Ok here is my system:
This system is based on the actual physical position the ball lands in for each spin. When I set out with this, I was building a system for actual physical roulette, not online computer generated numbers. However, strangely, it seems to work for RNG derived numbers as well.
Start by collecting 23 numbers.
For each number mark the number of pockets it has moved in a clockwise direction on the wheel from the last spin
E.g: 4 to 2 has a count of +2
24 to 21 has a count of -15
We now divide the count numbers into to sets, A and B
Ignore the first number as it has no jump count
The second number's jump count is in set A
The third number's jump count is in set B
The fourth number's jump count is in set A and so on
For each jump count in a set, subtract the previous jump count from it. Thisis the relative count
Eg
No
Eg
No Set Jump Count Relative Count
0
0 A 0 NA
12 B -4 NA
14 A -8 -8
7 B 6 10
10 A -13 -5
At the end of this you will have 10 relative counts each for A and B
Average the relative Counts for each set A and B, so that you will have one average for A and one average for B.
We now play for the next 20 spins (10 for set A and 10 for set B).
Determining where to play is based on the following calculation
x = next Jump count for the specific set
pr = previous jump count for the specific set
m = average for set
x-pr=m
To determine where to play (x), then:
x=m+pr
Eg
Average is 2
previous jump count is 10
x = 2+10
x = 12
We therefore count forward 12 pockets from the last spin.
Note, the last spin will belong to the opposite set, count from this spin, not from the previous spin within the set.
Eg if the last number was 33, we would play 35
I play 5 numbers either side of the count that is determined, so on 11 numbers in total.
I will post some results in the next hour or so.
Ok here are the results of one sequence:
N0 Jump Count RelCount Average Target No Win/Loss
SA SB SA SB Set A Set B
0
0 0
12 -4
14 -8 -8
7 6 10
10 -13 -5
3 17 11
13 14 10
5 7 -10
34 -10 13
31 17 10
6 -16 -6
5 9 -8
35 15 6
34 12 3
33 13 -2
16 -1 -13
18 8 -5
2 14 15
23 11 3
6 -7 16
31 16 5
27 -15 -8
33 11 -5 1 22 Loss -11
17 -14 1 3 6 Win 25
2 -2 -13 1 24 Loss -11
22 -15 -1 3 28 Win 25
7 3 5 1 9 Win 25
25 13 9 3 5 Loss -11
24 13 10 1 27 Loss -11
18 9 -4 3 26 Loss -11
34 17 4 1 2 Win 25
13 3 -6 3 16 Loss -11
25 -5 15 1 29 Loss -11
20 17 14 3 36 Loss -11
1 -1 4 1 24 Win 25
4 18 1 3 2 Win 25
28 -9 -8 1 4 Loss -11
0 5 -13 3 8 Loss -11
2 6 15 1 18 Loss -11
33 16 11 3 11 Loss -11
6 -12 -18 1 18 Loss -11
36 3 -13 3 22 Loss 25
Total 32
I have, as I said, tested this against the following:
5500 Numbers collected by myself from casino tables.
The reults for this test were +6952 Units
108000 Numbers from various RNG programs.
The results for this test were +65502 Units
I tested in batches of 664, the first 24 numbers weren't "played", but the rest were. The only real reason I tested in batches of this size was the spreadsheet was taking too long to update on larger amounts, and freezing me out enough to drive me crazy.
There is no progression here, all results are from flat betting.
I tested various variations to the rules I have explained, like a "rolling average", continuously calculating the average on the immediate previous ten, larger samples to average, smaller and larger playing time, playing more numbers, playing less numbers etc. I found that these variations didn't make a huge difference to the overall result. All provided a positive result and my utimate choice was a mix between what provided the best outcome and what I felt I could comfortably play.
I am looking for some help in testing this and any feedback would be appreciated.
I would like to verify one way or the other, whether I have made a massive error in my calculations, as I did not verify every single result from the 115000 that the spreadsheet gave me, although I have manually re-calculated around 10 000 results randomly to ensure I got the same result.
Secondly I would like to verify that I haven't got the weirdest set of 115000 numbers drawn from around 20 sources that all coincidentally fall into a specific pattern.
Thirdly, I would like (assuming the first two points don't shoot this down in flames) to know why it works.
I started out assuming what I was working on would only work on a real table. The only reason I ever even attempted this on random numbers was out of frustration at the lack of real spins I had available. I fully expected that when I plugged the first randomly generated numbers in I would an overall negative result. That the system continued to provide positive results has me completely perplexed.
Well, it's over to you guys now. Just one more point, please don't come back to me and say this has failed because you got a negative on 100 numbers.
Amongst my testing, the worst batch result I got was -930. The lowest an individual batch has come back from is -478. This system is, sometimes, over the very long haul
Cheers
Thanks for the explanation mate. Will do some testing but looks promising.
Neat system. I would love to write some formulas to help you test this system as it is interesting. I understand it all, but can not figure out 3 different numbers:
10 -13 -5
3 17 11
13 14 10
5 7 -10
34 -10 13
In this set of numbers above, when calculating the 3rd & 5th relative Count:
3rd, 14-(-13)=27 you have that it equals 10. How did you come up with this?
5th, -10-14=-24 you have that it equals 13. ??
another is toward the end, how did you get the relative count of 16 from (-7-14)=-21?
