This thread will explain one way to beat a math game using math :)
Stay tuned - the only delay is running as many examples as possible before posting.
Yep...This is gonna be good, I know it. I can feel it in my bones.
Yo TG...How ya been ?
Still keeping busy...I see !
Keep us posted.
Buffster
Well, let's try anyway!!
Quote from: Bago on August 04, 2008, 08:49:50 AM
The math of this game is you hit a number 1/37 and you are paid 36 (35+1), in the long term (what counts). Therefore, the player has a negative expectancy of 2,70%.
You cannot beat the math of this game with your gambler's fallacies on past numbers.
Then why are you here anyway?
I agree...let's try anyway. I think TG has much more experience than most...let's take a swipe at it!
All ears TG! Or eyes I reckon!
Hi Turbo
Thanks a lot for what you are about to share with us Mate.
I remember Victor telling me this once.
""when the student is ready the teaching will come ""
It would appear that there are some of us that want to learn & Improve.
And there are some that would Boo Santa Clause.
Good On Ya Mate
Lanky
Bago is correct - the house has the math advantage over the player.
However, there are ways to play so that the player's balance stays right
around that "edge" - That is what I'll show with this method - and it will
make more sense perhaps :) The player's balance will wander in both directions
and stopping when a profit is reached is possible.
The house "edge" is small - only 1 unit per cycle of 37 spins (or 38).
so 1/37th of 1 unit per spin (very small)
This is simple to overcome when done properly.
TG
Would that include taking advantage of wins after a loss to recover that 1 unit every cycle?
Keith
Hi Turbo,
Like everyone else, I'm looking forward to hearing this method.
Bago is obviously right about the math, but as nothing in the universe is truely random - as the great philosophers have claimed - it seems plausible that the math can be manipulated to create a system where you only lose if you are very unlucky (a method that is about as close to a HG as mathmatically possible).
Bring it on!
CD
;D Hi T-Nice 2 see ya back around-On average how many unit's per cycle does your new method average?? P.Mail await's u
It's coming Bago, roulette is on the major back burner at the moment :)
Glad to see you here though, you always bring the "anti-turbo" attitude which
does little but is indeed entertaining to read.
He does help keep the ideas in the real world.
Any news about system?
Yes, Bago is right and a little childish ;).
I'm very curious TG.
If it's true what you say. If every roulette player would play this way the casino's edge would be gone ;D.
Waiting patiently
I did it as a stunt. I wanted to prove that you could avoid hitting the zeros on a 0/00 wheel if you used a fixed rule to avoid them. I ran billions of spins in order to prove that the simulation was not in doubt.
These were the fixed rules of the Sim:
Whenever the zeros had hit in the last thirty spins wait until there are no zeros present in the last 30 spins and then begin betting again.
Bet on the low 18 only, EC bet to win.
Play 300 spin sessions that have stops at 10+ and 10- points that end a session.
The simulation lowered the house's edge from .0526 to .0500, on average, a little more than one quarter of one percent in the very long run.
In the short run, you can observe the zeros hitting very active, average, or below average.
I was always told that there was no way to change the house's advantage.
Whenever you play roulette on a 0/00 wheel you have two slots on the wheel against you. There are almost always two numbers in Roulette that are always in a state of resting or sleeping. It just moves around and changes numbers about every 40 to 80 spins.
IMO after observing changes.
Gizmo,
You're not changing the house edge
Here's a thread I think you should studynolinks://vlsroulette.com/reference-area/roulette-probability-made-easier-t2193/ (nolinks://vlsroulette.com/reference-area/roulette-probability-made-easier-t2193/)
Herb, this is the nice forum. I proved that I lowered the house's advantage on an other mean forum.
I don't want to be mean here, just ask Victor. I wrote this simulation to win the argument that there is a way to lower the house's advantage and to attempt to open the minds of those that say it can't be done.
Do you need the proof too? I have all day, week, month, year.
Here, the proof:
Sim includes the 10 plus, 10 minus stop session points, After more than 281 million actual spins, the Sim bet more than 48.7 million times, lost more than 2.4 million times for a new House Advantage of 0.049767. That's just a little lower yet for a test.
