Please take 200 decks of real cards and deal 200 nine handed games at a rate of, say, 45 hands per hour, per table, 8 hours per day, and tabulate your results.
Using a single deck of cards and dealing multiple single games at any rate that you are physically capable of is totally insignificant to the scale of Replay. It would be extraordinary if you were able to reproduce the results that you see here and well worth further investigation.
I and others have pointed you to some very useful sites that will help you understand probability in general and the odds of specific poker hands. I donât see how we can progress in this discussion while you insist that your observations, you still havenât provided any data, have greater weight than known mathematics.
We have data analysts and statisticians and several other maths âexpertsâ on this site as active players. We have one person who, if I remember correctly, completed a PhD based on some study or analysis of poker. It is not possible that some sort of âcheatingâ algorithm, some sort of non-random play, has come to your attention and these very well educated people, who make a living by detecting anomalies, have not noticed anything unusual.
again not dealing with 200 decks. its 1 deck 1 table. anywhere from heads up to 9 handed. my deck of cards is not affected by what goes on at other tables.
You need sufficiently large numbers to see any patterns.
In the time it takes for you to deal a single hand at home, Replay has dealt several hundred hands. There is no comparison and you can not expect to duplicate your perceived anomalies using the method that you are using.
Letâs imagine that you are looking for a 1/200 chance event. Coincidentally, getting dealt any specific pocket pair is a 1/221 chance event. In the time that it takes you to deal a single hand at home, that particular event has already happened on Replay. There will be, on average, 200 pocket Aces dealt on Replay in the time that it takes you to deal yourself just one.
also, ths you play faster online is a poor excuse.
you and I leaving the same location to go some place. we going same route. I am going faster than you and will arrive faster. you arriving much later. we both went the same distance.
same is for poker tournament. you and I playing poker tournament. you playing online which took 90 mins and took 200 hands. I play a live game and its took 180 mins, but it also took 200 hands. we arrived at the same amount of hands (distance)
I do see patterns and seen a lot of hands. I can see these patterns and fold based on my experiences on here.
how many darn times I got say every time I get KK there is a ace on board and we all know any Ax gets played.
how many times I got to say that every time I flop a straight that there is a flush draw and turn or river will give the player their flush
200 cars going to the same destination, each with a probability of NOT breaking down of 1/200 versus your car with a probability of NOT breaking down of 1/200.
If I had to bet, I would bet that at least one of the 200 cars would get to the destination and you would not.
Clearly, the game of Replay Poker is identical to the game of Texas Hold 'em in most regards. However, an essential rule of Texas Hold 'em is that the deck be shuffled prior to each hand. Stated in terms of probability, all orderings of the cards in the deck should be equally likely at the beginning of each and every hand. When this condition holds, I say that the deck is uniformly random (meaning that the set of cards is sampled uniformly at random, without replacement, in the deal).
I have not managed to locate any claim by Replay Gaming LTD that the deck is uniformly random in the game of Replay Poker. The company emphasizes that its random number generator (RNG) has been certified independently. What this means, to put it very loosely, is that the Replay Poker software has access to a source of 0s and 1s, and that for each and every bit (binary digit) emitted by the source, the bit is as likely to be a 0 as a 1. However, this does not tell us how the software uses the RNG. The fact of the matter is that the RNG might be used to approximate any non-uniform deck whatsoever. Then the dealing of cards would be biased, with some sequences of cards more likely than others.
I should emphasize that I am not raising the issue of fairness of the game of Replay Poker. A game with a biased (non-uniformly distributed) deck is fair, technically, as long as the bias does not favor any player over another. However, to use the word fair in a different sense, I will say that it would not be fair for Replay Gaming LTD to let customers (people who, unlike me, pay to play) believe that they are gaining skills at Texas Hold 'em poker if the deck were biased in the game of Replay Poker.
I am asking Replay Gaming LTD to issue a plain statement as to whether the deck is well shuffled in the game of Replay Poker. That is, are all orderings of cards in the deck equally likely? More technically, does the Replay Poker software deal cards by simulating uniform random sampling (without replacement) of the set of 52 cards? It would not be such a horrible thing if Replay Gaming were to announce that it biases the deck to make the game of Replay Poker more exciting than Texas Hold 'em poker, and that the bias does not favor any player over another (provided that the statement were true).
