end of such a tweak or the losing end, that too is a random occurrence.

If in Keno or the state Lottery, a certain ball keeps coming up more than others, something is wrong. It is inspected or just simple replaced. For all we know they simply change whole sets at "random" or at regular intervals to make sure nothing can influence the outcome. In cards this must be done within the software. Without such protection then numbers can vary dramatically if a glitch should ever occur. We are dealing with many variables and it is possible for a glitch to occur.

IS FT cheating and rigging the deck/game to get more rake from us? I hope not and really can't see anyone doing anything so insane. IS FT interfering with the deal? Of course. They have to monitor it and make sure it is meeting expectations. CAN this be exploited? No. Besides requiring a Cray or two and a whole team of braniacs to figure it out, they also will require a lot more information than is available to them and even if they had 99% of the info, there is no way to know where the "hit" will occur. They have to know EVERY hand dealt at every table over a period of time since the last "tweak", what the starting basis was from that point, what the current measure is, what the "statistical average" chart forms the basis of the program for what expectations to meet AND what the tolerance level is included in that. They can then know if x occurs more than y and can then bet x more until the "threshold of the tolerance point" is met. Avoid playing altogether while the the "adjustment/tweak" is made to bring things back within the programs functional parameters, then they start over measuring and have to play at random based on skill until they begin to see a trend occurring again that is measurable and consistent.

WHEN these tweaks occur and WHERE, what table and what level can not possibly be known and knowing every hand dealt cannot possibly be known due to all the "mucks" so it would be impossible to predict and exploit. That does NOT mean it does not occur, it only means it is still "random".

FT and their "deal" or "badbeat/suckout ratio" is widely discussed and experienced across a very wide range of players and on many forums. Bad beats occur, suckouts occur, so what, that is normal and expected to happen. The point is the frequency according to common or popular "perception" is extremely high at FT compared to other notable sites. Now EVERY site has people complaining how horrible it is to play there and their horrible bad beats. Sit at Bodog, Stars, UB and I here the same stories. Usually, I can look a persons stats over and see why they are getting "bad beats" confusing playing the wrong hand the wrong way and losing with it as a bad beat or suckout. If they lose all the time it can't be them.

But what I DO see that is different, is that type player and skill level of player making the statements. FT has a much higher percentage of solid players making the same type remarks as the weak players on other sites. From personal experience, I suffer bad beats and suckouts across all sites and all tables and levels on all sites. But I experience MORE at FT. NOT because I play FT more. Often I will play two sites at the same time. I have played 4 sites at the same time making comparisons.

Now PART of this can be due to the fact that there are more players making mistakes on FT and playing junk hands. Hmmm, that makes no sense. Because IF you look at a bad player LIVE making more mistakes playing junk, they LOSE more often. Now then, bad beats, suckouts, miracle one outers occur at random in real life all the time too. It happens and it's part of poker. In my lifetime to date, I have not played as many hands live as online. HOWEVER, that was not the case a year ago when it was about the same. In my lifetime playing live I never experienced the amount of suckouts live as I did on line.

I think there is a problem with the RNG at Full Tilt. Correction, there is a problem with the program parameters at FT. Let us say for example, JUST to keep numbers simple and round (they are very complex and this is for comparison only, they are NOT real numbers)... Let us say that Aces are supposed to win 8 times out of 10 times you play them. Now we ALL know that if you bet normal you usually just get the blinds. But if the program does not differentiate then it reads you were dealt aces 10 times and 8 times you won, therefore now you lose with them the next two times (no it does not work this simple). It is not seeing that 4 of the 8 times it was blinds won and hand not played, no flop even. So now your aces played will end up winning only 40% of the time.

Granted, this is WAY too simplistic an explanation and it doesn't work that way. But it does show that if using the wrong information to postulate from then the results are going to be skewed. It is a given that they have their program established to be "random" and "unpredictable" but it is also set to meet a certain "expected statistical average".

I still play at FT. I don't think they are trying to cheat "me" in any way shape or form. I think they are not fair to the masses and something isn't exactly right with the deal. But as to being rigged I don't see it as that either. It's "fixed" in that it fits within their functional parameters whatever those may be. It is streaky, variance occurs, suckouts happen etc... I have to develop my game and develop ME to play my best and that means dealing with suckouts, bad beats, my own mistakes, knowing the difference, tilt, and getting better with time always trying to do my best and be my best.

I still think FT will always be tarnished in this shady gray area of something really bad wrong until they get their act together like others and publish some info and carry out independent 3rd party audits. Perhaps there is something wrong in their numbers to back up this "bad beat dealt" theory of more at FT than other sites. If they publish the numbers it will become apparent. Perhaps when they get their problem corrected, they will then publish and have an audit done.

There is one aspect to look at, and that is Bad Beats normally occur against Good players. The good players generally play better hands and have a much greater propensity to be in a hand with a poor player running junk only to get lucky and "catch". The poor player won't have AA in position and a good player running 95o from UTG calling raising and hoping to get lucky by the end of the deal. So if you get hit by bad beats a lot, GOOD! That means you are playing right. That means we have to deal with them.

We cannot change the cards we are dealt, nor can we predict what is going to hit. We can only play knowing what we hold, and what is showing, the rest is skill and luck of what might come. Whether it is a perfect RNG identical to a live deal or an imperfect rng, the fact remains, you don't know what that next card coming is going to be and therefore it is in all likelihood some random unknown variable about to hit you.

Side note: FT has flaws in it's RNG/Deal/Software, but it's playable with reasonable assumption that solid play and good decisions long term will result in winning and not whining