Can you tell the difference between RNG and Actuals? - a challenge

[Tangram]

The idea is that there will be 1000 numbers posted in separate groups of 100 numbers, some will be RNG (not TRNG from random.org) and the remainder will be actuals (numbers taken from a real wheel). The challenge is to identify which groups belong to which category. Many people claim there is a difference, but no-one has ever stepped up to the plate to prove it  until now that is  Mr Chips has agreed to give it a try.

Mr Chips,

It looks as though Herb isn't interested. I can provide and post the numbers, but you might prefer a global moderator or administrator to do it - it's up to you. 

[TwoCatSam]

I'll be looking forward to the results.

[Mr Chips]

Tangram,
 
I should say from the out set that I do trust you to provide the numbers. Can you confirm that the RNG will definitely be PRNG.
Also will you be stating that 3 or 4 for example of the groups will be PRNG or will that give me an unfair advantage?
 
I look forward to the challenge. I have got my complex calculations ready and a back up crystal ball
 
Mr Chips

[rjeaton1]

I must say that I am definitely interested to see how this turns out.

To prevent any controversy on this challenge, you guys should discuss and agree on a few things before the challenge takes place...such as:

What qualifys as PRNG numbers?  - RXtremes numbers, numbers downloaded from a particular website...if so, what website?

How many sets of numbers are you going to use?

The question Mr. Chips just asked (does he get to know ahead of time how many sets of the numbers are PRNG and how many are live table spins)

etc.

Whether or not you guys do the above doesn't matter to me.  I just wouldn't want you guys to end up in an argument at the end of this....

But, as I said, I can't wait to see how this goes...

[winkel]

As I said in my threads: There is absolutely no difference.

[rjeaton1]

As I said in my threads: There is absolutely no difference.

I'm of the opinion that there is absolutely no difference as well.  I think that if a system works on RNG it will work on a live wheel and vice versa because the distribution of numbers is the same.  I believe that a good PRNG algorithm produces results identical to a live wheel...this is the reason that I can't wait to see how this goes...I would love to be proved wrong.

[Tangram]

Thanks for the interest guys. I too, am of the opinion that there's no significant difference, but it'll be interesting to see how Mr Chips performs. Mr Chips, I was planning to use 3 sets of PRNG spins. Whether this makes things harder or not I've no idea, my intuition tells me that it shouldn't matter, if you have a way of identifying them then it shouldn't make any difference what proportion of the sets are of one type or the other. 

The actuals will come from spielbank wiesbaden. The RNG spins will be generated from a programming language I use. If you're happy with this let me know and we'll start. Shall I post all 10 sets of numbers in one go? also is 100 numbers in each group enough?

One thing I'm not sure about is how to interpret the results, which is why I was hoping to get Herb's input on this. If this was a test of "predicting" the numbers (like a system test) there are techniques for telling you how "significant" the results are. ie; how likely it is that your results would have occurred due to chance alone, so if the results fall within certain parameters then you haven't really proved anything, because this could have happened whether you were using a system or not. But in this case it's not the same scenario; there are two "populations" (RNG and actuals) and the task is to match samples with populations. So if the sets of numbers from 1-10 are 1(RNG), 2(Actuals), 3(Actuals), 4(Actuals), 5(RNG), 6(RNG), 7(Actuals), 8(Actuals), 9(Actuals), 10(Actuals)
and Mr Chips' final decisions are that:

set 1 (RNG) are Actuals 
set 2 (Actuals) are RNG   
set 3 (Actuals) are Actuals 
set 4 (Actuals) are Actuals 
set 5 (RNG) are RNG 
set 6 (RNG) are Actuals 
set 7 (Actuals) are RNG 
set 8 (Actuals) are Actuals 
set 9 (Actuals) are Actuals 
set 10 (Actuals) are Actuals 

There are two ways of being wrong; you could identify an RNG set as Actuals or an Actuals set as RNG. The results above give a "score" of 6 out of 10, but you have to take into account that there are only 3 RNG groups. Another point which just occurred to me from doing this is that if Mr Chips knows that there are a fixed number of RNG sets then this is going to affect the results. If you look at the above, at set 7 there are 3 RNG sets identified, and whether this is correct or not is irrelevant, but at this stage Mr Chips would know there are no more sets of RNG spins because the quota has been used up! This should not happen, so it might be a better test to not tell you how many sets of RNG spins there are. There may be 10 sets, or zero sets, the challenge is supposed to be about identifying them, so you shouldn't be distracted by any info which might skew the results one way or the other.

[Mr Chips]

Tangram,
 
I think we are close to an agreement on the conditions for the test. I agree with the PRNG from the programming language you use
and the actuals from spielbank.
 
My preference is for more than 100 numbers in the group and therefore can I ask for 234 numbers, yes I know a strange request,
but if it's possible that would suit me.
 
