Online Poker Forum - Patterns ... and the flaws of Code

Kardsh4rk · Message Board Junkie Joined: 02 Sep 2006 Posts: 1279

Random is supposed to be just that � random. I've posted before about getting the same cards dealt to me 3 out of 5 times, etc. ... and wondered about the statistical probability of it. Now, I see other players are also finding it happening to them as well, so I thought I would make a post about the randomness and the patterns that I have noticed. Please note ... I am not saying it's rigged, so don't fire off a thread about 'it's not rigged'.

First of all, one of the biggest patterns I have noticed is card strength. On day X, I'll typically play 2 or 3 SNG games. During the course of the games, I'll be dealth strong pocket cards just as often as I am dealt weak cards. I'll see high pocket pairs, or solid starting hands that quite often, when played right - will take a pot. I'll also see marginal hands (8,9, 10,J) etc., and I'll catch. Of the 3 games played, I'll usually win 2 and place in the 3rd.

The next day I play - day Y - I don't see a single pocket pair and the hole cards are rags pretty much straight though, until I hit something like AK, or AQ ... and when played (with a pre-flop raise, etc.) ... they lose, usually to a set or draw of some sort. Out of 3 games, I'll bounce out 1st or 2nd. My game play hasn't changed between the days, how much I bet hasn't changed, etc. The only thing different is the strength of the cards.

After several rounds like this (X, Y, Y, X, Y, Y, etc.) ... I've adjusted my play, so when I see in one game that i'm not being dealt decent cards, it's my last game for the day. I'll take a day or two off, then come back and BAM, I'm catching cards again.

Now, onto the second (and even more obvious) pattern: Same pocket cards. Sure, there is a statiscally probability that it could happen. But when more and more players are stating they are catching the same cards 3 out of 4 or 5 times in a row ... that goes beyond a statistically probability and moves into the realm of impossibilities. Imagine yourself standing in a field in a lightning storm. Sure, you could be hit by lightning - the odds of it happening are actually pretty good. But, to be hit twice, the odds on that are very, very minute. To be hit 3 times � well, God must not like you and he's taking it personally.

I know there is people on the forum who will say, "it is random, and stuff like this can happen." ... but, when you are heads up against another player and you have a 99% chance ... and they catch their 1% ... are you just as calm and casual in saying, "it's random, stuff like this can happen?" - Or, do you lose your mind?

Computer code (the dealer) is just that ... code. Code can misfire (look at all the Windows updates - or even the FTP updates). Just because it is a computer - and supposed to be random - doesn't always make it so. -- Just my 2 cents.

2 high flush

junkbutton

JazzOne

junkbutton · Posted: Tue Dec 05, 2006 11:50 am Post subject:

Rumor has it that people have done this on the 2+2 forum, with no evidence of funny business... I haven't read any threads myself though to confirm this.

Paprika · Full House Joined: 09 Feb 2006 Posts: 194

Only thing I've seen that is strange to me is every time I play a HORSE S&G, during every single Razz round I always get a lot of premium pairs. I never get them during the stud hi round at all. Happens every time for me. Coincidence I suppose but I have played many and it is a regular thing for me. I almost expect it now.

CitizenCain · Posted: Tue Dec 05, 2006 1:32 pm Post subject:

Yeah, cardshark, you're suffering from selection bias - you only remember the "memorable" events (AK two times in a row, pocket rockets then QQ, etc) and not the unmemorable ones. Like that time you got 83(os), 95(s) and T6(os) in a row, which is also highly statistically unlikely.

jjpregler · Forum Pro Joined: 17 May 2006 Posts: 1452

Did you see last weeks WPT? I forget his name, but some guy got KK 3 times in the first 10 hands. It happens live too. I remember seeing 10 5 5 times in two rounds of the table last week. It happens. That's part of random. You only have 52 random cards to choose from so repeats are bound to happen.

griffinlord · Posted: Tue Dec 05, 2006 1:42 pm Post subject:

As others have mentioned, the human brain has evolved to detect patterns. The detection of patterns is a huge survival advantage in a hunter-gatherer or scientist.

But, our brains are sooooo good at detecting patterns we see them even when they don't exist. Ever see constallations in the night sky? That's pattern detection backfiring. There are no patterns in the distribution of stars, but we see patterns anyway.

The ability to detect patterns when none exist is one of the reasons that scientists use statistics to sift through their observations.

