Ok, hear me out. This might sound strange, but it's an idea I've had for a long long time. It wasn't feasible before, but I think bots and roulette programming are progressing to a stage where it might be.
First look at this article.
link:://:.overcomingbias.com/2008/06/average-your-gu.html (link:://:.overcomingbias.com/2008/06/average-your-gu.html)
To cut a long story short, it basically says that if you take multiple attempts to guess something (e. g. the height of a tree) and then average the guesses, you will get a more accurate answer. Makes sense right?
Now, we have many many roulette systems. Some are very good and work very well most of the time, but always a situation will arise that will defeat a given system. What if we created (using a bot or a program of some sort) a master system? The bot would incorporate many many systems - so we are talking hot, cold sectors, dealer signature, EC, dozens, law of the third, sleepers, any system on here. . . whatever. Just any system with any merit. Based on all the systems, the bot would identify which numbers have the most likelihood to come and only bet those numbers. So if we have 10 systems, and 7 of them predict 18 will hit, then we bet this along with any other number which many systems predict will come. If only one system predicts number 4, then we don't play this.
What this means is we are not betting on many numbers - only the numbers that are most likely to come from the joint power of all the systems included.
Possible? Of course a bot, or some program would be needed for this to work as manually it would be a nightmare (and you definitely wouldn't have enough time to figure it all out before needing to place a bet). Is there anyone on here with the programming knowledge to accomplish this?
I think this is what Victor service is all about :)
Quote from: The_Force on Aug 16, 08:11 AM 2010
Possible?
Absolutely.
I'm applying this principle myself in the Tipping Service. With the sole difference that instead of full-blown systems I'm applying the most likely numbers from a collection of triggers.
It would be really good to see this implement to compare and contrast with what I'm doing.
I know of a group from Germany that is only waiting for me to complete multi-casino OCR so they can use their database of methods (mostly using statistical extremes) with an array of casinos at the same time... but I got them in stand-by, since my programming of the tipping tool for the fellows here is #1 for me right now.
Also, the multi-casino OCR is definitely a must for me to code anyway, to have both optical recognition (for reading numbers from any casino) and configurable clicking for using the tipping tool automatically with any casino the users take their time to configure.
Do know how to program yourself? (*ANY* language, preferably .net family)
Quote from: Jordan on Aug 16, 08:41 AM 2010
I think this is what Victor service is all about :)
The difference being that I don't use full-blown methods, but straight-up number triggers according to the current scenario and filter the eligible numbers.
Similar principles, but not exactly the same to what TF is suggesting.
Actually, a lot of people end up thinking this way (or similarly) as the lonely isolated triggers being bet over and over are doomed; the shift must them be done from the triggers to the best temporary
filtering of them.
Focusing in the moments of
application among the
temporarily enabled triggers.
Sorry, I hadn't read about what you were doing in the rolettetips section, Victor. It does sound like our ideas are very similar.
Actually, what is the difference between what you are doing and what I proposed? When you say "multiple triggers", what does this mean?
Unfortunately, I'm a cretin when it comes to computer programming so I'd have to rely heavily on someone to do this - I really would like to do this if I could.
I think there's actually a very easy way we could test this without having to do any programming. We could get people on this forum who have systems to take part in a challenge, but collectively. So I (or someone) would give some numbers and each person using their own system, make the selections. Then we see which numbers are predicted the most, and we play those. A simple test like that could see if it actually works.
Quote from: The_Force on Aug 16, 08:57 AM 2010
Actually, what is the difference between what you are doing and what I proposed? When you say "multiple triggers", what does this mean?
At some point I started asking:
- What is the most common happening at spin #2
- What is the most common happening at spin #3
- What is the most common happening at spin #4
- What is the most common happening at spin #5
- What is the most common happening at spin #6
... all the way to
- What is the most common happening at spin #37
And I started working out my array of "almost certainties".
