Surely "streaky" is rigged!?!

First I must clear up a misunderstanding about RNG's.
I have fully confidence in RNG's used by ALL software providers. My ONLY worry about RNG's are that they are predictable(not random enough or RNG result can be pre-determined), but that would be a player advantage then! What can happen is cheating software. Ie. the software ask the RNG for a result and the software looks at the result and decide it didnt like it, and then it asks the RNG again and use that result instead. This scenario is indeed possible and is what has happened in the past with cheating software. The RNG has no idea what game(slot,BJ etc) the result is used for, therefore it is pretty obvious that the RNG can not cheat in any way. But in the intermediate SOFTWARE mapping the RNG result to game event, it is possible to cheat. This is another way to cheat instead of asking the RNG for another result, but again you have to blame the software. It is important to distinguish between the RNG and the software.

I have observed that BJ players are most likely to talk about suspicious winning/loosing streaks and blame the RNG/software. And when I play BJ - I often get the same feeling and therefore fully understand why a lot of people feels this, but I have a rational explanation for this. (It has happened at MG software which I am sure is not cheating). Because BJ is a kind of 50/50 win/lose game - players that lose 10 times in a row start shouting rigged. BJ is the best hand and people subconcience believe they won, but it is not really that rare it ends with a push instead and it often does and often several times in a row. Dealer hitting 21 starting with a 3 happens. Dealer having face card is common. Therefore there will be some sequences where dealer starts with a face card a lot more than expected. Also players does not seem to notice their winning streaks in the same way.
So basically because BJ is a low variance game with very few outcomes a lot of seemingly suspicous patters will inevitable emerge all the time. (*)

However getting 5 scatters/5 wild/5 retriggers in a slot game is much more rare but it does happen also and people at not complaining then. People blaming slots of long losing streaks should know better. Slots are designed this way and they are working as intended. This is the cost of being able to hit big win.(and facing a gigantic house egde at 5%+...)


(*) Most famous example is from 'Ramsey-theory' explained simple this way: Draw 6 dots (symmetric in a circle, to make it easy). Then connect ALL 6 dots with eigther a BLUE or RED line. When you are done there will ALWAYS be eigther a BLUE or RED 'triangle' Ie. there will exist 3 dots that all are connected with the same color.
Another way to express is that given ANY 6 people chosen, there will always be 3 that all know each other or 3 that all does not know each other. ('Know each other' is defined so the
both knows the other one). This example shows that given complexity enough, some patterns will always emerge.

Zoozie
 
Last edited:
Incidentally, I just remembered one of the more famous exploits out there, Ron Harris, a computer programmer employed by the Nevada Gaming Control board, realized that the RNG used for some keno software could be predicted, and built a software program to determine the pattern. He and his accomplish were able to pick ALL of the numbers in a keno jackpot after just a few rounds.

Old / Expired Link
 
Incidentally, I just remembered one of the more famous exploits out there, Ron Harris, a computer programmer employed by the Nevada Gaming Control board, realized that the RNG used for some keno software could be predicted, and built a software program to determine the pattern. He and his accomplish were able to pick ALL of the numbers in a keno jackpot after just a few rounds.

This was also the case for some B&M VP machines. They could predict what cards would come on the draw. Not as usefull as the keno prediction though.

I know poker sites are now implemented so draw cards are decided 'on the fly'. (think of it as picked random from the deck instead of top cards, but the random factor now depends on the time you request the cards so you can not predict it)

So whenever any independant party approves a RNG for a software company, it really means NADA! They should test the software instead. English Harbour
probably had a perfect RNG, yet the software was rigged/flawed.

Zoozie
 
Last edited:
(*) Most famous example is from 'Ramsey-theory' explained simple this way: Draw 6 dots (symmetric in a circle, to make it easy). Then connect ALL 6 dots with eigther a BLUE or RED line. When you are done there will ALWAYS be eigther a BLUE or RED 'triangle' Ie. there will exist 3 dots that all are connected with the same color.
Zoozie

I was not precise enough it seems. You have to connect ANY two dots with each other, so you have to draw a total of 15 lines.

So the example shows that a predetermined pattern will always emerge. In BJ you play a sequence of hands and then you examine the data and find
patterns from the data. This is a much weaker requirement and therefore a lot more probable since you can try to find a pattern that match the data.

I have another more 'BJ' like example of patterns that will always emerge.