That leads us to: what is randomness?

Randomness is unpredictability. As simple as that idea is it is beyond most people's skills to comprehend. When people are asked to generate random numbers they always impose a pattern, especially too few repeats.

That is, asked to generate a string of random digits 0-9 a human will only rarely include "333" or "777" in the string. But strings of 3 digits should happen one time in 100 if there is true randomness. Heck, people will rarely include repeats ("22" or "55") in their "random" digits. The absense of those apparent patterns is in fact a pattern.

Statistically the lack of randomness shows up as an ability to predict. If there is a pattern you will be able to predict what comes next with greater than chance accuracy. If a person is generating a string of random digits and the current digit is "2" I know that they will very seldom choose "2" or "3" as the next digit, so I can in fact predict their next digit with an accuracy of ~12.5% which is better than the 10% I should be able to get with true randomness.

But with true randomness "2" and "3" are back in the mix and my prediction drops back to chance levels.

If you can predict future hands/cards with any degree of accuracy then you have found a profitable situation--and evidence of RNG failure. But just because you think you can predict doesn't mean you can. Remember, we humans see patterns even in the absense of patterns.

This finally brings us to "long-term" versus "short-term." If you have access to a statistics book with a table of random digits you can see short-term and long-term in action.

In such tables there are typically lines of 40 - 60 digits (0-9). For each line you can count the number of "0's" "1's" and so on. In a given line of the table (the short-run) the number of "0's" will usually be close to its long-term percentage. But sometimes you'll get a row with no 0's and sometimes you'll get a row with as many as 20% 0's. That is short-term variance.

(Aside: humans will generate digits with too little short-term variance, they will almost always produce 40 digit strings with 3-5 of each digit.)

After counting up the whole table, the percentage of each digit will be very close to 10%, probably between 9.5% and 10.5% or even closer depending on the size of the table. As the size of the table grows the difference between observed and expected values becomes vanishingly small if the digits are truely random. (Note, this could also be true with non-random digits which is why some sort of pattern test is so important.)

UFO1947 · Posted: Tue Dec 05, 2006 1:43 pm Post subject:

I have had one of my worst weeks playing poker and the funny thing is I only suffered 1 bad beat, it turned out to be my play. I was trying to buy too many pots.

today I was in a SnG where nothing good was coming and when i did flop a straight he gets a full house on the river. the next SnG after that everyone was getting pocket pairs, one guy had AA 3 times!!!!

S&%t happens, i know your concern though, before i put money online I asked the same questions.

griffinlord

JazzOne · Posted: Tue Dec 05, 2006 1:59 pm Post subject:

There is a difference between "random" and "stochastic." I tried to look these terms up on wikipedia to initiate a discussion on the topic, but the articles were unhelpful. I took a class on theoretic physics in college, and one of the things we talked about was what it meant for quantum events to be non-deterministic and how that relates to randomness and stochasticism. I honestly cannot remember the sublteties that differentiate the two. Would anyone like to clarify these terms?

Regarding the class I mentioned above, the professor gave us a very strange assignment. Everyone was randomly given one of two assignments. One group was supposed to flip a coin a couple hundred times and record the results. The second group was supposed to fabricate the results without actually flipping a coin. Without being told, our professor was able to accurately determine which assignment each student had been given. How did he do it? Because the people who made up the results did NOT include enough repeats. They failed to realize that a long stretch of repeats would probably occur if the coin was flipped that many times. So, when you lose 8 times in a row with a pp against overcards, it does NOT mean that FTP conspires against short stacks or anything like that. It just means you caught one of those nasty stretches of repeats.

craigo6x · Posted: Tue Dec 05, 2006 2:09 pm Post subject:

There is also a difference between rigged and flawed. While I do not think online poker is rigged, I certainly think the RNG is flawed. I remember at pokerroom, they published a chart showing the distribution of cards to back up the rng as truly random, and the chart showed an even distribution of cards. But the chart didn't show an analysis of hand v hand. In my opinion the randomness of flopped sets is outweighed by the frequency with which it occurs.

Tell me how many times you had aces against two underpairs and you knew you were going to lose to an underpair set. I don't get that feeling live.

JazzOne · Posted: Tue Dec 05, 2006 2:28 pm Post subject:

KGBlovesOreos · Posted: Tue Dec 05, 2006 2:32 pm Post subject:

Online Poker Forum at Full Tilt Poker