Good thing is *any point* of the numerical stream is good to use as the "start" and "ending" boundary.
The focus being strictly the amount of numbers used in the "spin window". The numbers themselves became meaningless, the only important thing being the
events.
Let me illustrate. A clear sample being simply the "neighbors" bet.
In general, you can start by considering it like:
- Do you get a 37-spin cycle without neighbors? (so far I haven't*)
- Do you get a 36-spin cycle without neighbors?
- Do you get a 35-spin cycle without neighbors?
- Do you get a 34-spin cycle without neighbors?
- Do you get a 33-spin cycle without neighbors?
- Do you get a 32-spin cycle without neighbors?
- Do you get a 31-spin cycle without neighbors?
...to
- Do you get a 4-spin cycle without neighbors?
- Do you get a 3-spin cycle without neighbors?
- Do you get a 2-spin cycle without neighbors?
The answer is a very clear bell-curve.So when you re in the most likely spins for this event to happen, these neighboring numbers are added to the "eligible numerical pool" and are then
filtered according to criteria like:
- Is the neighbor bet currently trending? = YES
- Has it been active for too long? = NO
- Have we reached the most common point for it to balance-out? = NO
Etc. etc.
As you can see, filters are as important as triggers.
Like I said in another thread: a 1000-piece jigsaw,
but it usually pays to move yourself only in the gross of the spins where the hits are realized (i.e. if you are using the spins where +98% of the instances are realized, then it is likely you'll get a hit within these spins).
All being said, I feel confident to be selling a good tool based on solid grounds. I see no harm in me disclosing this.
Several of the triggers
I have shared already, in fact, I think I can share all of them and still by not divulging the exact filters, I wont' be "disclosing the tool".
So, there lies the difference: I have embedded a lot of these
triggers (not full-blown systems).
*: no doubt in "infinity" there is every counterpart realized, but what are the actual sequences we get to witness in our limited experience of the roulette game? This is what I focused in.
I don't deny in infinity there is the counterpart for every sequence bet, so the tool doesn't claim to beat roulette 100% (i.e. to beat infinity, I'm not claiming to have the holy grail, this isn't). But by using only the most likely statistical range and activating/deactivating by filtering the triggers according to their current performance
in the exact numerical sample being played, I can tell it does give the player a fair chance; and if roulette keeps on being roulette as we experience it, the averages will provide true in most samples and hits are inevitably going to happen...
...Unless of course, all of a sudden roulette starts changing and we only get 37-spin cycles without neighbors, or samples without a bell-shaped normal distribution, or 37-spin cycles without repeats, etc. etc. Are you following me?
Chances are roulette is going to keep on presenting the game as it has been doing so for the last centuries from the times Blaise Pascal did his studies and even millennia to the Chinese monks and before, to the times our grand-grand-grand[...]-grand children get to play it... at least in all of our current lifetime there's an almost-certain chance that we won't witness 100 reds in a row and the likes.
Granted all of those adverse sequences are realized in infinity, this is why I simply can't say "this is the holy grail, this beats roulette", because that would be akin to say: "This puts a limit on infinity, constrains it"... but how many of us will experience the true ranges of the game? The whole set of numerical possibilities realized as in a boundary-less sample? Not many of us for sure :D That's why I'm only interested in winning while I'm alive, not in infinity ;)
Regards.
Victor
winning wile u are alive or in infinity is the same thing....and u know that
Quote from: Jordan on Aug 16, 09:45 AM 2010
Winning wile you are alive or in infinity is the same thing....and you know that
I won't deny a single number is an independent event on its own.
In fact, the same number has the POSSIBILITY to come each and every spin for a roulette wheel's lifetime.
Imagine a roulette where the only number spun is "23", and the math gurus confirming that it is nothing unexpected, the game is a game of independent trials and the number's pocket never closes, so it's all within the possibilities of the game.
