You can now download the program I used to collect this data here if you'd like: nolinks://vlsroulette.com/downloads/?sa=view;id=401 (nolinks://vlsroulette.com/downloads/?sa=view;id=401)
Alright, so I asked in the Roulette Coding section if somebody could make me a program that would analyze spins in sets of 37 spins. Tangram took on the project (and finished it very quickly) and I figured it would be helpful to everybody here in the forum to see just what happens in any given set of 37 spins. That excel file I have attached is the results of analyzing roughly 66 thousand spins from an RNG. The spins were collected from a PlayTech Casino.
You can check the thread out here: nolinks://vlsroulette.com/roulette-coding-zone/need-a-program-that-analyzes-roulette-spins-in-sets-of-37-spin-sequences/ (nolinks://vlsroulette.com/roulette-coding-zone/need-a-program-that-analyzes-roulette-spins-in-sets-of-37-spin-sequences/) to see just what it is looking for and analyzing.
If you don't go to that thread, I'll tell you what the values at the top of the excel file I've attached mean, but you'll get a much better understanding if you read that thread.
So, across the top of the excel file you'll see the following (I've explained what each means here)
Set of 37 Spins Number - The program analyzes sets of 37 consecutive spins as if they were one block. So, if you see the number 2 in that column, it means the data to follow in that row is all of the data for what happened in the second set of 37 spins analyzed. Now, it analyzes data in consecutive sets of 37 spins. So that means that if you analyzed 38 spins, you would get TWO sets of 37 spins analyzed. This is because obviously you have 1 through 37 to analyze (thats the first set) but then you also have 2 through 38 (the second set of 37 consecutive spins)
Number of Different Numbers that Appeared - This is pretty simple. It just means how many numbers showed themselves in a given set of 37 spins. It does not show you WHAT numbers came out, just how many differnt numbers. A number is only counted once regardless of how many times it came out. This is a good column to look at to see the "Law of the Third" in action.
X2 - Any number beneath the heading "X2" is the number spin on which a number appeared two times. You'll see a lot of these as numbers appear twice quite often in a set of 37 spins. Keep in mind that again, the number underneath that heading IS NOT the number that came out out twice. It is on WHAT NUMBER SPIN a number appeared twice. So, again, if you see 23 underneath an X2 that means that on SPIN 23 a number doubled.
X3 - Same as X2, but for on what number spin a number appeared 3 times. Now, keep in mind that when a number appears for the third time on what number spin it appeared is recorded in the X3 column, but the spin that it doubled on (X2) is not removed. As for a number to appear 3 times, it obviously had to have appeared twice.
X4 - Sames as X2 and X3 but for on what number spin a number appeared 4 times.
This data has a lot of VERY valuable information inside it. Whether that information is of value because it will PREVENT you from trying to design a system, or give you new insight on how to design A NEW system there is a lot that goes both ways in this data.
For instance, you'll be surprised how often in a set of 37 spins a number DOES NOT ever appear 3 times.
Enjoy!
Oh, and I would love to have some collaboration on what (if anything) patterns/systems can be extracted from this data. It's pretty cool stuff. Since it is in excel you'll be able to sort the data any which way you can think of to see what you can come up with.
Alright, the link to download the file is: nolinks://nolinks.mediafire.com/?sharekey=b320ba1087b5a97400d27174b47c665736be3272e67c3ad85be6ba49b5870170 (nolinks://nolinks.mediafire.com/?sharekey=b320ba1087b5a97400d27174b47c665736be3272e67c3ad85be6ba49b5870170)
There are two excel files there, both contain the same data. I just uploaded two in case one format didn't work for some of you.
Sorry guys, the file will be posted momentarily. I tried attaching it to the post but it is too large. I'm uploading it to MediaFire and will post the download link in a second.
Alright, for those of you who read the first post in this thread and then were waiting for the download link (and didn't feel like re-reading the first post to find the download link) here it is: nolinks://nolinks.mediafire.com/?sharekey=b320ba1087b5a97400d27174b47c665736be3272e67c3ad85be6ba49b5870170
RJ
As usual your generosity and hard work are very much appreciated.
Thank you mate
TSK
Quote from: The Spiders Kiss on June 17, 2009, 05:28:42 PM
RJ
As usual your generosity and hard work are very much appreciated.
Thank you mate
The Spiders Kiss
Thanks Spider! And it's no problem, as we're all here working towards the same goal.
Just as one example of something that I extracted from that data in the form of a system is the following (I have included a BR balance trend graph)
Track 10 consecutive spins of the wheel. If a number appears 3 times in any given set of 10 spins then you bet on that number as well as any other number in the next 26 spins that appears 3 times. It isn't foolproof, but again it seems to be a solid system.
