Put the MG RNG to the test!

steinhaug

Senior Member
Joined
Feb 21, 2006
Location
Norway
Well, some of us in here are software developers and we sure as hell have people with math skills. Why do we not do the following:

We recreate Thunderstruck and put it to the test with a real RNG!

Zoozie could easily give us a random alrorythm, we already have the reels and we know what the win combinations are. We do not need to do anything, since the reels are already 95%. This would be perfect, and so easy for us. All we need is just a playable environment with a bankroll and different denominations.

Then we can start playing using high and low wagers and see if it plays out the same as usual! If that happends - I'll shut up for life regarding this topic. Do we still loose as much, do we get the same bad spells? I would love to see this in action, heck maby I'll even pay some flash dude to do this for my own personal entertainment!

When I think of it I could most likely create this game in PHP and MySQL which is my field, however I would think a flash version would be more "dynamical".

What about this idea? That would really stop all our mouthes when I think of it, mystery solved! If we chip in together here, I guess we spend a few hours each on the project - it sure as hell beats complaining in here for next 50 years!

Anyone?
 
Well, some of us in here are software developers and we sure as hell have people with math skills. Why do we not do the following:

We recreate Thunderstruck and put it to the test with a real RNG!

Zoozie could easily give us a random alrorythm, we already have the reels and we know what the win combinations are. We do not need to do anything, since the reels are already 95%. This would be perfect, and so easy for us. All we need is just a playable environment with a bankroll and different denominations.

Then we can start playing using high and low wagers and see if it plays out the same as usual! If that happends - I'll shut up for life regarding this topic. Do we still loose as much, do we get the same bad spells? I would love to see this in action, heck maby I'll even pay some flash dude to do this for my own personal entertainment!

When I think of it I could most likely create this game in PHP and MySQL which is my field, however I would think a flash version would be more "dynamical".

What about this idea? That would really stop all our mouthes when I think of it, mystery solved! If we chip in together here, I guess we spend a few hours each on the project - it sure as hell beats complaining in here for next 50 years!

Anyone?


It wouldn't be a good test.

Why? Hardware vs. Pseudo RNG.
 
It wouldn't be a good test. Why? Hardware vs. Pseudo RNG.

I do not agree, I am sure Zoozie or KasinoKing can give us a more than good enough mathematical equation combining with pseudo-rng which will render a more than good enough RNG for this experiment.

I could hook up a webcam outside my office which streams in realtime, we could do a SHA1 checksum of the image and use that as seed? How complicated do we need this?

If Zoozie can verify the machine being random with his own slot analyzer, surely the same RNG must be good enough for a test like this - if not you are saying pretty many things at once!

A pseudo-rng, also, works just aswell unless you know what kind of rng it is and therefore have knowledge of how to predict the sequences coming out, something we are not interested in doing in this experiment. The point here is for us gambling addicts wagering at a machine that is truly random, will we still loose, will we still win? I do not know, I for one sure would love to try this!
 
thanks KK, the reel postion map.
--
You do not have permission to view link Log in or register now.


Sorry to say it, but we need more information on RNG.

the most important is to know which method is used to eliminate "bias sequence".
 
Sorry to say it, but we need more information on RNG.

the most important is to know which method is used to eliminate "bias sequence".

Well, I call for KK and Zoozie to enlighten us with a good enough random equation for this experiment.

If it's down to a seed of some sort being the magic trick, I am pretty sure we could hook up a radio somewhere and let the static be used as a seed though the mic into a soundcard. (You know the noice, just like on the TV where there are no channels).

To be honest, we could even precalculate a million random numbers, aslong as the players in the experiment don't know them. The point here is just let us play in a controlled environment where we know it's random.

However I have a feeling most operating systems today are capable of creating a good enough RNG engine for this purpose.
 
nvm, it's a pay service for anything they don't have listed for 'free...', ie: nothing that could be used here.

(
You do not have permission to view link Log in or register now.
)

"The randomness comes from atmospheric noise, which for many purposes is better than the pseudo-random number algorithms typically used in computer programs."

Anyone care to explain how you can get randomness from atmospheric noise?
 
Last edited:
"The randomness comes from atmospheric noise, which for many purposes is better than the pseudo-random number algorithms typically used in computer programs."