In Sep 2015, Gatzby (then one of the devs) said thisâŚ
âHereâs how the dealer works: For each hand, we create a new deck deck of cards ordered lowest to highest. A card is then randomly selected from that deck and put it into another deck. This process is repeated until every card has been randomly selected from the first deck and moved to the second deck. After the second (entirely random) deck is complete, the dealing process begins from the top of the deck just as you would have in real life poker. As mentioned before, this process has absolutely no knowledge of anybodyâs hands or previous decks.â
Thatâs a good description of an unbiased shuffling algorithm. My question is now whether Replay Gaming LTD will commit publicly to dealing cards by this algorithm.
ETA: Oops, the statement of the algorithm needs to be tightened up a bit, to satisfy me. Each card transferred from the first deck to the second should be drawn uniformly at random from the cards remaining in the first deck.
Replay Gaming LTD has told us that its random generation of a sequence of 0s and 1s is certified to be good. But that is only part of what players need to know. I have asked the company to state publicly how it deals cards (presumably using the sequence of 0s and 1s that comes from the random number generator).
How would you define âuniformly?â One of the devs told us how the decks are shuffled and dealt. The devs represent Replay when they post to the forum like that, and they would know better than anyone exactly how it is done.
If them telling you once didnât satisfy you, would telling you 2 or 3 times suffice?
If they lied the first time, they would just lie again. If they were telling the truth, their answer would remain the same, so whatâs the point?
Someone asked how they do it ,they said how. Thatâs good enough for me.
You will not be satisfied until Replay can accurately describe each and every card that will be dealt which kind of defeats the purpose of your âquestionâ. That would, indeed, prove your âthoughtâ, if I can call it that, that the deck(s) and dealing are not random.
You remind me of a god-botherer ⌠you believe in something that can not be falsified and then ask that we falsify it.
It is not possible to prove that Replay are honest. There is absolutely no way ever of proving this.
A code review will lead to questions such as âhow do we know that this is the code that is actually running?â. An observer, a skilled software tester, can observe the running system for how long? Replay is a constantly running system. Are you prepared to pay 3 observers to cover all 24 hours? What if Replay outbid you and offer the observers a âlittle bonusâ if they just report that everything is correct?
It really doesnât matter how the system is tested or observed, there is an argument that test was unfair in some way or there are some doubts.
Please collect the data and post them here so that people who have the skills to analyse it can do so. As I have said before, I suspect that 100,000 hands, given that Replay deals many millions of hands per week, would give us an insight as to whether further data collection is reasonable.
âUniformly at randomâ is a common term. Google, and ye shall find.
What a self-identified developer says is unimportant to me. Replay Gaming LTD has made a legally binding statement regarding the sequence of 0s and 1s that comes out of the random number generator used by Replay Poker software. Iâm asking that the company similarly make a legally binding statement regarding the sequence of cards that comes out of the deck. Itâs not a huge thing to ask.
I request that Replay Gaming LTD make a statement that can be put to statistical test, and you respond by speculating that some unstated belief of mine is unfalsifiable. Thatâs a tad ironic, donât you think?
LOL. Knowing now that Iâve asked for a statement that can be put to statistical test, perhaps you have guessed that I intend to do the testing. Youâre not exactly brave in telling me to collect the data. The only way I know to do it is by screen-scraping the visual replays. The programming is not hard, but it is tedious (and Iâm presently working on another project). More importantly, the visual replays are very slow.
Iâd love to hear of a better way to collect data from Replay.
Real-money sites enable players to download their entire hand histories in textual form. That facility is easier to implement than the visual replay of hands at Replay Poker. Perhaps you will join me in asking that Replay Gaming LTD allow us to download our hand histories.
How much data you need depends on what youâre analyzing. There was a period of several weeks, back in late October and early November, when I saw four-of-a-kind three or four times daily. (Other players noticed that there were lots and lots of quads â how about you?). A run like that is ridiculously improbable if all permutations of the cards in the deck are equiprobable at the beginning of every hand. If I could download my hand history, it would be easy to do some convincing analysis. But my hands are lost in a veritable ocean of hands at Replay. I have a good idea of how to improve the efficiency of a search for them. But Iâm not sure, even then, that Iâll manage to find the majority of them.
For what little itâs worth, I collected data on how often I was dealt a pocket pair and then flopped (or would have flopped) a set. I quit after 1,023 instances because the frequency was within 0.02% of statistical expectation. Is this conclusive? No, but it was enough to satisfy me.
I know of 2 other data collections, one by ILoveTacos and one by Puggywug. Both of these were also within margin of error for their respective sample sizes. All of the data are on this forum, but I am far too lazy to dig it out.
As far as Iâm concerned, you have every right to ask for a statement. I was just saying that they have made such a statement, and that the limited amount of data collected seems to support their statement. Do with this as you will.
