hi
How many spins do I need to test a system?
Quote from: nkmagic on Sep 17, 06:18 PM 2013
hi
How many spins do I need to test a system?
Apparently 10,000 PLACED bets is the benchmark..... :thumbsup:
good luck 8)
O0
Quote from: ddarko on Sep 17, 09:04 PM 2013
Apparently 10,000 PLACED bets is the benchmark..... :thumbsup:
good luck 8)
O0
and what if 10.000 spins will be very good for my system and other 10.000 will be killer?
Quote from: biagle on Sep 18, 03:36 AM 2013
and what if 10.000 spins will be very good for my system and other 10.000 will be killer?
-------------
after 10k placed bets...quite unlikely...even thou, how frequent will it happen?
vundarosa
Quote from: biagle on Sep 18, 03:36 AM 2013
and what if 10.000 spins will be very good for my system and other 10.000 will be killer?
Just curse your luck. Seriously. You can go on and on testing billions of spins, but the first time you play with real money a killer sequence comes through. You cant really help it. I always do 2 types of testing for my ways of playing.
Testing based on real life scenarios- how many spins would you ideally play/I play normally 100-250 spins, worth around 4-5 hrs in a casino/test 7-10 such sessions on your system with 100-250 spins each. Testing for extreme conditions â€" 1000-2000 spins will give a good idea of extreme conditions that you should not ideally cross. you can play 10000, 1mn, 1bn spins to see more extremes, but if you try to play within the extremes that you have faced within this 1000-2000 spins, good enuf. Ball doesn’t stop there. Every day you play in real is a test bed on its own/you need to be able to learn from those everyday play and reverse feedback into your system of play. above all these, the system should have multiple facets. It cant just chase either hot or cold, or it cant just use a static progression, it cant bank in on one betting position. it should be able to adapt based on whats happening in the table. 10-15% of the times, whats happening in the table might completely reverse when you play but you will come a winner the rest of the sessions. 4 out of 5 sessions winning? Good ratio.
Many test i did with my both... 10k spins, great result, select 2k extra spin, failed result... so 10k can give indication... but next 1k can kill it all...
Quote from: nkmagic on Sep 17, 06:18 PM 2013
hi
How many spins do I need to test a system?
all systems fail so whats the point of massive tests. A system shouldn't be mechanical...or formula based. You will die if it is.
So you need to bet when conditions are favourable. Say...a repeating number, or a returning sleeper, or something else based on what random does....like waxing and waning SD's (I think thats a new phrase :smile:)
You shouldnt be saying "I am going to do this...then that...then the other" without seeing whats happening.
So the right amount of spins to test an
idea is "
some spins after a favourable event" and not "
hope to catch a favourable event (fingers crossed) with a mechanical repeating system"
I think that just about sums it up really.
Quote from: biagle on Sep 18, 03:36 AM 2013
and what if 10.000 spins will be very good for my system and other 10.000 will be killer?
I got the figure from "Herb" over @ VLS. He is/was a pro player with Advantage-play & ALWAYS poo pooed ANY mechanical system & in fact anyway of playing apart from Advantage-play.
So, if he says 10,000 placed bets that's good enough for me.
O0
Realistically 10,000 is nowhere near enough. He was probably referring to bias analysis for which 10,000 spins is enough to realistically determine if a "realistically" legitimate bias exists, assuming you aren't using any kind of secondary verification.
But for system testing, you actually realistically need hundreds of thousands of spins. Many players contact me claiming they can beat 1m RNG spins for the challenge at link:://:.genuinewinner.com/100k-challenge/ and then I send them 100k more spins to test the exact same system, and the system fails. This is because they probably built their system around their 1m spins and if you build a system around specific spins, of course your system will win.
Will the average player ever play more than 50k spins in their life? Unlikely. that's why some people say you don't need a system to beat 50k spins. Well, true. but what if 100 players all with the same system say this? Then you will have 55 winners that all think their system works, and 45 losers that believe the system loses. And in the middle is the casino who understands there will be both winners and losers, but the majority are losers.
Basic fundamental facts that most players don't understand.
Anyway the original point is to be sure about a system, you really need two completely independent groups of about 1m spins. If you build the system around one of the 1m groups, then that same system should work against the other 1m spins.
Another thing to understand is even if you test over 10k, 1m or even 2m spins, does it mean the system definitely works? No. It just means the more spins you beat, the higher the probability that the system is effective.
Quote from: Steve on Sep 18, 06:19 PM 2013
Realistically 10,000 is nowhere near enough. He was probably referring to bias analysis for which 10,000 spins is enough to realistically determine if a "realistically" legitimate bias exists, assuming you aren't using any kind of secondary verification.
No, I can assure you he wasn't referring to bias analysis
Another thing to understand is even if you test over 10k, 1m or even 2m spins, does it mean the system definitely works? No. It just means the more spins you beat, the higher the probability that the system is effective.
Couldn't agree more :thumbsup:
O0
If he really said 10k spins then he's wrong. Even just betting red over 10,000 spins can give you a profit.
Quote from: Steve on Sep 18, 07:18 PM 2013
If he really said 10k spins then he's wrong. Even just betting red over 10,000 spins can give you a profit.
There is no IF about it Steve..... :thumbsup:
& I'd doubt somebody would test a "system" of just betting red do you ?
O0
A system being "just bet red" would be pretty bad yes, but I was just making a point.
Could some one send me the amount of spins you want me to test it on and I will get right on it and tell you how I did... ?
A system may give u a profit after 1 m spin on one wheel and still be a loser on another wheel.