Write the sequence 1,2,3,4,5,6,7,8,9 where each number is eigther with a red marker or a blue marker.
Then there will always be a set of 3 numbers in the same color on a arithmetic progression. An arithmetic progression is a sequence of numbers that
differ by a constant amount. Ie 3,4,5 or 12,14,16 or 21, 26, 31, 46 etc.

Example: 1 2 3 4 5 6 7 8 9
Then you can find the arithmetic progression 1 5 9 in the same color.
Good luck trying to color the numbers 1-9 so there is no aritmetric progression of size 3 :)
(if you only use the numbers 1 to 8, then it is possible to color them so there will not be an aritmetric progression of size 3).

So what does this have to do with BJ? If you play 9 hands of blackjack and consider them won or lost (push also lost etc.) then there will also ALWAYS be this artimetric progression of won or lost hands. ONLINE BJ MUST BE RIGGED - IT HAPPENS EVERY SINGLE TIME! :thumbsup: Generally the theory says that an artimetric progression of any size (3 in this example) if the sample size is big enough (9 in this example).

Found this online article with more information and not too much math. (picture of nerd included)
Old / Expired Link

Zoozie
 
Last edited:
As an aside, Ramsey Theory is accompanied by rigorous proving and does not, as far as I am aware, have any mention of any on-line casino's Blackjack algorithm no matter how many coloured dots I have joined up.

Interestingly, no online casino as far as I am aware has any rigourous proving therefore it would not be correct to have an acceptance of, for example, a Blackjack algorithm in any online casino as it would be correct to accept Ramsey Theory.

Furthermore, if we are talking mathetical proofs, surely anyone with a healthy questioning of any unproved system (i.e. An online casino BJ) are surely less delusional (in the scientific sense of the word you understand) than those that accept the systems and berate others for not doing so.
 
As an aside, Ramsey Theory is accompanied by rigorous proving and does not, as far as I am aware, have any mention of any on-line casino's Blackjack algorithm no matter how many coloured dots I have joined up.

Sorry, but I dont think you got the idea.

The whole point was when you play 9 hands of blackjack you can ALWAYS
find 3 hands that are all loss or wins that also form the arithmetic progression pattern described.

Instead of the red and blue numbers think of them as wins or loss.
Hand 1 won: make a red 1.
Hand 2 won: make a red 2
Hand 3 lost: make a blue 3

When you have 9 hands, there will always be the arithmetic progression of size 3 in the same color. That this pattern will emerge has nothing to do with blackjack, which was excactly what I wanted to state. Just a message to
those people that are looking for conspiracy everywhere.

Zoozie
 
Zoozie,

From my understanding there are two main types of processes used to generate a base seed.

In laymans terms:
1/ A base seed is extracted randomly from a frequency wave / atmospheric noise.
2/ A base seed is manufactured.

In regard to (1) it's my understanding that this type of seed produces a truly random outcome.

Where as (2) is know as a quasi RNG, where the outcomes are not truly random.

From what I've read RNG's just spit out gobble gook, it's the software that interprets the data and applies it (in this case) to cards, slots, dice or two flies crawling up a virtual wall.


A few Questions if you don't mind?

Which type of seed do casinos generally base their RNG's on? And which type of RNG do casino's use?



Cheers
 
Which type of seed do casinos generally base their RNG's on? And which type of RNG do casino's use?

Sorry, I do not know what kind of RNG the casino software does use.

But I have have heard of another type that uses radioactive decay to
generate the numbers and this is considered truely random also.

It is never the RNG I distrust, it is always the software.

Zoozie
 
I've heard that a popular seed for software RNG's is the system clock measured down to a very small value, like milliseconds. That usually at least has the advantage of changing frequently.
 
Just a message to
those people that are looking for conspiracy everywhere.

Zoozie

But surely the results of the Ramsey Theory would be applicable whether the RNG was random, non-random or the casino software was dealing a crooked game - so, in a sense, it's also a message for those looking to believe any software that is laid in front of them as being totally honest.
 
Rainbow ramsey theory

Hi Zoozie,

in the article in Old / Expired Link Jacob Licht gave this example:

"EXAMPLE (n = 35): 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
This example has a four-term monochromatic (blue) arithmetic progression: 6, 13, 20, 27, with a common difference of 7."

I cannot figure out why he chooses this 6, 13, 20, 27 progression because looking into the data I think the following arithmetic progression is much better here:
15, 18, 21, 24, 27, 30 with a common difference of 3.
It gives me 6 terms here.

Another would be 10, 15, 20, 25, 30 with a difference of 5 and 5 terms.

3, 9, 15, 21, 27 with a difference of 6 has also a better spanwidth.

