Guilty Gear EC...

[birdhands]

The more I think about this the more confused I get.  Up until now I thought I was the voice of reason, but now I think it's all a little more complex than I originally thought. 


So we know that in 12 million spins the longest run for an EC was 23.
We all agree that an EC is equivalent to any 18 numbers.
That run of 23 reds in the data could have only had 14 unique numbers and 9 repeats (for example).
Oh my, I get it now: so if there are 18 unique numbers out of 30 then the other 12 are repeats of the 18, so it's equivalent to an EC running 30 in a row.


Well now, I feel like an a**.  You were right all along.  My humblest apologies.  How embarrassing. 

[iancloud2001]

birdhands,

Precisely! You are correct. Don't apologize! Your own curiosity allowed you to clear things up for anyone else that might not have understood this discussion. You explained some of my theory better than I could.

Now that we have things cleared up, I'd like to know just how long a grouping of 18 numbers can actually sleep. I know that a true EC can sleep for up to 29 spins in 12 million spins... how about a virtual EC? Honestly, I'm starting to think that we'll need 50 million spins to get a better picture. I should get started on that with RX soon... It could give us insight on a lot of things...

amk,

The problem with waiting for only 13 virtual ECs is that you can't know what to bet against. You need 19 unique numbers (on a single zero wheel) to get the opposing 18 numbers you're supposed to bet on.

[birdhands]

This is a very exciting idea.  I'll be testing it whenever I can.

[birdhands]

Yes, we may need 50 million.  So far this is easy to test.  In 50 sessions I've already had 34 numbers with 19 uniques.  Incredible.  I had an idea like this last year involving dozens.  I calculated how many groups of 12 it was possible to split 36 numbers into- the answer blew my mind- it was 500 million+ groups!

[iancloud2001]

Yeah I know... There are so many combinations. As of now, I have the RX program running trying to get 50 million spins. It should be finished by tomorrow. I'll post the .rmd file for anyone interested. Have a good night everyone, and please do try this, or at least something similar to this. I feel like the only real way to beat roulette is to truly know what random is capable of...

[warrior]

Martigale ha ,18 unique numbers, inside numbers at bm casinos where i play max bet on1# 25$ you will get 5 bets in then you have reached the max ,sure you coul use the split bets but whats the max on that.

[Optimist]

Hi,

I have coded and tested it on my RSS bot. With marty progression u will be killed. I have attached one of the spins files I have tested (this is playtech RNG file from real money session). I have seen 19 unique numbers in last 43 spins  (its around spin 9190).
maybe another MM could help? Guetting?
I have tried short marty like 1,2,4 have tried various options. Sometimes I was in + after 10K spins - sometimes in -
Anyway - it is not playable manually (I think). The best results I had when marty is short and u bet trigger is 19 unique numbers in last 32-34 spins.

Regards
DL

[catalyst]

dear iancloud
martingale with 18 unique numbers!   your days are numbered!
thanks
catalyst

N.B. your system is as old as our civilazation itself. ask forum members for a playable progression.

[birdhands]

Hi,

I have coded and tested it on my RSS bot. With marty progression u will be killed. I have attached one of the spins files I have tested (this is playtech RNG file from real money session). I have seen 19 unique numbers in last 43 spins  (its around spin 9190).
maybe another MM could help? Guetting?
I have tried short marty like 1,2,4 have tried various options. Sometimes I was in + after 10K spins - sometimes in -
Anyway - it is not playable manually (I think). The best results I had when marty is short and u bet trigger is 19 unique numbers in last 32-34 spins.

Regards
DL

What about a fibo starting at 19 unique numbers in 36 spins?  Or Hermes' leveller prog.: 1,2,4,4,4,4,4...until profit.  In either case, I think the key is to set the trigger late enough.

[keel44]

Each spin of the wheel, when covering 18 numbers, has a 48.64% of hitting. 

I don't care how many times in a row.  Yes, you would think it is extremely rare for an even chance to hit 30 times in a row, but by your own discovery you see it happens all the time.  Any string of 19 unique numbers just proved to you that 30 times in a row happens quite frequently.  This type of thinking is based on gambler's fallacy that something is "due". 

The previous 19 numbers have a better shot to show than the "sleeping" 18 numbers.  You need to think the other way.  Take advantage of streaks and repeats.  You know there will be some of those.

[keel44]

I guess what I am trying to say is that on every spin there is an even chance bet that is hitting anywhere from 1 time in a row all the way up to 60 or more.  (I am just guessing)

Why choose against an even chance selection that appeared 34 times in a row?  It seems to me that it should be taken advantage of, not go against.

[birdhands]

Ian, what were the results of your 50 million test? 

[iancloud2001]

birdhands,

Hi, I still have Roulette Xtreme running. It's at 30 Million spins at the moment, and still rising. Should be finished by the end of the day. And guess what?? I'm getting the EXACT same results as the 12 Million spin sample.

No ECs repeating for more than 23 spins.
No Dozens repeating for more than 15 spins.
No Single Numbers repeating for more than 5 spins.

Plus, I know for a fact that I have continuous spin data, unlike the 12 Million sample where I'm not really sure...

Honestly guys, I'm going to save the session at 50 Million spins, then I will continue letting it run for 500 MILLION spins! I need to see what roulette RNG is truly capable of...

[birdhands]

Yes!


I've been testing a lot and I've seen it go to 43 spins myself. 

[iancloud2001]

What exactly have you seen go to 43 spins?