I believe that I am not the only one that 1)gets alot of gut shot draws after flop or turn 2) And the board or hand pairs with turn. The management is aware of the deficits in its algorithm. May not be a huge deal but I had gut shot draws on flop (10) and turn (5) and the board paired 4 times and my hand paired 5 times out of 15 in the last tourney I played. The room gets a kick out of me predicting paired board.
What are you noticing?
1 Like
Iām noticing a lot of people proclaiming to have found some anomaly because they just assume that common occurrences should be rare for some reason.
The odds of either the board or your hand pairing on the turn is 1 in 3, which seems to be pretty much exactly what you observed?
1 Like
The chance to have a pair on Turn is better than the chance to have nothing. This is math, brute force, without RNG, without to think āwhat has to beā.
A big help for Notmathematicians (?) and Notbelievers. I made me such things for learning poker. German numberformat again, ā.ā = ā,ā and reverse. You can save the picture to your Desktop.
1 Like
Right. When I said 1 in 3, I was talking only about making a pair on the turn when there wasnāt one already. The 3 cards on the board and the 2 in our hand make 5 cards that can pair. Thereās 3 remaining suits of each, and 47 unseen cards = 15 / 47 = ~1/3.
Itās also pretty hard to arrange 5 unpaired cards and not have at least a gutshot.
1 Like
Ok, no pair on flop, 1 pair on turn, chance 1 in 3.02 hands.
And: On short time / not many hands it can be many near this numbers.
2 Likes
I have noticed that lone opponents often draw a three-outer (usually an ace) on the river against me, and I have an idea why it happens. Itās not the algorithm. Itās bc Iām playing good hands, betting aggressively pre-flop and/or when I flop top pair, and the only time people chase rather than fold is when they have a draw or an ace.
Opponents also criticize how I raise them if I flop a flush draw along with an open-ended straight draw (two shots at 15 outs to make my hand) when they have top pair, and I hit to send them to the rail. Never mind Iām a 56% fave at that point, have a 1% edge if theyāve flopped two pair, and will still win two out of five times when they have pocket pairs and flopped a set.
1 Like
Poki, youāre obviously a very talented programmer. Strange, Iāve always suspected someone will always have at least a pair when playing whether based on the board on not on the board. The odds seem to be in line with what Iāve expected except for the two which Iāve noted below.
Just for my clarification, what do you mean āwithout RNGā? How did you program this without using or developing an RNG to calculate the results?
Your results also highlight a concern of mine about Replayās shuffle RNG regarding the frequency which Iāve seen for Quads. The table shows for 7 cards; it indicates 1 in 595 hands on average. Iāve mentioned many times; over the last 5 years Iāve noticed at least 1 or more times each session which Iāve played HE and I usually play only around 200 hands per session and seldom more than 300. This sort of supports my observation that something is off with the RNG because this table indicates that Iām seeing it at least 3x more than I should be observing it.
I donāt use anything like your table, but I do use my own probability % estimates. The table indicates Quads would occur about 1/595 = 0.1681% of the time. In other words, I should never expect to see them even though i know they will occur since Iām only playing around 200 hands per session.
The other exception Iāve noticed is the high frequency of a Flush which seems to occur more frequently than either trips or straights on Replay which both should occur more frequently and shows in the table.
Iām just hoping someone (Replay) will just review the RNG to verify that itās still working correctly. If thereās an error, they should just correct it. No need to announce it since it wasnāt intentional.
If there is a flaw, Iāve stated before that everyone is facing the same flaw, so no harm was done.
Thatās how often you will make quads. If youāre playing 6 handed, youād expect quads to show up roughly 6 times as often, or about every hundred hands. (Itās not quite 6 times as often due to card removal effects, but itās close enough)
Also, not to speak for @Poki65, but an RNG is not needed because you can just iterate through all possible 7 card combos.
I guess I donāt understand this or how he is doing this table.
I guess Iām thinking how often these hands would occur.
Your reply above conflicts with your previous reply in post 3 about making a pair, unless youāre saying you would always expect to make a pair by the Turn when there wasnāt one already by the Turn.
I donāt believe you. You either grossly underestimate the number of hands played or you grossly overestimate the number of quads and flushes that you see. You don;t actually keep any data. You never present any at least.
All that you have to do to prove that the deals are not random is record 10,000 or 20,000 hands in a spreadsheet, your hole cards and the runout and count up how many sets, trips, quads, straights, flushes you hit by the river and compare those numbers to pokiās numbers. Only the hands that make the river because your folded hands would have hit stuff too. I would think that your numbers would be a little bit higher than pokiās because you will tend to fold hands with less equity (chance of flushes and quads etc) and play out hands with more equity (chances of flushes and quads etc). If your numbers are lower, or waaay higher, then you might have a case.
YOU do it. Youāre the one who wants to prove it, so you do it. I just told you how.
You need to multiply the number of hands that you play by the number of Villains at the table. If you play 525 hands per day, YOU should get quads roughly once per day but only if you see the river all 525 times. If you see the river say 100 times, then you should get quads every 5 days or so. Now, if you are playing 525 hands of 6max all day, the TABLE will hit quads roughly 6 TIMES if everyone sees the river, but if players are seeing the river roughly 20% of the time there should be better than 1 quads per day. And if you play 600 700 800 hands and you donāt see quads at all, that would be quite rare. Thatās what you are seeing. ; )
Run through every possible combination of cards exactly once and just count the different hands and convert the frequencies to percentages. Quite simple.