Basically no mechanical system can give u a profit after consistent betting for many spins on many different wheels. You will always lose in a long run. Your only hope is not to be such naive as to look for a system which will work consistently.
Talk to any person who really played roulette a lot in a real casino, and he will tell u that playing to many spins consequently is mostly a suicide. Well, for your bankroll of course.
All said about mechanical systems, not AP of course.
For example if u found a bias in a wheel u may play lots of spins and still be well ahead. And even if u found a bias, it is still not a guarantee.
And some of you still hope to defeat on a long run many unbiased wheels with mechanical system consistently and consequently. That's so naive.
I knew one guy in VLS forums which claimed he won 2k with martingale...
he had been winning for some time, not just for a day or two...
It's very possible that he passed 10K spins...just to find our later that it was not worth it.
Okay let me put it like this then.
Send me any 10000 spins from a real wheel of course and I will bet you that I beat them all.
what do u mean "i will beat them all"? you get 10 K spins, and show us them beaten?
Isn't it too easy when you already have those spins? In a casino u don't know even the next number to come, let alone 10k numbers which will come next :)
There was one guy here, he tried to sell his roulette systems. He said: "Send me the spins and i will show u my results". When asked "how the hell can we know that these are real results you would bet? Show what you gonna bet next when u get spins you need to track (but not the next ones)", he said: "I can't show you what i bet, but i can show you my winning results. We are gentlemen here, I am not gonna lie".
Okay then how do I test a system without using real money?
I mean that I will come out with a profit on any 10000 spins that I get ( if the spins are real and from a live wheel and not RNG)
OK. My advice to you. Most here won't agree though.
but try this. Take german casino data and work with all the wheels there. And give up any idea about consistent winning.
try to make small profit then quit the game for a while to come back later. As most of real casino smart players (not Advantage-play) do.
when you see it's working well on different wheels, play free at those online casinos:
smartlive, dublintbet, castle casino.
It may take years though before u come up with something worthy. It may never happen as well.
I have used german casino data and dublinbet and my system is showing consistent winning.
I just ask how many spins should I test in a row to know if it will beat the long run?
nkmagic - I suggest you play it in sessions as you would when playing with real money. Decide which casino you will play at and use their spin rate to test.
Some online live dealer games average 1 spin per minute = 60 spins per hour. @ 5 hours of play = 300 spins. I play at Party Casino at that rate, although they are not test/free play friendly as they block the image with no money in the a/c. Their table limits are better than many. 1 - 250 GBP for inside bets & 2 - 2,000 GBP for all outside bets.
Smart Live is test friendly but their plays are slow = about 1 spin/2 minutes = 30 spins per hour about the same as a light/medium busy time at a real chip casino.
I don't see the value of testing for thousands of spins when most players only usually play daily in the hundreds +/- ? Test it like you'll play it.
Using Party casino as an example (their play runs 24/7), in a year the spins will exceed 1/2 million @ 1,440 per 24 hours.
Even if you play for 2 x 5hr sessions = 10 hours, the other 14 hours of spins aren't too relevant if you won't be playing them.
Runs from HELL can happen at any time that will kill most plays/systems. A good system may include a safety valve where you stop betting before losing too much & virtual bet until the usual spin patterns return.
no mechanical system can beat roulette continously on a long run. People who think it can -- are wrong. You will see it later. Just friendly warning. Don't hold your breath and don't bet too much because u will finally lose. On a long run -- house always wins. This is a fact.
It does not mean though that it's impossible to win in roulette. You just have to adapt your tactics to reality.
nkmagic
There are two schools of thought:
1. If you test enough numbers, you will find what your system does over time.
2. We already know what it will do. The run from hell will get you.
Let's say you run 10,000 spins 100 at a time trying to find out which hundred contains the run from hell. OK, so what if you find it? It may have been in test 900. If you try a new set of 10,000 spins, it may be in test 9. The run from hell comes randomly just like everything else; it cannot be predicted.
So the game is not to beat the good times or even the normal times--the game is to beat the run from hell. How do you do that? I wish I knew.
I can tell you this for certain--endless testing is futile.
THE ABOVE IS MY OPINION ONLY; I HAVE NOT BEEN UP THE MOUNTAIN!
Sam
i will try to put it this way. Many people just can't embrace the reality. The reality is that roulette can behave totally different at different times. 10k or even a million spins on the same wheel may not encounter things which will happen at another wheel (or even at the same one) just 5 minutes later.
There is NOTHING that can beat all the wheels mechanically on different wheels for long periods of time frequently.
I will try this allegory.
Let's say that you have a distant relative which u visit once in a two years -- for a few days. And when u come to see him once in two years -- he is as nice as an angel. But in reality he is mentally ill and may kill or injure someone when hе got attacks of his mental disease. His attacks doesn't come too frequently but they may happen any time. So otherwise he is a very nice and sweet person and loves you. So when you come to visit him once in a couple of years for a week, you see a nice guy, and that's it. It happens that you will never in your life see him during his crazy "attacks".
But then instead you start visiting him once a month for a couple of weeks. Guess what will likely happen sooner or later?
But again -- it may happen that you will be lucky to never see him in that "state of mind". But if you are unlucky he can take a knife and kill you some time. And it may happen twice or three times in just a few days, but then again he will be nice for long time.
When you test roulette on million spins in a row and expect that it will be always like your million of spins -- think about what I told u.
So roulette is a personality disordered knife wielding maniac!
Next time you go play you’d better take a gun.
LOL.
you got it