Anyone care to explain how you can get randomness from atmospheric noise?


Sure:

However, there are many other ways to get true randomness into your computer. A really good physical phenomenon to use is a radioactive source. The points in time at which a radioactive source decays are completely unpredictable, and they can quite easily be detected and fed into a computer, avoiding any buffering mechanisms in the operating system. The HotBits service at Fourmilab in Switzerland is an excellent example of a random number generator that uses this technique. Another suitable physical phenomenon is atmospheric noise, which is quite easy to pick up with a normal radio. This is the approach used by RANDOM.ORG. You could also use background noise from an office or laboratory, but you'll have to watch out for patterns. The fan from your computer might contribute to the background noise, and since the fan is a rotating device, chances are the noise it produces won't be as random as atmospheric noise.

You do not have permission to view link Log in or register now.


btw, they've been around since 1997, so I think they know what they're doing.

You do not have permission to view link Log in or register now.


Interestingly enough:

Chapter 0: How It All Began... (1997)

RANDOM.ORG started during the summer of 1997 when I was working for a startup that was building an engine for online gambling. It was a small operation: just four guys constructing a prototype with the intention of raising money for further development. The engine was designed such that there was a core of functionality for which various game modules could be written. We implemented a prototype of the engine itself and a series of game modules that allowed you to play lottery, blackjack and slot machines in your web browser.
 
I should have been more specific. How do you get a random number TO your computer from atmospheric noise. Actually, never mind on that last question. I did a little research and this is way over my head. I did find this interesting

A physical random number generator can be based on an essentially random atomic or subatomic physical phenomenon whose unpredictability can be traced to the laws of quantum mechanics. An example of this are the Atari 8-bit computers, which used electronic noise from an analog circuit to generate true random numbers.

ANyone have an Atari 8-bit computer laying around? :D
 
Hey, I'm sure there's enough slotjunkies on here to submit real statistics for real money over a given period of time.

You could even do an experiment where, say, ten people sign up to do ten hours solidly on autoplay at 9c denomination - then compare and contrast the statistics...

Or maybe something a bit more scientific/logical?
 
Just out of interest, the RNG used in the first proper electronic real fruit machines (pre processor controlled) was usually derived from white noise
which was used to clock a shift register,the output of which provided the random number. This was used to time the solonoids which stopped the reels
ans provided the nudge and hold chances.Was a bit crude but it did the job.
 
thx for the help guys......... *run like a mother-*snip* * cuz these math guys have no idea either!
 
Last edited by a moderator:
I READ THAT THERE IS NO "TRUE RTG"........AND.....

Also, after playing online for over 5 years at Micro. casinos I can attest from personal experience that it is not a true RTG in the meaning that we players think. The outcome is not truly random if the outcome changes based upon the amount you bet per line. I guarantee, though not scientifically but basedon experience, that the outcome does differ on the amount you bet! I would bet me life on this!.

Also, there are too many noticible variables such as some games on a given session pay more when you bet only 8 lines not 9 or 24 lines not 25 yet you hit a combination that would have been a winner at either line bet. (but probably wouldnt have got the payout at 9 lines)

I have no proof but only thousands of bets to come to this conclusion.

I did read however on a Mathematics website that there is no "true" RNG as it does not exist.
 
Also, after playing online for over 5 years at Micro. casinos I can attest from personal experience that it is not a true RTG in the meaning that we players think. The outcome is not truly random if the outcome changes based upon the amount you bet per line. I guarantee, though not scientifically but basedon experience, that the outcome does differ on the amount you bet! I would bet me life on this!.

It's definitely doable.

Take Mega Moolah for instance. Of course you can't hit the 'big one' unless you're betting the maximum amount....

I think we'd all love to see proof that backs up these theories..:)

The problem is, IMO you'd need millions of spins to come to any concrete conclusion, not thousands.
 
YOU HAVE TO CONCEED THAT IT DOES MATTER HOW MUCH

YOU ARE BETTING PER LINE, RIGHT? Of coarse to win a progressive you have to wager max. But I can guarantee you will have more winning payouts betting less than the max when playing progressives, for the very reason you state WINBIG.

Per spin, you will hit a payout less on progressives when betting max.