I had another very cool post on this thread...but apparently it has been lost....
I'll re-post when I have the energy to re-type it all, haha.
Quote from: rjeaton1 on June 19, 2009, 12:20:06 AM
I had another very cool post on this thread...but apparently it has been lost....
I'll re-post when I have the energy to re-type it all, haha.
Crappers, deep apologies.
This was a very hard merging, had to deal with thread versioning trying to get us the older ones which got corrupted back in place.
I'll try to see if there is a copy left around so you can post manually... I really don't want to add another database to the merge-up! :swoon:
@ Victor - If it is possible that would be great! Definitely don't worry about trying to do another merge up, you (I'm pretty sure) have done a TON of work already trying to get it back to where we are now. If you can't find a copy I'll re-type the post sometime in the next couple of days.
Working on something at the moment.... ;D
Thanks again Victor for all your efforts, and I'm sorry to hear you'll be leaving us as "administrator" but I'm glad you'll be staying around in another facet.
I'll stay as a member RJ, but -for instance- I won't be able to log in for days whlie I devote to hardcore programming, right now I need to have "tunnel vision" and only focus in what I need for my family's maintainance and support.
I'm very sure TwoCat or Lanky will do a great job as administrators.
I posted what I'm about to say once already, but it unfortunately was lost during the last fix. I'll do my best to reiterate the same info here.
First off, I'll say this is (in my opinion) pretty amazing stuff. I have posted a BR balance trend at the bottom of this post, but first I'm going to try and build some anticipation about what the results were and why they're so astonishing. I'll also try to offer some insight as to why I continued to try and come up with something better than what I already had.
Alright, so here we go. I got the idea that I put into play with this system from looking at the data extracted from the program Tangram made for me. You can download that data using the link in one of my previous posts on this thread.
I strongly encourage any input on what anybody else can get out of that data (I.E. patterns, trends, or any other "attackable" repeating scenario...etc.)
I felt as though I had a very successful system that I was ready to play with using real money and having the bot I recently had made for me play it for me so I didn't have to spend hours in front of my computer. The system (and this is a vague explanation) involved betting on 16 numbers with a basically infinite progression (using .01 as units) until a win occured. In my test the furthest it went was I think roughly 12 bets into the progression. That however, using .01 as units is roughly a $150 unit drawdown (which is actually 15,000 units). At the end of the 66k spins I was roughly $80 or so dollars in profit.
I got to thinking though..."have I really come up with successful system?" What I mean is, I was wondering if I was just "brute forcing" the table into producing a win. I mean, the system involves betting on 16 numbers, almost half of the numbers on the table. It would bet on those 16 numbers until a win occured. So, I thought it's possible that the system I came up with wasn't actually winning because it was a good system, but simply the amount of numbers I was betting on combined with the length of time I was betting on them. I mean, technically after a long enough period of time, one of those numbers should theoretically HAVE to hit but not because of the system....simply because it is a lot of numbers.
So, I decided to go forward with my search...and here is another example of a system using the information gathered from my countless hours in front of RXtreme, using the bot, and the data just recently extracted using the program Tangram made for me
Details about this system
First off, it too involves progression betting. What's the largest drawdown incurred using this progression you ask? Ready? 112 UNITS!! The first time I posted this it was actually 203 units. I changed just a couple of things and SO FAR the largest drawdown is 112 whopping units.
So what you might be saying, while asking these questions - "Over how many spins?" "How much was made?" "How many numbers are you betting on?"
I'll answer each of these.
How many spins? - 103,597
How much was made? - 1,090 units
Alright, now here is the REALLY awesome part.
The last system that I felt was successful enough to put into live play was the one that involved betting on 16 numbers. I talked about why I was wary about that actually being a successful system as compared to "a brute force" system.
This system involves betting on....ready? AN AVERAGE OF 5 NUMBERS!! The most amount of numbers I've bet on now is 12 (in the previous post it was 8, but I've gone to 12 now).
But like I said, an average of 5 numbers are bet on! That is amazing in my opinion.
So, to recap - The system involves progression betting, you bet on anywhere from 2 to 12 numbers (with the average being 5), the largest drawdown I've had in over 100,000 spins is 112 units and the profit I've made in those 100,000 or so spins is 1,090 units.
Overall, I'd have to say that is an incredible result and definitely not a "brute force" or "luck" system. I'm not saying this is the Holy Grail, nor am I even suggesting it. I'm simply saying that the more information I obtain about what the wheel does and the results "random" produces, the better the systems I'm able to create get.