So this shows that just the zeros being avoided and the stop loss, stop win can lower that house advantage just a little.
total lost = -2423989
Total number of bets placed = 48,706,527
New lower house advantage after real long run test = 0.049767 the number to beat "0.0526"
Actual number of spins in test = 281,782,130
on mouseUp
__put "" into field "f1"
__put 0 into totL
__put 0 into runtot
__put 0 into runtotTot
__repeat with i = 1 to 1000000
____put "" into zamOO
____put "yes" into holdZero
____put 0 into totLhold
____repeat with y = 1 to 331
______add 1 to runtotTot
______put random(38) into zipX
______if zipX < 19 then
________put 1 into zip
______end if
______if zipX < 37 then
________if zipX > 18 then
__________put 2 into zip
________end if
______end if
______if zipX > 36 then
________put 3 into zip
________if holdZero = "no" then
__________subtract 1 from totLhold
__________add 1 to runtot
________end if
________put "yes" into holdZero
______end if
______put zip after zamOO
______if y > 30 then
________delete char 1 of zamOO
________if holdZero = "yes" then
__________if zamOO contains "3" then
__________else
____________put "no" into holdZero
__________end if
________end if
________if holdZero = "yes" then
________else
__________if totLhold = 10 then
____________exit repeat
__________end if
__________if totLhold = -10 then
____________exit repeat
__________end if
__________add 1 to runtot
__________if zip = 1 then
____________subtract 1 from totLhold
__________else
____________add 1 to totLhold
__________end if
________end if
______end if
____end repeat
____add totLhold to totL
__end repeat
__put (totL / runtot) into cck
__put totL & " -- " & runtot & " -- " & cck & " -- " & runtotTot into field "f1"
end mouseUp
after a billion spins:
-2,423,989 -- 48,706,527 -- 0.049767 -- 281,782,130
-2,431,805 -- 48,634,899 -- 0.050001 -- 281,790,284
-2,424,882 -- 48,681,252 -- 0.049811 -- 281,908,415
-2,424,818 -- 48,639,756 -- 0.049853 -- 281,793,286
The one quarter of one percent improvement holds up.
Read it and weep.
*sigh*
Anyway, read the article.
Herb, *sigh*
What am I supposed to learn, I read it, it's basic probability.
You have no clue, I showed you the proof. Others are proving it too. They are writing their own computer Sims that show that the probability math can be broken in billions of trials.
Do you need an aspirin? I'll use the same math as your primer advises. These are the results of someone else trying my solutions to beat it. The real number to beat is 0.0526 in probability. Look at these.
Total bets placed 407073
Total number of bets won on low numbers 194768
Total number of bets lost 212305
212305 - 194768 = 17537 -- (17537 / 407073) = 0.043080
----
Total bets placed 1026608
Total number of bets won on low numbers 490929
Total number of bets lost 535679
535679 - 490929 = 44750 (44750 / 1026608) = 0.043590
----
Total bets placed 1020335
Total number of bets won on low numbers 488119
Total number of bets lost 532216
532216 - 488119 = 44097 ( 44097 / 1020335) = 0.043218
----
Total bets placed 4094557
Total number of bets won on low numbers 1956288
Total number of bets lost 2138269
2138269 - 1956288 = 181981 (181981 / 4094557) = 0.044444
You can have the last word.
"Don't trash this forum the way you and Spike trashed GG." --- It will SLOWLY happen. Ken
Hello Gizmotron,
You wrote:
"After more than 281 million actual spins..."
May I ask: Where did you find that 00-spins-collection?
A lot of people ask for 00-spins, you know. Please make a post in the "Actuals / Permanences" area.
Please...
TIA
KFS
I don't know about others but I thought Turbo was going to show us something. I've done all the work and you provided all the improper remarks, and started it here.
That's OK. Almost all roulette players are going to believe the basic premise of this thread, that it's impossible. I'm not going to apologize for you attacking me and calling it me trashing this forum. I know how to act here and have only tried to discuss something.
On this topic was a wonderful discussion by several others that wanted to get to the bottom of it and never attacked each other. Yes there were people that came into that thread to disrupted it, to take it off topic. I resent those that can't handle an objection on the basis of really learning something. I presented you with the proof and you have produced a belief topic to refute it. That's not enough proof for me. The RNG is not suspect even if you think that others will stand with you that it must be.
