PRNG, TRNG, Actuals

[Mr Chips]

I admit I am a number addict lol and in addition to roulette numbers I have a great interest in Prime Numbers.
 
I haven't found any solutions to the many Prime Number problems, that would upset the mathematicians, but
I have discovered some interesting patterns associated with those numbers and it can't be a bad thing looking
at Prime Numbers not from a mathematicians perspective, but from someone who has an interest in numbers
generally and likes to design roulette systems.
 
The point of mentioning Prime Numbers is that I saw a very interesting programme on BBC tv recently, which
referred to prime numbers and that a discovery has been made, which shows a relationship between the theoretical
world of mathematics and the characteristics of matter and energy.
 
The frequency of prime numbers apparently is the same as hitting a crystal and the vibrating waves that
show on a screen match each other. Really incredible.
 
There have been endless discussions about random numbers and the types of numbers and can we distinguish
between Pseudo random, True Random and Actuals.
 
The first question is it possible to distinguish between say TRNG, which you will find at Random.org and PRNG, which
I understand can be generated from a computer.
 
If it was possible to distinguish between the two RNG's, would it help us in any way to try and understand random
numbers, even just a fraction of comprehending the random complexity? Also would it help in any way to devise
"intelligent" systems?
 
Do PRNG's for example produce a complex wave pattern very different from TRNG'S that we could identify.
 
If anyone finds all this interesting and can produce approx 100 PRNG numbers I will try and discover certain characteristics,
which I have noticed before about such numbers.
 
Please only PRNG and their source.
 
Mr Chips

[Ulysses]

I watched that BBC program with Alan Davies but I missed that wave in primes and crystal correlation bit, getting a cup of tea, and came in at the tail end of it. Caught something about Rheiman graph and searched for it with no success. You don't know what the graph is do you or where I can find it on the net. Will get you 100 PRNG results that's a fair swap lol. Uly

[gizmotron]

on mouseup
   repeat with i = 1 to 100
      put random(37) & ", " after field "hotNumberList"
   end repeat
end mouseup

14, 29, 33, 9, 29, 7, 35, 5, 12, 22, 7, 32, 29, 31, 1, 12, 1, 16, 26, 20, 12, 2, 2, 27, 26, 8, 23, 26, 12, 36, 15, 36, 9, 5, 8, 16, 36, 32, 32, 10, 28, 24, 16, 28, 21, 1, 30, 31, 18, 35, 12, 28, 20, 10, 15, 28, 16, 6, 20, 3, 21, 21, 23, 29, 13, 30, 11, 32, 22, 11, 23, 10, 3, 13, 3, 35, 17, 28, 10, 19, 33, 25, 13, 29, 32, 16, 12, 12, 8, 20, 14, 11, 21, 29, 33, 9, 23, 18, 24, 8,

These 100 spins were created with a very modern RNG that reset the random seed by checking the number of milliseconds since 1970.

[Number Six]

Bear in mind that TRNGs can produced bias in the binary digits if they are not filtered. You would have to crunch a lot of numbers to distinguish between true and pseudo random.  But as the states of PRNGs and TRNGs are different and the numbers are generated by different processes, they will create different sequences of numbers that can be told apart. In a nutshell, the TRNG results will be "better". However, if the TRNG is flawed, it would corrupt any study and render the exercise worthless.

[gizmotron]

...However, if the TRNG is flawed, it would corrupt any study and render the exercise worthless. 

Modern PRNGs start to repeat after one half a million spins. The programmer needs to reset the random seed before that occurs. For short sets like 1000 spins it doesn't matter. If you reset the random seed with a secret set of random seeds every 100,000 spins then it should be bomb proof.

[Number Six]

I'm talking about TRNGs, or hardware random number generators, not software PRNGs. If you are trying to distinguish between the results generated by a TRNG and a PRNG, and the TRNG was flawed, the results would be corrupt because any patterns in the TRNG outcomes would be formed by bias.

[gizmotron]

But as the states of PRNGs and TRNGs are different and the numbers are generated by different processes, they will create different sequences of numbers that can be told apart. In a nutshell, the TRNG results will be "better".

How do you know they will be better for roulette? If you are running a mechanical rule based betting system for millions of spins then perhaps that might be true. If you are following charted trends, patterns, and dominances from past spins then perhaps it doesn't matter. The context for usage matters. I for one believe it does not matter for short termed runs, like a real playing session at a real casino. That holds true for using a fair RNG or real spins too. In my own experience I see the spins as the same, producing the same kinds of swings and changes.

[Mr Chips]

I watched that BBC program with Alan Davies but I missed that wave in primes and crystal correlation bit, getting a cup of tea, and came in at the tail end of it. Caught something about Rheiman graph and searched for it with no success. You don't know what the graph is do you or where I can find it on the net. Will get you 100 PRNG results that's a fair swap lol. Uly

If you go on the BBC tv website and put Horizon in search and scroll down a bit
you can see it again. I think it's well worth a second view. I bet a few
mathematicians were shocked when they saw the match, with the waves from
the crystal.
 
If you do happen to have a spare 100 PRNG that would be great.
 
Mr Chips

[Number Six]

Gizmotron,

It's a generalisation, whether or not TRNGs are better for roulette than PRNGs is a moot point...but I'm telling Mr Chips that basically any study of random and pseudo random under these circumstances would probably be a waste of time because there are too many values and variables to consider. The study could be corrupted and you wouldn't even know it. I really don't think there any exploitable and/or detectable patterns to be found in a roulette RNG because you're dealing with such a small amount of numbers. So, you're right that it probably doesn't matter, but why settle for pseudo when you can have the real thing?   