Granted, in infinity that sample is expected to be realized; but I prefer to go with the more "normal" samples I
experience.
I know for a hardcore mathematician this is a perfectly viable 37-spin cycle:
23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23.23
In a game of certified independent trials, mathematically this sample is not different than any other sample (you can't deny that, can you :))
But -again- if you ask me
as a player, I certainly won't bet on the next 37-spin cycle to be conformed only of the number 23.
More likely I'll *never* see such a cycle in my lifetime and I'm counting on it.
Of course it is possible to happen, I can't say with a 100% degree of certainty and a 0.00% failure range I won't get it when playing (and nobody can either). On every time I witness the number 23 being spun, there is the possibility of such string coming. as possible as it is, I still would prefer not to bet on it and instead to use the events in a more regular-looking string such as -for instance- this one in wiesbaden:
link:://:.sbwb.de/Permanenzen/tag.asp?tisch=3&tag=2010.01.01 (link:://:.sbwb.de/Permanenzen/tag.asp?tisch=3&tag=2010.01.01)
[attachthumb=#]
I guess you can't blame me for working with what I have in front of me rather than theorizing how many extremes will I experience in a gazillion-zillions numerical sample :D
Victor
I just read through all the threads about the tipping service and I very impressed. . . I didn't even know any of that was going on :D. Although it is in principle very similar to my idea, I very much like the logic behind your triggers and I think they have a higher chance of predicting the correct numbers than an amalgamation of systems (which may or may not be based on logic).
Certainly, I would be interested to see how your tipping service goes first before actual systems are tried.
I wait for the release of your tipping service :thumbsup:
Quote from: The_Force on Aug 16, 10:28 AM 2010
I would be interested to see how your tipping service goes first before actual systems are tried.
Yes mate, that's the right way to go. See first the principle applied and the results and then try to pursue an implementation of your own.
Quote from: The_Force on Aug 16, 10:28 AM 2010I think they have a higher chance of predicting the correct numbers than an amalgamation of systems (which may or may not be based on logic).
I have found my way with using very specific straight-up triggers, making the filters add the "situational malleability".
I encourage you to EXPLORE that path; you never know where the next "roulette breakthrough" may come from :)
Kind regards,
Victor
Interesting thread. I found another article on the power of averages here (link:://arstechnica.com/security/news/2008/01/face-recognition-app-reaches-high-accuracy-by-being-average.ars).
Of course, you can't take the analogy too far, because in roulette we're interested in predicting events, not merely recognizing them.
It really doesn't take too long to learn that in roulette, sticking to a single strategy will get you into trouble sooner rather than later. I've been doing rather well for a long time by using a combination of "systems" or "triggers" or whatever you want to call them. My approach is similar to Victors, but I prefer to only make bets when the majority of events "point" to them. It's quite difficult to explain in detail, but take a simple example of R/B. Probability theory "predicts" that the outcomes will be distributed in a certain way, but there are many ways of interpreting the stream of R/B - you can see it as number of B vs number of R, number of Streaks vs Chops, number of repeated patterns of length 3, etc etc. All of these interpretations have their own probability distributions and there are times when they're in conflict and times when they're in harmony. My strategy is to bet when the majority of independent "interpretations" are all "indicating" the same outcome (based on averages), which might not necessarily be a simple R or B, it might be a series of bets.
So in a sense, I use a kind of averaging technique. The only times I've ever come unstuck is when I've become fixated on a particular event which fails to materialise. :D
Over time I've learned that the more events you have at your disposal, the better. This makes "manual" operation practically unworkable, so a software "tracker" of some kind is a must.
Yes, I KNOW the maths says this shouldn't work, but I'll continue to play this way until it actually doesn't. :)
Interesting...very interesting. So we've all separately arrived at the same conclusion.
After the tipping service is released, I'll hold little competition to test my idea. I'll offer Ã,£10 per person, up to a maximum of 10 people who TOGETHER, can make it through my 1000 dublin bet spins and come out in profit.