I agree don't say how many of the 10 groups are PRNG or actuals for the reason, if I am undecided on a group, which I mentioned
in a previous post about the "mirror effect" on groups of numbers, then I will leave it to one side and decide at the end of the test,
or even before, as I won't know if it's PRNG or actuals.
 
As I have requested more numbers, perhaps we could do it one group at a time and I will try to give a reply as soon as I can and you
can award your  or 
 
Mr Chips

[bombus]

As I said in my threads: There is absolutely no difference.

In my humble opinion, at the end of each sample of 100/234,etc spins (too small really) there will be minimal if any identifiable difference between the two. But the wheel scatter patterns that develop for each as each number is considered may well show different measurable properties.

[Tangram]

Bombus,

I think you're right about the scatter patterns. I've no idea what methods Mr Chips is going to use or whether 234 spins is enough, but he seems fairly confident. Perhaps he will share his knowledge if the test is successful.

Mr Chips,

Ok, let's get the show on the road!  Here is your first set of numbers (234) :

Set #1

14

20

10

35

13

13

14

8

12

34

36

12

2

28

0

23

0

20

18

0

20

11

20

9

30

15

31

3

7

21

4

35

20

31

13

1

5

17

14

31

22

28

34

32

15

31

25

36

3

13

16

23

3

5

28

5

15

33

25

35

31

2

4

14

22

25

8

20

5

26

18

5

3

31

22

5

32

29

18

15

22

23

27

3

3

23

0

28

10

10

11

29

2

24

8

23

3

18

9

24

20

31

31

17

31

36

5

26

11

33

25

3

12

32

4

21

17

8

35

1

34

23

35

3

27

35

2

22

34

1

29

30

17

2

33

34

31

17

35

3

4

2

10

26

31

13

5

24

36

25

6

29

3

23

6

33

0

27

4

2

28

14

5

25

33

17

5

14

0

12

34

21

11

0

34

10

21

27

10

12

32

10

25

6

35

28

20

1

29

35

30

29

29

27

26

10

21

6

2

7

31

3

10

35

15

20

36

0

17

31

31

9

21

7

36

24

28

19

22

10

1

20

8

6

16

22

34

2

7

10

20

11

2

13

[Mr Chips]

Let's break the ice set 1 = PRNG
 
Mr Chips

[Tangram]

Sorry Mr Chips, they were actuals.

Set 1 (Actuals) identified as PRNG 

Next set:

Set #2

9

5

8

24

30

23

10

28

8

2

6

8

7

19

35

8

2

11

11

5

17

34

30

25

20

8

8

2

24

31

1

31

30

23

7

14

19

25

25

30

5

31

28

3

35

24

16

10

31

8

1

8

17

23

22

11

36

13

27

10

35

4

28

18

13

21

2

2

6

11

23

15

9

30

20

1

13

34

31

6

25

2

7

27

28

1

11

10

31

8

26

30

28

11

5

17

14

6

4

24

26

13

32

3

34

33

27

18

26

22

25

7

1

17

12

5

8

32

1

14

3

5

27

4

24

18

19

2

34

16

21

28

12

30

29

10

31

14

12

28

1

34

24

2

15

3

28

16

30

25

19

26

29

36

5

14

8

2

29

15

16

30

36

21

33

8

31

12

21

18

25

10

9

22

15

5

14

23

30

21

27

33

14

18

33

15

18

34

15

11

24

36

34

33

32

35

34

5

31

27

17

27

35

12

28

11

12

35

12

5

26

10

15

33

22

6

10

16

20

32

8

11

18

20

19

28

6

14

26

30

27

34

29

20

29

23

5

22

29

30

7

12

14

16

21

28

36

24

28

13

13

9

9

7

18

13

15

21

21

18

8

15

32

29

5

29

21

14

12

28

12

36

25

19

9

32

4

24

6

2

11

26

27

9

2

15

35

35

12

24

1

2

18

13

13

16

22

19

30

5

12

16

26

26

33

18

11

18

27

34

1

9

9

24

23

34

20

27

35

28

35

1

20

29

23

15

36

32

9

15

36

5

18

14

[Mr Chips]

Tangram,
 
I am genuinly surprised by that result, but will carry on to set 2.
 
Mr Chips

[Tangram]

Mr Chips,

I've been using a set of spins from spielbank but I wouldn't be able to say which table they're from or when they were recorded (they have been merged into one file), so if you like we can re-start the test and I can get fresh spins from wiesbaden making sure to note the details (table no and date) that way everything can be verified (at least for those sets which are actuals). I'm not suggesting that you think I may be "cheating" or anything like that, it's just to keep things as "above board" as possible. 

[Mr Chips]

Tangram,
 
Don't worry I will carry on with the test set 2. In my estimation set 1 conformed to PRNG. Wouldn't it be funny if all the  results I gave for PRNG
were actuals and visa versa, I would definitely have to book into a rest home for people suffering from a numerical breakdown
 
Mr Chips

[Mr Chips]

-