I am probably just missing something, but all the other numbers worked out. If you just made a few errors, let me know and I will attempt a test with some real played numbers. Also, if these were errors, how many more are there in your test you have accomplished? Hopefully none, but just a thought. Either way, I will test it without errors.
Cage
Sorry,
I get what you are doing now to get the 10, 13, and 16 relative counts (subtract from the 37 numbers). However, how do you determine all of these to be a positive value? Shouldn't 10 & 16 be a negative value representing the direction of the first jump count (similar to all the other positive and negative relative jump count values)? seems like it would have to be this way to get a true average jump value?? Please help clarify, thanks
Cage
Cage77
No, the numbers should be oposite.
From your start point, you can move forward 18 numbers (clockwise) or negative 18 numbers (anticlockwise). -18 and 18 are actually next to each other so when you physically count -21 numbers, you go past -18 and into positive 16.
Another question: My numbers are from double zero wheels. How do you treat a jump count of 19? positive or negative (on a double zero layout)? Or would that not matter?
And to make sure that I have this correct on your example:
1. the clockwise 27 is a -10 Negative Correct?
2. the counter clockwise -24 is a 13
3. the counter clockwise -21 as you stated is a 16
It makes sense to me know.
I have started some formulas, but may take me a week or so with the Thanksgiving holidays this week.
Cage77 I have never played a 00 wheel, but you can treat it as either +19 or -19, whichever you prefer.
You can, infact treat them all as positive if you want (0 to 36), it makes no difference. The only reason I have divided it up this way is when you actually have to count to the position to play, it is far easier to count -1 than +36, all the way around.
Hello Iwonder,
Do you always bet on 11 numbers? I don't understand it yet but trying.
Thanks
Bobco
Personally, I always bet on eleven numbers. The target number and five either side. This really is a personal choice. It doesn't matter whether you play 1, 3, 5, 7....17 etc numbers,so long as you play the same amount either side of the target number. I play eleven because I am comfortable with getting my chips down without having to rush.
Hey,
Sad face here. With my double zero wheel real personally played numbers, I tested your system exactly how you explained that you play it. I did it in an excel spreadsheet with a ton of formulas to make it easy to test and error free. My results were not good at all = -1471 units for 1786 total numbers tested in many different sessions. Not sure why the results are so much worse than yours. Curious on the accuracy of your test, maybe the small error factor is what you need to come out ahead, dunno.
All in all, I would like to see some more ideas on jump count type systems similar to this one. I will be glad to test systems when I have time for anyone. I have seen a several and have lots ideas that I might continue testing on this one after the Thanksgiving holidays.
Happy Thanksgiving to all,
Cage
Cage77
Can you post a session, so I can have a look? I'd be interested to see the results.
When you were handling counts over 18/under -18 did you subtract 37 or 38? For a 00 wheel you;d need to use 38, as the total number of pockets. You'd also need to handle +19 as a ligitmate result, so the limits would be less than -18/greater than +19.
As I said, I haven't ever tested a 00 wheel, so I would really be interested to see the results
Hey Iwonder,
did you track your results on a excel spreadsheet? If so, could a get a copy of it (the spreadsheet with formulas)? I could then also assist you in testing (both online RNG and live play).
Cheers,
Matt
Can't speak for him, but either of you are welcome to me numbers. Theyare all in Excel and i will be glad to send it to you. We just need to find a way to do it electronically or I will have to mail you a disk. I have a 3rd set of numbers on a diff file also that i did test yet due to the results. I have a total of around 5000 trusted numbers.
How big of a file can you recieve in an email? The two diff files that i have tested are 7kb and 11kb. I would prob have to talk to you via my email and explain the testing to you.
In exchange, do you have any real played and trusted double zero numbers? The ones I use are from my own playing in real land based casino's, so I trust them and know there aren't any errors from pickups and things like that from downloaded numbers. They are also broke up into each diff time period that i played. So you can test similar to if you actually went and played for a few hours at a time. I still have the original cards and notebook with the numbers.
Cage
I don't have any 00 numbers, as we only have single 0 tables where I am. If you could email me what you have already tested, that would be great. I'll see if I come up with the same outcome.
I can't paste it in here because it is too wide and overlaps....scrambled looking. If you post your email, I will be glad to email a text file of one example. Or if your email can handle the entire files, I can try...but they are 7kbs and 11kbs.
Yeah, I used 38 instead of 37. I used the real value for -19 or +19 in the relative number results.
Example:
23
6 -5
12 -5
29 -1 -4
21 7 12
32 16 17
15 5 -2
and so on. later for a -19 example:
30 -11
10 17
3 -7 4
1 4 -13
5 -8 -1
4 -15 -19
and so on. so, does this -19 get treated as a -19, 19, or deleted and divide by 9 instead of 10? I treated it just like it calculated above.
What is your email address? The files are large though due to the formulas.
I have to go out of town for the holidays. Send me that email address and If I can get you the test.
Take Care,
Cage
i decided to dig this post up it looks interesting
just remember the post is a few years old, these members might be long gone people
The title of this system looks similar to Mr. J's My Method. It might confuse people.
Quote from: Proofreaders2000 on March 04, 2010, 04:04:54 AM
The title of this system looks similar to Mr. J's My Method. It might confuse people.
It appears i cant change the title of the thread
Maybe iwonder dropped this in favor of Mr J's system, which he posted in RE:an old system?