Do you know why excactly this one?

Franz
 
Hi Zoozie,

in the article in Old / Expired Link Jacob Licht gave this example:

"EXAMPLE (n = 35): 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
This example has a four-term monochromatic (blue) arithmetic progression: 6, 13, 20, 27, with a common difference of 7."

I cannot figure out why he chooses this 6, 13, 20, 27 progression because looking into the data I think the following arithmetic progression is much better here:
15, 18, 21, 24, 27, 30 with a common difference of 3.
It gives me 6 terms here.

Another would be 10, 15, 20, 25, 30 with a difference of 5 and 5 terms.

3, 9, 15, 21, 27 with a difference of 6 has also a better spanwidth.

Do you know why excactly this one?
Franz

I must compliment you in your findings, you definately understood it.

First of all the number 35 is the minimum number such that there will always be an arithmetic progression of size 4. So take the challenge to find a 2-colouring of the numbers from 1-34 such that it does not happen, but a such does exist or 35 would not be the minimum number. It is not mentioned that 35 is the minimum number needed to find one of size 4 like 9 is the minimum number needed to find one of size 3. But it is so.

The 2-colouring in the article is not a good example since there even exist an arithmetic progression of size 5. Maybe he chose a random coloring and did not notice the one of size 5 or believed it was not important as this can also happen. The important fact was that there will always be one of size 4 (at least).

Generally these minimum numbers are damn hard to find and not many are known. To always find one of size 5 (for a 2-colouring) the minimum number is 178. Some years ago
it was not know what the mimum number was if you wanted an arithmetic progression of size 6, but maybe it has been found since.

So
N=3: 9 numbers is needed
N=4: 35 numers is needed
N=5: 178 numbers is needed.
N=6: ??? numbers is needed.


Zoozie
 
Last edited:
an interesting thread so far

firstly i would just like to thank those that have contributed to this thread so far - very interesting even if my brain is hurting a bit.

Much of this thread is about empirical evidence running against detailed RNG theory.

It is clear that it would be really hard to have any sort of satisfactory proof of whether a game has greater varience than what it should. It is also the case that "streaks" are hard to interpret by the player even if they did exist.

I am a BJ and Poker player primarily although i do dabble in VP and slots now and then but when it comes to streakiness I have only really got enough playing time to look at BJ.

If we take the software AND the RNG and take the combined effect of them then impirically there are some sites that seem "as a player" to be streaky.

I will say that 99% of talks of streakiness are about people complaining about losing. I do not see streakiness if it really exists as a bad thing but rather something that the player can exploit.

On some softwares myself and several other BJ players I speak to have experienced apparent streakiness on several sites. When we are in these negative streaks we bet less and in the positive streaks we bet more. Over the course of wagering 20,000 in bets on such a site between 1-20$ the "streakiness" factor does seem on some providers to have a positive player effect. This seems to be the strongest basis for supporting the idea that streakiness is real that some players (myself included) seem to be able to exploit in on certain sites for profit.

If it exists is it something that casinos do deliberately? I would say probably not but whilst casinos etc are often audited for the extent to which they are paying out appropriate amounts I do not think that much of the effort goes into checking that the varience is 100% as expected.

The combination of the RNG and the software that creates a BJ hand is not the same exactly from one software provider to another. If we say streakines does exist (it is highly debatable that there is sufficient proof) then it could be argued that it makes the software "imperfect" but this is akin to a roulette wheel not being 100% balanced. It is something that can only really affect a knowledgable player in a positive way.

Finally as a player i would say that streakiness is a bit like the "god gamble" i.e. if it doesnt exist then it has not done me any harm but if it does exist....

What this means is that if you sense the casino is really streaky then it does no harm to adjust your bets accordingly. It cannot effect your EV in a negative way. If however there is some merit in the notion of streakiness on a small number of sites then you have improved your EV.
 
Yes, it is an interesting discussion.

My point was in the original post (and explained further about three posts down) is that a player cannot say that one casino's set-% game is streakier than another casinos equivalent game - by default you are saying one of them is rigged. Doesnt matter whether you can take advantage of the streaky nature, doesnt even matter if one is actually streakier or it is just perception, doesnt even matter if one of the games is set to payback 200% - you are saying one of the games is rigged "in your opinion".

If I run a totally above board BJ simulator you can show Ramsey in the results, if I run a smart BJ simulator, say dealing smart seconds based on the player betting strategy, you will still find Ramsey in the results. This is what I understand!
 

Users who are viewing this thread

Meister Ratings

Back
Top