Hi All,
I hope you all had a great Christmas!
Lately I have been thinking about what is "the game" of roulette. Many people use that term as if it is obvious. I think however that it isn't obvious at all. There are inherent losing games and winning games, so I believe it should be worthwhile to ponder for a moment on the "game" of roulette. What is a game? There a several definitions, but all agree that games have rules and goals.
Most common roulette game
After reading many posts, i conclude that the standard game definition is most common: goal is to "predict" the next number spun, or the next number characteristic spun. I think it to be common knowledge (by now) that this is a losing game proposition: if the wheel or the RNG are truly or pseudo random, than we can forget it, prediction is not possible (- true there might be wheel bias, or there might be a really simple and silly RNG algorithm; but I'm looking for the holy grail that is in dependant from other factors).
All statistical proofs, 99.99% of the systems deal with this roulette game interpretation, and indeed all are losers in the long run (for the truly stouthearted there are some systems with which you can reach a couple of kEuro in no time; but of course, in the end you will lose, it is not a hg method; for example a labouchere on the high and low at the same time - the starting row has 1 element and is 1 unit long)
Other games
Do not get me wrong, I do not know if other games will lead to winning systems/methods. This list are just examples, and might lead to nothing:
1. 3 dimensional tic-tac-toe (we have 4 3x3 planes)
2. 2 dimensional tic-tac-toe, we play on a 3x3 field per field we cover 4 numbers
3. ...
do you have any ideas?? (hmm maybe this is a subject for brainstorming, on the other hand I believe it is fundamental stuff for all students of roulette) ??
reddwarf
OK,
I give it a shot:
- higher or lower than the previous number (check out "optimal stopping" (link:://:.americanscientist.org/issues/pub/knowing-when-to-stop) you will be amazed)
- ehhh, (indeed this is a difficult exercise, it forces you to think outside the box)
reddwarf
roulette is addiction..i thought i would quit..but i cant...thats y foks we should never gamble with the money we cant afford to loose..even i am in loss in my 10 years play..still i am happy..that i have used the money which i can afford to loose
Hi Darrnyf,
Yep, you are right, roulette is enticing. But in those 10 years, what roulette game did you play? What goal did you persue? Was it the same game allover: "predicting" the next spin? Or was it a radical different game?
reddwarf
The reason why I ask is: it can be proven that (OK, to avoid discussion I agree that there might be something in VB), when the games goal is to predict the next spin, you will lose on the long run. However, it has never been shown that playing another roulette "game" is also a losing proposition.
Although I have to admit that I do not know if such another games exists...
reddwarf
Thanks for the article on "Knowing when to stop".
That's what I love about this forum, you never know when something might spark an idea.
Am testing the "N = 2 Surprise" to see if it can be applied it to roulette and preliminary results look promising. More testing in progress and if it continues to look good will post system.
Nick
Hi Nicksmi,
Great someone tries something new: good luck :thumbsup:
Quote from: Nickmsi on Dec 29, 08:44 AM 2011
Am testing the "N = 2 Surprise" to see if it can be applied it to roulette and preliminary results look promising.
There's nothing to test, because that "paradox" does what it claims. So, either it can, or can't, be logically applied to roulette.
However, something can't follow from nothing, so-called "paradox", or not. The problem with picking numbers at random from an infinity of integers is that those numbers can just keep getting larger. And by artifically imposing a "reasonable end-pt", we are left with the simple prospect that the larger apart the two numbers, the more chance that those two will straddle those (reasonable numbers') overall mean. The tip-off is that true randomness is randomness, whether we so pick just the first or second number (necessarily along with a guessed number in between), or only the number in between.
Hi Garnabby,
You might be right that there is nothing to test as it only applies for unboundend numbers (aleph-null numbers??), I could not make a winning system out of it, maybe Nicksmi can, or maybe it will lead to other brilliant insights no-one thought about, who knows?
The essence here is that, looking at number streams from a different angle might give new insights.
reddwarf
Quote from: reddwarf on Dec 29, 06:06 PM 2011
The essence here is that, looking at number streams from a different angle might give new insights.
And hi, reddwarf.
Yes, and try as might for a couple of hours before replying, i couldn't find a way to rule out this possible "other streams" approach for roulette. I've done some similar baccarat-research to try to "hedge" some of the lesser-known card-counting routines so that whichever way and degree the counts went for some given scenarios, mostly one outcome would be indicated.
Both of you are right.
I'd like to think a little out of the box sometimes because it's fun and can be rewarding.
Here's how how I am testing this N2 surprise. I've tweaked it a little bit so it's now exactly like it is in the article.
If the first number (Casino) number is <19 then get a random number.
If the random number is greater than the first number then what are the chances that the next Casino number will be greater than the first number.
So far in over 2500 spins (RNG), 75% of the time, the second number is higher than the first number. This appears to be mathematically expected as we limiting our trigger to LOW numbers.
So if we know that 75% of the time, the second number will be greater than the first number, can't we construct a bet to take advantage of this?
So I constructed two bets, one on HIGH and the other on remaining Streets.
So far the streets are hitting at a 10-1 rate, 429 wins, 41 losses. All flat betting.
There are many, many 5 â€" 10 units streaks in those 429 wins. Plenty of times to get on and off the win wagon with a profit.
Testing continues.
