2 pair

[New2vp]

I did see the Beat Vegas TV show a few years back (travel channel maybe?) of the programmer that triggered jackpots on slot machines by making a particular sequence of coin bets.   But there was another such show that showed a more direct exploitation of the random number generator being used.Another way to look at a random number generator is to think of it as a long, long, long list of numbers that has a beginning and an end, which marks the start of a new cycle where the "random" numbers would repeat the previous cycle.  The computer does not keep this long list in its memory, but calculates the next number in the list from the previous number or a few previous numbers, often using fairly simple calculations, such as multiplication or taking the remainder after a division.  Your place in the list is often determined by what is called a "seed", a number that the programmer supplies or often some function that the current time being kept by the computer clock.The show that I recall involved the selection of keno numbers.  This was a 2-person operation with one person on the keno floor in radio contact with a programmer in a room in the hotel with a laptop.  By using a computer and looking at the last two or three keno draws, the programmer was able to find where in the list those particular sequences occurred and could identify what numbers would likely follow.  Then he would signal his confederate on what numbers to play.  Even here the programmer could not know for certain which numbers would come up, since there were many places in the RNG's list that contained places where the particular previous draws occurred.  However, eventually the choices of places on the RNG list were narrowed down and the team got a winner.  I think they were found out because the winner used falsified ID to claim his prize.One solution to this type of scheme is the practice of continual cycling through the RNG list until an event like pressing a button occurs.  If that had been the case, the programmer's method would have been not appreciably better than blindly guessing at the next keno draw.  Though I do not know what is used by various chip manufacturers, some of the current RNGs used by scientists for simulation have a cycle so long that it could spit out several billion numbers each second from now until the current age of the universe and still not deplete the list.  Even those of us who have worked with and understand how RNGs work should be hesitant to ascribe patterns that are noticed by the human mind solely to "selective memory," which is why I started my last post saying that I believe that Lizzy has in fact seen 2s and 3s more than any other combination.  She's a long time contributor to the forum with a history of being sincere, funny, and reasonable plus I remember her having a pattern of unreasonably big hits on Quick Quads on successive days when she regularly seemed to get Quad aces and kickers dealt to her.  When competing against her, I also thought there must be some sort of pattern in the cards dealt to her that were different than those dealt to me.However, if this message was posted by a few other members (including those recently suspended again, but not naming names),  I would possibly not be so charitable in my beliefs.  But anybody keeping track will see one particular combination a lot more than they will see others and it has nothing to do with selective memory.On the other hand, of course, selective memory certainly does happen.  There are highly unlikely things that do occur and are noticed.  That is part of the fun of video poker.  If you noticed the hit in the current contest by BIGBOSS, he not only got the requisite dealt royal, which seems to be necessary to win so many of these on-line contests, but his dealt royal was sequential (TJQKA of clubs), increasing the odds from 650,000 to 1 to nearly 78 million to 1 (or only 39 million to 1, if you consider sequentials going both ways like AKQJT and TJQKA).  Why doesn't that ever happen to me?  I'm sure many of us can ask ourselves that question.Over any finite time period of a few thousand deals to a few million deals, even with careful record-keeping (which avoids any selective memory problems), something unusual is bound to happen.  And if you look at what happened after the fact, you can often find something with pretty long odds of happening under an assumption of randomness.  Thus most valid tests for nonrandomness would (1) state the suspected deviation from randomness first, (2) determine how many observations will be made in the future, disregarding what has been seen in the past, (3) determine specifically what percentages or statistics you will use to make your decision, and then (4) determine how unlikely it would be to see a result as extreme as the value that you got when running the experiment.Of course, it's a lot more fun and less work to just let your mind wander and wonder what's going on, as we hit the buttons for hours on end.  As challenging as video poker is, it doesn't engage our entire mind, so we have plenty of time to ponder theories about what's going on with the machine we're using.

[cddenver]

Nice post!  I remember Lizzy's hits, I think dealt Aces w/kicker in TDB on consecutive days.  I've played QQ's on the site pretty much daily since the end of 2007 and in that whole time I've been dealt Aces w/kicker just once (DDB).
 
Lizzy said that she'd noticed this in her play on the site as well as for-real play (different RNGs) and over a period of time, so I think this would be more of a case of "selective memory", though that might not be the best term to use.  "Individual perception" might be better.
 