Do you ever actually read and understand a post before responding?
And you think this is done without using an RNG?
Maybe you can describe how itās done since Iāve already stated I donāt know how this is done.
Fan out a deck of cards face up. The first 7 cards make the first hand (no cards are removed, weāre just making a note of what those cards are). Then the first 6 and the 8th card. The first 6 then the 9th, etc all the way up the first 6 and the 52nd. Then select the first 5 followed by the 7th & 8th cards. Then 1-5 & 7 & 9. If you repeat that pattern until you get to selecting the last 7 cards in the deck, you will have gone through every single possible 7 card hand exactly once.
It doesnāt matter what order the deck is in to start out, and we never change the order, so we donāt need to shuffle and no RNG is required.
(NOTE: do not actually try to do this by hand. A computer can run through all 130+ million combos in no time, it will take you a fair bit longer 
Is this what was done or are you just guessing?
Because this isnāt the same as shuffling and dealing out a hand and I wouldnāt consider it the same.
Your first question, yes I understand you completely. You do not present any actual data - ever! All you give as evidence is anecdotal observation. You need to actually keep track of the hands and actually do some math to prove that something is out of whack. You donāt do that.
For your second question, it is very simple. What poki has done is he has listed out EVERY POSSIBLE 5 card combination, 6 card combination, 7 card combination, and counted how many pairs, 2-pairs, trips/sets, quads, etc etc happen, and given a percentage/frequency that each outcome should happen. This is the baseline for how many should randomly occur. After several thousand randomly dealt hands, the outcomes should match these percentages quite closely.
For instance, in a 7 card run out (Heroās pocket cards plus the 5 on the board) every 33rd runout should hit a flush. So what you need to do is start a spreadsheet and track your hands, the hole cards and the runout. If you play 200 hands per day, you should probably see roughly 40 runouts. 40 per day for 250 days (approx 8 months) will give you roughly 10,000 runouts tracked in your spreadsheet. Then you get the spreadsheet to count up how many flushes you made. If itās every 15th-20th hand you are getting too many flushes and something MIGHT be up. If itās every 50th hand or less, you are not getting enough flushes and something MIGHT be up. Although 10k hands might not be a big enough sample size. 20k would be better. 50k would be good.
This is how you prove it. This is how the RNG testers who issue the certificate do it (although with millions of hands) instead of 10k.
If all that you (and a host of other conspiracy theorists - to put it mildly) are going to say is, āI donāt need to do math, I can see it,ā then you have nothing. You cannot claim mathematical shenanigans if you do not understand the math, and certainly not if you just arenāt willing to do the math.
Nop you might say, āwhoās going to do all that work?ā Well, people do. Of course itās a lot easier to do on a site that letās us download hand histories (cough-cough-Replay!!) and run them through a tracker program, but you can do it nonetheless, and that is what you must do if you want to continue this argument. This is what several of us have told you for weeks/months and it is the one thing that you refuse to do.
You say that quads happen waaay too often, but you havenāt actually counted any quads, let alone tracked them against normal distribution numbers.
Go do your math.
Iām not guessing. There are more efficient ways to do it, but theyāre all equivalent.
Itās not supposed to be the same as shuffling and dealing. It represents the ideal case, where there is zero bias and all possible run outs have equal probability. Itās just used to produce the ideal odds to measure against.
If you actually used a RNG to shuffle and deal cards to generate odds and you found a difference between your results and Replayās, you wouldnāt be able to tell if your RNG had a bias or Replayās did, so thatās not useful.
1 Like
You play around 200 hands. How many do you fold?
How to find all 5 card hands (Flop) in a 52 card deck without RNG (Random Number Generator)
Code in Pascal, Lazarus Free Pascal Compiler.
*** TOP SECRET ***
Output:
Number of 5 card hands = 52 x 51 x 50 x 49 x 48 / 1 / 2 / 3 / 4 / 5 = 2.598.960
The FOR command:
Output:
It starts here at 1 and goes up to 10, step 1.
I can find on this way all number combinations what can come on Lotto too, 6 out of 49 here in Germany, 13.983.816 different.
Pascal code is easy to read, Pascal is much faster to learn than C, Pascal projects are faster created than C projects. I saw Basic Code (most the same) first time in age of 18 years, my brother did it, and I understood. To read/write Pascal code is easier than to read/write English text, lool.
To find what I have on all this hands (Pair, Flush,ā¦) is not so easy. My function to do this has a length of > 300 lines of code. It can read (evaluate ?) millions of hands in 1 sec. Fast enough for me.
The Deck, sorted:
I could use random dealt cards/hands too, but there is then the luck factor on it, and I have to look how many hands I have to use to get a correct result. Here can be only 2.598.960 different hands, and this is fast to calculate and the result is perfect. On 7 card hands is most the same, variable f and g needed to create them..
If I want to know the win chance of AA vs 8 players, then I better use random dealt hands, because the number of all what can be (hole cards of the other players, table cards) has sure a length > 15 digits. Much too many. On 100,000 random hands it starts to give me good results, very near at the numbers for the chance what I can find on poker websites too. Standard RNG, not certified. The RNG of another Compiler, 20 years old, do the same, all OK on it.
I heard the words āTexas Holdemā first time 2017. I looked what this is, the rules. Then I had the idea to play it. I programmed me then my own poker game for this, fixed limit, had many to learn. Later I started then to play online bingo poker, won from begin on.
Other peoples like Crossword or Sudoku, I like programming.
3 Likes