I'd like to see if, as a team of between 5-10 people using different systems, they can collectively predict my stream of numbers. Send me a pm if you're interested (max 10 people and must have different systems)
In the SR forum I suggested to make an option to let several systems work simultaneous and then go for the "majority" of bets. So if 5 systems say bet Black and 2 say bet Red, it'd bet Black.
I'm interested in your offer The_Force. :thumbsup:
This is an interesting read and I have thought about it to. I'm up to test this with you TF, if you want to :)
Thanks Anima and Fripper - got you two noted. Need probably min 5 five people for it to work.
You just need to decide which system you will use. It has to have some kind of bet selection, so it can't be "just keep betting red". And obviously you'll have to use different systems.
I want to see how the tipping service gets on first before I do this, so will have to wait a while. Oh, and in the event that you make it through my set, I will need paypal accounts to transfer money.
I'm just curious to test my way of play with your spins.
Is possiible to have the txt file pls ?
Many thx.
A possible way to go with something like this (guaranteeing transparency both sides) would be:
a) To publicly provide as many hash signatures of the text file file as possible (MD5, CRC, SHA... etc). This guarantees changing a single bit of the file would change all of the hashes, and it would be impossible to change the file to fool all of the hashes via a (theoretical) collision attack.
b) To provide numbers in the file 1 by 1, in a thread destined for it. Once the bets have been called the thread is closed and remains locked until the next number and bets.
c) Once the challenge ends the actual text file is released to the public and everybody can re-generate the hashes to be confident it is the exact same file they have been dealing with.
It doesn't get more transparent than going like this!
You can find several freeware hash generation programs at the link below:
link:://free-hash.qarchive.org/ (link:://free-hash.qarchive.org/)
The "Hash Calc" program should do a good job:
[attachimg=#]
Regards.
Victor
Yes, the numbers should definitely come one at a time, because with with 5-10 different systems all in play, a single number could trigger multiple systems. But getting through 1000 numbers like that is gonna need more than 1 number a night, otherwise the challenge will last 3 years lol.
The hash idea sounds good. I mean, I don't have to be the one who gives the numbers. I can send the file to anyone not involved and they could post the numbers. Really don't mind.
The most important thing is that the people involved agree on the numbers to play and it's up to them how that's gonna work. So if all the players post their numbers based on their systems, someone needs to decide which ones to play (so it can be like "any number that is predicted by 5 or more of the systems") and then play them.
I also think that it should be flat betting, as we are looking to see if this increases accuracy, and not just whether you make a profit or not.
TF,
Is there any particular reason why you want to use your personal numbers for this? I'm not suggesting anything fishy is going on, but wouldn't it be better to have some kind of link up with participants and play live spins in real time? just a suggestion. Or perhaps I'm misunderstanding the point of the exercise.
There is absolutely no reason why I need to use my numbers. I just suggested my numbers because I have quite a lot of Dublin bet live spins available, but of course we could use anyone's numbers (as long as they're definitely live spins).
The reason why we couldn't do this live is because as we have at least 5 people playing and at least 5 different systems, I can see that it will take quite a while for each person to figure out which numbers to place, then it will take more time to collectively decide which numbers to play. Remember, we are not just playing all the numbers that the systems predict, but we are only playing the numbers that the systems predict will be most likely and filtering out the least likely numbers. I don't think there would be enough time in between spins if we were to do this live (hence why this would need to be programmed into a bot or something for it to be practical live).
But I'm open to suggestions of better ways for this to be done.
Ok, I see your point. I don't see why anyone would "cheat", even if they happened to have some or all of your spins (which is extremely unlikely) because the idea is see what the results would be taking this approach.
Yeah, it's in no one's interest to cheat. Although I stand to lose some money if they do well, I'd rather lose the money and find something useful.
Still waiting for the tipping service to get up and running first though.