The one thing you'll see in the BR balance trend below is that it does not bet very often. So, am I bound to get a loss here (and probably soon)? yes, I'm sure I am. So, I have built into the system (although it hasn't had to be used yet) a stop loss of 200 units. It doesn't actually stop playing when that loss occurs, it just starts the progression over again without trying to recover those 200 lost units.
Also, one other "safety measure" I've built into this system is this:
If it goes past 7 bets into the progression without a win, it stops betting and starts tracking again. The next time it goes to bet, if it hasn't hit the stop loss of 200 units, it will continue to try to recover any previous losses. If when it went beyond the 7 bets into the progression it DID lose 200 or more units then the next time it goes to bet it DOES NOT try to recover previous losses.
So far, neither of these safety measures has been used.
I have attached the BR balance trend, as well as the full spin/betting history.
MODIFICATION: I tried to upload the spin/betting history and it was too large for this post. Since there is no downloads section anymore, I've uploaded the .txt file with the betting/spin history to MediaFire. You download that file here: nolinks://nolinks.mediafire.com/?sharekey=b320ba1087b5a97400d27174b47c66579ff70cf123e90c69c95965eaa7bc68bc
I would share the details of this system here, but it is awfully complicated.
It works based on a number of different things, which is why I'm using a .DGT file in RXtreme to test it. I have basically re-written the file I uploaded in the "ultimate RXtreme .DGT file" thread to test this system.
I will be uploading the new one soon, I just have to make it so that it will be understandable by everybody (as of now, I have it sloppily coded and most of the inputs I used/changed aren't able to be changed on the input screen. I've been changing them all manually in the coding.) As soon as I have it all nice and neat, I'll upload the file for everybody to use.
Also, as I type Tiago is adding the new additions to my bot so I will be able to test this against an online casinos RNG not just RXtremes.
The reason for my playing with a bot (as it doesn't do anything I couldn't do myself) is because of the long amount of time I would have to sit in front of my computer. At playtech, in my last test using the bot, it takes about 12 hours for 50 to 60 thousand spins. I'd much rather have the bot do it for me than sit in front of my computer for 12 hours.
Again, if anybody has actually downloaded the data file from the beginning of this thread and found anything useful in it, please share it.
Incredible stuff RJ. Too bad it's not playable manually due to the large number of spins before another bet occurs, but I guess that's what bots are for.
Quote from: celiza427 on June 20, 2009, 11:17:00 AM
Incredible stuff RJ. Too bad it's not playable manually due to the large number of spins before another bet occurs, but I guess that's what bots are for.
Thanks Celiza! Yes, that certainly is a drawback. I'm trying to see if there is a way to do it manually... You'll be sure I'll let everybody know.
Almost done with the .DGT file everybody! (For anybody interested anyway, haha)
Quote from: rjeaton1 on June 20, 2009, 09:22:13 PM
Almost done with the .DGT file everybody! (For anybody interested anyway, haha)
Ooh ooh, me!! *waves hand frantically* Just letting you know your hard work is appreciated ;)
Quote from: celiza427 on June 20, 2009, 09:50:39 PM
Ooh ooh, me!! *waves hand frantically* Just letting you know your hard work is appreciated ;)
As always Celiza, you're a huge source of encouragement! Thanks! ;D
Glad to see at least some of the stuff I put up here is read, haha.
I've decided to post the prgram Tangram made for me. I'm sure there are others that might find it useful. You can download the file here: nolinks://vlsroulette.com/downloads/?sa=view;id=401 (nolinks://vlsroulette.com/downloads/?sa=view;id=401)
Im still eagerly waiting on the RXtreme file aswel! Fantastic stuff again
Hi
I checked your "66K analysed spin"-file
There is something wrong
pls check:
Set #1
21 #s appeared
11 #s 2x
4 #s 3x
1 # 4x
11*2 = 22 spins
4*3 = 12 spins
1*4= 4 spins
That are 38 spins needed in 37spins?
pls check
br
winkel
Quote11*2 = 22 spins
4*3 = 12 spins
1*4= 4 spins
These are 16#s that appeared. That means no number appeared just once.
I checked more lines. No Line gives exact 37 spins if you calculate appearance and spins.
The whole file is wrong.
sorry for that
br
winkel
I appreciate that you took the time to check this thread out Winkel, also that you took the time to see if the file is correct.
It is correct however. I would type out everything again showing you that it is correct, but I'm just a bit too lazy at the moment to do that (I'm sorry).
If you'd like to understand exactly what you're looking at in that file, I suggest you not only read what I posted on this thread, but also the thread linked to in the beginning of the thread.
Also, just to make it a lot easier to see that the file is correct, I've attached a set of 37 spins in the form of a .txt file.
For you to see that the data produced by this program is correct you'll need to do the following.