I'm not about trying to find people that want to discuss Roulette and have it turn into an argument. I had no idea that this petty arguing would pop up here. It should not have. Just look at the thread title.
It demands proof. Is that too difficult to consider?
Hello dears.
We all want our roulette sharing & exchange place to remain.
Remember there is a section for "that other type" of discussion right at this very page to use:
nolinks://vlsroulette.com/uncensored-pit/ (nolinks://vlsroulette.com/uncensored-pit/)
Shall you feel you need it ;)
Regards.
Victor
They are RNG spins that I wrote the code on so there is no tampering. If you can't use a computer to confirm long term results then I guess you will have to find your proof in a step by step walk through each objection. I'm using one of the most modern RNGs in the industry. It that uses microseconds since 1970 to establish the random seed. My Sim asks for a new session to start after each 331 spins. Each spin is independently asked for. The process for running this many tests takes more than 15 minutes to process over 300 million spins. In that large of a sample I reverse engineered my Sim with the same RNG and ran it for 300 million spins and got the correct 0.0526 house's advantage. The RNG produces accurate enough results to get a correct reading when no conditional selection is offered.
I did what all good scientists are required to do to prove the RNG as good enough.
Please tell me about any objections with RNG that are suspect. I'm here to listen and learn about what you believe.
The results just aren't relevant.
Victor, I feel like a heretic. I've had nothing but trouble with this topic. It's like a sacred cow or something. I thought this would excite interest, but not anger. I don't want to go to the pit because it produces and helps nothing.
go to the chat. We can talk there.
This way we won't create a thread 100 feet long. lol.
Is that your function Herb? I'm going to wait for Turbo to see what he has. I know the significance of what I have done. Every book on gambling says you can't change the odds. I think I will let someone you like let you in on it. Stuff like this doesn't stay under a rock no matter how many try to stop it.
You know something? Turbo just might have more fun and games for you to consider. I really hope so.
Here's the problem.
1. Roulette the game is a game of independent trials. Past results simply don't affect future results. Simple logic tells us this is true.
2. Since roulette is a game of independent trials, then using past numbers that have or have not hit in our bet selection is of no benefit.
Now here's a big exception,
Roulette wheels that are not random will tend to have certain numbers that clump, or come in when conditions are just right due to physical defects in the gaming device.
Regarding your testing: You may have simply slightly exploited a small flaw within the RNG. Try retesting. I'm impressed that you tested so many trials. It's really something to see.
-Herb
Quote from: Herb on August 25, 2008, 02:59:07 AM
Regarding your testing: You may have simply slightly exploited a small flaw within the RNG. Try retesting.
-Herb
I created a version that only bets without the conditional selection and showed that code as proof also. It returned a confirmed 0.0526 number that could not be more than a thousandth level off in the extremely large sample. This was confirmed by other computer programmers. The RNG works perfectly as expected for proving the probability Why would it stop doing that once the selection process is added? I was forced by others to use such a large enough sample at to prove it was beyond the risk of RNG error.
Have you tried running the test again on a different RNG?
This may be chat to you but nobody has come along before and changed the odds, albeit by selection and waiting to avoid the zeros. It's just an argument for the fact that Math is supposed to be the absolute final answer here. I have found that that is just not true. Now if that is the case then what other conditional selections can be employed to lower the odd even more. Is it possible to find a strategy that uses conditional selection that works on really abstract things like patterns within patterns and patterns within trends?
I apologize if this is a can of worms and that there are others that must slam me down for saying what I just suggested. Never the less I want to discuss it and think that on a forum dedicated to discuss ideas is much better than a chat room. Also, let it be known in all this. I don't care if I'm wrong and have made an error. That would be informative too.
Please suggest a place, Victor?
The math is absolute, however, you may have:
1. Found a statistical anomily in a random trial
2. Found a small flaw in a particular RNG.
I would be interested in discussing it further.
Herb - "Have you tried running the test again on a different RNG?"
Jugs Monthly created his own sim in another language and got lower results than me. I still don't know why. Bliss created a version of his own but it has problems with optimization and execution order with regards to use of operators and the find function. (sorry for the tech speak) Even Bliss's version produced almost perfect results for 0.0526. We all used massively large numbers in order to isolate natural drift from test to test.