Hi Nicksmi,
Ha, that is an interesting approach! I'm really curious to the testresults, concerning the remaining streets, does one of the streets include the previous number?
And, what is the function of the random number, does this enhance the probability? Many questions, maybe I'll give it a shot today.
But, keep in the back of your mind that we still are looking for "other" games to be played...
reddwarf
Hi reddwarf . . .
The street system I am testing is just betting the 11 streets that do not include the previous number.
I am also testing to see if adding the random number as a trigger makes any difference.
Will keep you posted.
Nick
It would be easier to focus on the "good guessing" element of that "paradox" than to precisely find/recreate its numeric intervals here. What sorts of guesses can be of value; how to design your system around being right, if only by luck?
Playing for a solid shoe-mixture of baccarat- P's with B's is another simple example. By far, the most-likely shoe, hence the best way to guess the most of its outcomes. (BBBB... will very-likely net you only about half the outcomes.)
When you do just win, by other than progressive-betting forces, you want to make the most of it.
I did a 1500 spin test (3 sets of 500 spins) to see what the results would be if we eliminated the random number. This is all RNG.
Win% Without Random #: 68.5%
Win% With Random #: 73.04%
High Bet Win% Without Random #: 44.88%
High Bet Win% With Random #: 53.09%
Street Bet Win% Without Random #: 92.91%
Street Bet Win% With Random #: 93.63%
I used the identical spin data to test both ways.
While this is only a small sample, it shows in each case, we would do bettor by utilizing the random number trigger.
So I will continue to test with the Random Number trigger and see if this is a worthwhile method.
Happy New Year to All . . . Nick
so at the end of the day roulette remains game of chance.... >:D hope it turns good way of play nick...
Hi Nick!
How many units do you lose if you lose, and how many units you win?
I saw you play against the last street? You put high or low and 5 street?
Thanks.
Hi Nick
I see where you are coming from but hate to say, I have tested random against random, the win/loss ratio is just the same as flipping a coin, I have tested it against PHP random(), random.org C++ ranodm() and Playtech they all have the same ups and downs, these are of course just my findings and I know other programmers have done the same, even trying two open game windows and using one against the other. Your results may vary.
Streets: Theres a member on the old VLS forum called John Gold, he had/has a blog section where he told us about Streets of Gold, very very similar to what you are doing, I have coded his system into a bot and it fails, you will be suprised ho woften that street hits, when you don't want it to, it's a good read you should check it out, sorry I'm to lazy to link to it lol
Hi superman,
If i'm not mistaken nicks approach is a little different. It is based on a nice piece of maths, you can find it in this tread, i put the link on it.
It can be proven, for unbounded numbers, that you can increase the probability of guessing if the next number is smaller or greater than the previous number by using a gaussian random number generator. The rng's you mention Generate uniform distributed numbers, so this in itself is already different.
I think that nick understands that for bounded numbers this might not be true, but his experiment is at least an indication that he might be on to something.
Reddwarf
Nicksmi,
Maybe it is an idea to open a new thread where we can discuss and follow your experiment?
Grts reddwarf
Hi Donik . . .
In my testing, I checked only the RNG data spins to see if they would win for just two types of bets, ie. Betting HIGH and Betting 11 Streets. There may be other bets that might even be better.
I did not bet them together, just kept the results.
For example, in 2000 spins we would have had 537 bets on the Streets. We would have won 494 times and lost 43 times. Each loss would have been 11 units or 473. Our flat bet profit was 21 (494-473).
In the same 2000 spins we would have had 744 bets on HIGH. We would have won 386 and lost 358 for a flat bet profit of 28.
What intrigued me was how close to the expected win/loss values these bets were and how stable the ecart (variance) was, albeit only 2000 spins tested. Might changed drastically with more testing.
My thinking was that with this kind of stability perhaps we could get off the train at the profit stations more often. Whether you want to call it “hit and runâ€, timing, luck or “knowing when to stop†or whatever, I think riding the ecart is one of the best ways to make a few bucks.
If this does not pan out, perhaps John Gold (thanks Superman) might have a few nuggets to mine.
You are right reddwarf, that this testing is not the same as the experiment, but it was just a thought I had that might be applicable to roulette. Maybe it is just random vs random but I do enjoy the journey.
Spins | Wins | Losses | P/L | Profit % |
2750 | 386 | 358 | 28 | 3.76% |
3000 | 423 | 385 | 38 | 4.70% |
3250 | 459 | 412 | 47 | 5.40% |
3500 | 488 | 447 | 41 | 4.39% |
3750 | 522 | 486 | 36 | 3.57% |
4000 | 553 | 521 | 32 | 2.98% |
4250 | 587 | 551 | 36 | 3.16% |
4500 | 622 | 580 | 42 | 3.49% |
4750 | 652 | 615 | 37 | 2.92% |
5000 | 684 | 643 | 41 | 3.09% |
Still testing. I am now just concentrating on the HIGH bet and have discarded the street betting.
The above is the results from just the HIGH bet. It still shows remarkable stability. It almost passes the Van Keelen Test, ie. if in 1000 placed bets you generate 100 plus profit. We only generated 40 Plus profit in over 1300 bets.
Again, this is all flat betting with RNG spins.
With time permitting, will now look at some reasonable progression that might enhance the profits.
Nick
Hmmm,
no more games to be played within roulette? I think I will lock this topic.
grts reddwarf