Import the spins I've uploaded into RXtreme. Open up the wheel frequency tracker of RXtreme (unders - Statistics > Wheel Frequency)
Then, before you run those spins, analyze the same file with the program I've uploaded here. Open up the results in microsoft excel and make sure you are viewing all of the data (a lot of times, the data of X3 and X4 are WAY to the right and you have to scroll to find them)
Alright, now keep the excel file open and use the STEP button of RXtreme to spin one spin at a time (this is VERY important). You'll see everytime a number doubles, triples, or quadruples it will match exactly to the spin number in the analyzed data. Also, the number of different numbers will be exactly correct as well.
The way you've done your math in the examples you've posted is flawed. (I'm not picking on you, I just don't want people here in the forum thinking I've been giving out inaccurate information).
I also encourage everybody who thought this file was wrong to go through the steps I've just explained. You'll see that it is in fact correct.
look at this. It can´t be right.
Set of 37 Spins Number 0 spins needed 2 3 4
18791 21 54 11 8 2 11*2 8*3 2*4 22+24+8 = 54 spins needed
18792 21 54 11 8 2
52207 21 32 8 4 1
2581 20 50 12 6 2
6530 20 40 9 6 1
6531 20 37 9 5 1
6532 20 36 10 4 1
6533 20 37 9 5 1
13259 20 40 10 4 2
14134 20 50 12 6 2
17126 20 45 15 5 0
17127 20 47 14 5 1
17128 20 48 13 6 1
17129 20 49 13 5 2
18786 20 52 11 6 3
18787 20 50 12 6 2
18788 20 50 12 6 2
18789 20 50 12 6 2
18790 20 51 11 7 2
18793 20 51 11 7 2
18799 20 52 11 6 3
23271 20 47 14 5 1
23272 20 49 13 5 2
23273 20 49 13 5 2
23274 20 37 12 3 1
23275 20 26 10 2 0
23276 20 27 9 3 0
23277 20 27 9 3 0
23278 20 27 9 3 0
23279 20 27 9 3 0
23280 20 28 8 4 0
23281 20 41 9 5 2
23282 20 37 9 5 1
23283 20 37 9 5 1
23284 20 37 9 5 1
first row your index
second row = unhit numbers
third row = needed spins for the result (should be constantly 37)
Here is a quick example:
Lets say the wheel only has 10 numbers and the file analyzes sets of 10 spins.
The next ten numbers are the wheels spin results:
s. #1 - 8
s. #2 - 8
s. #3 - 1
s. #4 - 8
s. #5 - 2
s. #6 - 8
s. #7 - 3
s. #8 - 4
s. #9 - 5
s. #10 - 6
The results would read as follows:
# of Different #'s that appeared: 7
X2: 2 (a number doubled on spin 2)
X3: 4 (a number tripled on spin 4)
X4: 6 (a number appeared 4 times on spin 6)
So, your example would look like this:
1x2= 2
1x3= 3
1x4= 4 +
-----------
9 spins. See, it doesn't add up the way you've done the math because of the way the program keeps track of the data and what data it is actually keeping track of.
If you do the steps I mentioned above exactly as I described them you'll find that there are no errors.
Here is another example building of off the first:
Lets say the wheel only has 15 numbers and the file analyzes sets of 15 spins.
The next fifteen numbers are the wheels spin results:
s. #1 - 8
s. #2 - 8
s. #3 - 1
s. #4 - 8
s. #5 - 2
s. #6 - 8
s. #7 - 3
s. #8 - 4
s. #9 - 5
s. #10 - 6
s. #11 - 6
s. #12 - 6
s. #13 - 2
s. #14 - 11
s. #15 - 6
The results would read as follows:
# of Different #'s that appeared: 8
X2: 2, 11 (a number doubled on spin 2 and spin 11)
X3: 4, 12 (a number tripled on spin 4 and spin 12)
X4: 6, 15 (a number appeared 4 times on spin 6 and spin 15)
So, your example would look like this:
2x2= 4
2x3= 6
2x4= 8 +
-----------
18 spins. See, it doesn't add up the way you've done the math because of the way the program keeps track of the data and what data it is actually keeping track of.
Have you managed to complete the .dgt file? Thanks
Sorry rejaton
I obviously read the statistics wrong way.
br
winkel
@ Sean>> Still working on it. I didn't end up paying to have it completed, but I still may. Sorry about the wait...just messing around with about 40 different things all at the same time. That and I've been working a lot more lately...
@ Winkel>> No problem Winkel, like I said originally, I'm glad you came to check this out and make sure it was correct. I've read a lot of your threads and have the utmost respect for you and your ideas. Hopefully I didn't come off sounding like a jerk, as I certainly didn't mean too. So, thank you again for reading this read and taking the time to comment.