I would love to see some others try it here. I never ever believed that anyone would not want it to be researched here.
I agree, long threads are a pain.
Gizmo, thanks for stopping by. Please feel free to open your own topic or shall you wish, I can split (separate) your posts at this thread and relocate them somewhere else you specify.
A good place to start posting this roulette-related topics is here:
nolinks://vlsroulette.com/gambling-and-roulette-related/ (nolinks://vlsroulette.com/gambling-and-roulette-related/)
Also, remember there's the testing zone, where you are totally welcome to post your charts/results:
nolinks://vlsroulette.com/testing-zone/ (nolinks://vlsroulette.com/testing-zone/)
Be certain there are many (including myself) who are happy to know you are around with a productive mind and for sure loving the fact you are willing to contribute your personal tests and take feedback.
Just tell me if I shall split & move your contribution to our "gambling and roulette-related" section or you'd want to open a new thread. Perhaphs one named "Gizmotron's beating a math game with math thread" to separate them from current Turbo's "beating a math game with math" one.
As you wish it is fine. Just let's work this out the best way we can, and thanks again for being around at the productivity side.
Your friend.
Victor
Quote from: TurboGenius on August 05, 2008, 09:46:41 PM
It's coming Bago, roulette is on the major back burner at the moment :)
Glad to see you here though, you always bring the "anti-turbo" attitude which
does little but is indeed entertaining to read.
When is the release turbo ;)
OK, while we wait for Turbo to begin again, thanks Victor. Just a short update. Three of us have written three different Sims in three different languages and they all worked correctly. We set out to attempt to avoid the zeros by waiting 30 spins for them to not appear. However we all had this extra 10+ & 10- stop point and we did that so different that we got three different results.
If this goes no further then there is no need to start another thread. It just ends here unless the reason for the test in the first place were to continue. We needed to prove you could in fact lower the odds. It turns out that when you attempt to miss the swarms of zeros you still lose many of them anyway. When you add those fewer losses to zeros to the normal losses on the 50/50 lower and upper 18 bets you get the 10- stop point kicking in more often than the 10+ does. That appears to have a small effect on the odds. What it means has not been decided.
If anyone wants to talk about this then tell Victor to separate out this Sim stuff that I brought up and we will delve into why it works and discover it together. Somebody wondered why we were bothering with all this. I answered him that with two little insignificant rules we could get a small percentage loss what would happen if we used more intelligent conditional selection rules. I told him that the door swings wide open if the Sim is operating correctly. We have already proven that the RNG are not suspect.
still waiting for Turbo.
Gizmotron,
I think Turbo has done quite a bit of testing in this area as well. You may find that he has some additional information to help you with your research project.
TG, when are you going to follow-up on this topic? It's been about 30 days, we are waiting. What is the appointed date to submit the remaining input?
Bloom
:) Funny how John has been waiting years for Turbo to deliver....yet it's his first post here. New guy huh? Or old guy new name?
aaaahhhh man you guys just do not let up. Get a F'n life. If you do not like Turbo's posts, do not read them. Simple.
I am on this forum because I am a fan of Turbo's work over a number of years.
In my opinion he has nothing to prove but always has something interesting to contribute and his posts tend to make me re-think some of the strategies I have been playing profitably in live Casino's all around the world.
Macca37
Quotehis posts tend to make me re-think some of the strategies I have been playing profitably in live Casino's all around the world.
That is nice to know. That you actually take some time to rethink. To me it is always an evolution, there is always room for improvement.
At this game we never get to the "I know it all" point. There will always be left another system, another tweak, a suggestion, etc. and that is what makes the game so appealing.
Regards.
Dumdeedum....dumdeedum....
It's a beautiful day today...
Maybe something new will be waiting for us today.
Maybe wishful thinking.
TTT
Tickety tackity toe,
your comments around here blow.... ;D
come on Turbo
Waiting!!!
ginger
This phenomenon seems very common in these forums...
Someone announces grand news and plans... Makes a few posts and then nothing happens
It's been half a year now and this "Turbo's Place" has gone nowhere
He must have 3. I have 2 and I'm still here. ;D