[gizmotron]

...but why settle for pseudo when you can have the real thing?  

There are people that hang onto magical beliefs and superstitions. They think that they can tell the difference and use that as an excuse for their systems and methods doing better or worse. They are not experienced enough with real and RNG enough to know that each time a new random sample of spins is used they act naturally, like the basic nature of randomness. So they blame the string of numbers. It's a mistake to blame the random source. They should know that wild swings are normal. Why should a stretch of spins be ready to go just the way you hope or expect it to?  If you use RNG to practice and gain experience you can use that experience to play a real wheel in a real casino. I would count on playing experience and less on superstitions.

[Number Six]

You can tell the difference between true random and pseudo...the point being that pseudo isn't random, it's a fake imitation and by nature imperfect. But you'd have to crunch millions and millions of numbers and then keep re-analysing them, and any patterns that occur in pseudo aren't going to materialise over 5k or 10k or even 1million spins. You might start to see anomolies after a few hundred thousand, but when was the last time anyone played 300,000 consecutive spins on a roulette RNG? In roulette the sequences of numbers in play are simply far too short to be able to draw any conclusive findings. It would be better to join the numerology section and explore the issue of randomness there.

And Gizmotron, I agree that systems should be tested, put into practice, against an RNG, but only an independent hardware-based one such as random.org, certainly not the casino's play money wheel. If a system can hold up fairly well against true (and honest) random, then it'll be able to handle a real wheel.

[Mr Chips]

Gizmo,
 
Thanks very much for the numbers they are what I would expect from PRNG. I will list the numbers and explain
my take on them. It is not always easy to identify the difference between PRNG and TRNG and Actuals, but a friend of mine
was working on a number problem and he was relying on being given TRNG, but as it turned out the person instead gave him
PRNG and he couldn't get the same results as he got previously.
 
When I design systems I often use 'sections', which I used for the 4Selecta system. For me they are essential and I wouldn't
have been able to make sound constructive systems without them. They have many uses and one use is particularly interesting
and you may say is impossible and therefore dismiss it. However I have found it useful and the saying speak as you find fits the
case here.
 
[table=,]
sections,fcst,sections,fcst
14,,35
29,,12
33,7,28,11
---,,---
9,,20
29,,10
7,11,15
---,,28,9
35,,---
5,,16
12,13,6
---,,20,13
22,,---
7,,3
32,11,21
---,,21
29,,23
31,,29,14
1,6,---
---,,13
12,,30
1,,11,7
16,,---
26,11,32
---,,22
20,,11
12,,23,11
2,,---
2,,10
27,,3
26,18,13
---,,3,8
8,,---
23,,35
26,13,17
---,,28
12,,10,17
36,,---
15,12,19
---,,33
36,,25,7
9,,---
5,,13
8,23,29
---,,32,6
16,,---
36,,16
32,,12
32,21,12,10
---,,---
10,,8
28,,20
24,,14,10
16,8,---
---,,11
28,,21
21,,29
1,,33
30,,9,13
31,,---
18,12,23
---,,18
,,24,14
,,---
,,8
[/table]
 
You will see that for example 14,29 (5 +2) digit sum to get a forecast of 7, the actual number that came in was 33. I use these
forecasts in various ways and in TRNG and actuals there is as you would expect, a fair distribution where the forecast is correct,
just misses by +1 or -1 or appears as the next number. For some reason whenever I have used PRNG I get few correct forcasts
and many more +1,-1 or is the next number!
 
The above shows :
 
12,13
 
13
---
12
 
14
---
13
 
11
---
10

10
---
11

There is perhaps an explanation that I have been unlucky whenever I have used PRNG when the above occurrence has happened,
but I have yet to come across PRNG numbers, that produce the same results as TRNG or Actuals.
 
Mr Chips

[gizmotron]

And Gizmotron, I agree that systems should be tested, put into practice, against an RNG, but only an independent hardware-based one such as random.org, certainly not the casino's play money wheel. If a system can hold up fairly well against true (and honest) random, then it'll be able to handle a real wheel.

So if I include an RNG from my desktop application for running 50 to 300 spin competitions, will that be a problem trusting it to be random?

I'm also adding in the ability to use real spins for competitions. That should take care of those that think it's not random.

I just don't want losers blaming the randomness of the numbers for their losses. It has to fall on the player.

[Number Six]

Gizmotron,

It's an interesting question but if you're only running short cycles of numbers then, no, it really wouldn't matter that those numbers are generated by a software PRNG. It would, of course, be better if you could take numbers from random.org and run those through the application to test the systems in the competition, if only just to be able to say that the competition was founded on true random...and no one can refute that when they lose.

People often blame the RNG when their system fails, and I myself have fallen into this trap once, but it's more a matter of suggesting that the RNG has inbuilt pattern recognition software and so is rigged by the casino. PRNG isn't a problem for testing, it's just that you can't rely on the test results to be purely accurate. 

[gizmotron]

Roulette Ride 1.2 will include the addition of loading real spins or spins from random.org for competitions. The two methods will be line delimited and comma delimited text files or clipboard based. RR1.2 will automatically encrypt them for the sponsor of that competition. Then the sponsor can tell the users what type of spins were used. My guess is that real spins will be the only ones that everyone will accept for any competitions. It's purely baloney. Their betting methods are far more flawed than any true form of randomness perception. You watch. Somebody will post RNG spins and claim they were real. Several people will say they knew that they were real spins. People play better when they think the spins are real, I guess.

I know this because I read it in the entrails of a two foot long baloney. 

[gizmotron]

-