Let's do a little thought experiment.  Suppose we could set a machine to deal/draw the same cards over a run of (say) 5,000 hands, and got a group of VP players together who played exactly the same strategy.  Each player sits at the machine and plays out 5,000 hands.   No "oopsies" are allowed, or hunches that affect play, so each player ends up with exactly the same score, same number of quads, straights, flushes, etc.  No player is allowed to actually take notes during the test of what's coming up during the run.  After each player is done, they're asked one question - over the course of the test, did they notice anything unusual about this run of 5,000 hands?
 
In her run of 5,000 hands, Lizzy might mention the frequent appearance of two pairs with 2's and 3's.
 
In my run, I might not have noticed an unusual number of two pairs with 2's and 3's, but might say that I noticed more or fewer Aces than usual.
 
In your run you might not have noticed either, but might mention more or fewer dealt 3oak's than what you'd expect.
 
Oej might mention more or fewer dealt 3 to RF than normal.
 
Any testers who were primarily Deuces players might notice runs or non-runs of two's that others didn't pick up on.
 
Etc., etc., etc.   Some events during the 5,000 hand run we'd all notice, such as consecutive dealt Full Houses that might have come up in that short run.  On others we'd differ based on our individual expectations and past experiences.  A mathematician might examine the 5,000 hand run and say that we're all correct...or even all wrong.    

[EDC1977]

CD; still random or noticeable coincidence? I think new2vp and I watched the same show as they put it this way, with a slight edge over the house per keno card, numerous hands would increase the edge by that much. It's been awhile since they took away our Travel channel but I seem to recall it being a video keno machine and not parlor keno.

[cddenver]

CD; still random or noticeable coincidence?
 
I think both, Ed. 
 
Thanks to you and New2vp for the TV show info - I think I was combining two different stories.

[faygo]

new2vp, I am very close to understanding your dissertation on randomness.   Help me with this: When and how often does a "seed" number occur. When we first deal or does it select the "seed" again while we are playing ?  I might add that while playing in the current contest (DWB) 2,7,7 (while I didn't keep a physical count) showed up an inordinate amount of times.  In TDB it has been a pair of nines.  I refer to these occurances as the  " card du jour" of the day. When playing quick quads on the site when I am dealt a non-suited K,Q,J within 6 hands I will get a good hit.
 
I need to get a life.

[New2vp]

Lizzy said that she'd noticed this in her play on the site as well as for-real play (different RNGs) and over a period of time, so I think this would be more of a case of "selective memory", though that might not be the best term to use.  "Individual perception" might be better.Might be, but we cannot know for sure.  Agree that this tendency is strong in most of us; but Lizzy asked us NOT to tell her it's selective memory, so I thought we should honor that wish and look for other explanations.
 
Let's do a little thought experiment.  Suppose we could set a machine to deal/draw the same cards over a run of (say) 5,000 hands, and got a group of VP players together who played exactly the same strategy.  My favorite type of experiment.  At least, since I've got old enough that physical experiences hurt more.  Who knows, a few more years and thought experiments might hurt, too.And, faygo, I know you understand it all.  You like to pretend to know less than you do, but I'm on to you.  You couldn't know as much about medication without a mind like a steel... a steel... a steel...I can't remember the saying any more.  Mine is like a steel sieve.

[damule]

All video poker machines are rigged with gaffs and losing patterns. Only special plays and martingale denomination increases will overcome this. Just ask the bestest guru player in the world...he'll tell ya. Oh, and BTW RS, my real name is Jim .

[Lizzy15]

 She's a long time contributor to the forum with a history of being sincere, funny, and reasonable....Why...thenk ewe.....  I am not worthy.When I started this post, I wasn't looking for any explanations. I think it's just karma that I get dealt this 2 pair combo more than any others.

[oej719]

All video poker machines are rigged with gaffs and losing patterns. Only special plays and martingale denomination increases will overcome this. Just ask the bestest guru player in the world...he'll tell ya. Oh, and BTW RS, my real name is Jim .

Way to bust that dope head chops damule. I love it when that happens. Then I put up his emoticon like this

[cddenver]

 Help me with this: When and how often does a "seed" number occur. When we first deal or does it select the "seed" again while we are playing
 
I believe that the WM said once that the RNG is seeded when a game is loading, but that no reseeding is done during play.  That's determined by the game programmer.  If the RNG were to be started with no seeding, then you would get the same sequence of cards every time you played.