Some games ALWAYS pay more frequently if you bet less lines than the max, Cashapillar is one such game as is Gladiator and a few others that are 40 line slots. You have to play around with your bets for a bit to determine what combo of lines and bet per line are paying off. So, if it was truly random, and you are playing 9 line on a 9 line slot should it matter how much you are betting per line? I say no if truly, purely random. Yet I guarantee you have a higher chance of hitting any payout betting less than the max per line.
\
What do you think, anyone!
 
YOU ARE BETTING PER LINE, RIGHT? Of coarse to win a progressive you have to wager max. But I can guarantee you will have more winning payouts betting less than the max when playing progressives, for the very reason you state WINBIG.

Per spin, you will hit a payout less on progressives when betting max.

Some games ALWAYS pay more frequently if you bet less lines than the max, Cashapillar is one such game as is Gladiator and a few others that are 40 line slots. You have to play around with your bets for a bit to determine what combo of lines and bet per line are paying off. So, if it was truly random, and you are playing 9 line on a 9 line slot should it matter how much you are betting per line? I say no if truly, purely random. Yet I guarantee you have a higher chance of hitting any payout betting less than the max per line.
\
What do you think, anyone!

I totally agree footdr, but you might as well be pissing in the wind to try and prove that theory and then you will also have several here that will come along and swear on their life that this software is truly, purely random and not weighted...

The other thing I would like to know is who exactly tests the software providers RNG's, who are these people and what type of tests do they run on the RNG's that MG and the other software providers use in their programs...
 
I totally agree footdr, but you might as well be pissing in the wind to try and prove that theory and then you will also have several here that will come along and swear on their life that this software is truly, purely random and not weighted...

The other thing I would like to know is who exactly tests the software providers RNG's, who are these people and what type of tests do they run on the RNG's that MG and the other software providers use in their programs...

Well here is my theory:
Slots are not random, and were never supposedly that either from the mechanical point of view. There are weighted rules and balancing going on all the time to meet expected payouts. Each and every purchaser of the gaming system want different payouts, and are able to do this. Some have different regulations, and need to do this.

So - here comes the internet, and people can sit down and realy analyze. The perception is that it's random, so let's make it look like it's damn close!

I belive this is what they are doing, they make it close, however it is not more random than a vegas slot!

Did a tourney last nite on 3Dice, and was amazed playing that blue butterfly thingie. Play 1 line only and keep looking at all the wins you are "missing"... Up to 25 line and you suddenly get lots of naked spins, down to 1 line again and wins all the time. This is just a little example on how the system balances out what your expected wins would be. Also when you have close wins, like 4 out of 5 - I belive many systems are buildt to "randomly" give you theese aswell, to make you want to play a little harder.

A random environment is easier to win in given you are allowed to +/- the bet sizes, only at a fixed betsize will the house have it's edge.

It's like the old roulette trick, tripple your bet and you will always double, however you reach the house limit in a while. Well, at a lower level the same works in the slots, that is, it doesn't work at the slots since some balancing scheme is in affect and wont allow this kind of betting!

We all know it, play one line play 9 lines, bet 0.09 jump to 9 - it's 4 different games right there.

Doing thunderstruck in my own served random numbers would be a great exercise for us all, but only doing +/- on the betting can you skew the house edge, and only of the numbers are truly random and not streaky random or balanced.

Anyways, the more I testplay at real random slots - the more I know on how right I truly am of my above statements!

In the end, we should't complain? It's entertainment machines, atleast that is what the makers are changing their name to in IGT, we do get out expected payout. We got what we paid for! In online terms, that should mean 92-96%

Coffetime!
 
YOU ARE BETTING PER LINE, RIGHT? Of coarse to win a progressive you have to wager max. But I can guarantee you will have more winning payouts betting less than the max when playing progressives, for the very reason you state WINBIG.


Did you take the time to see the slot I was referring to? Mega Moolah has different progressives, and each goes by how much you are betting. IE: Major, Mini, etc. If you're betting .50 a spin, you have no chance of winning the Mega progressive, but I'm sure you have one hell of a chance of hitting the mini. :rolleyes:
 
Did you take the time to see the slot I was referring to? Mega Moolah has different progressives, and each goes by how much you are betting. IE: Major, Mini, etc. If you're betting .50 a spin, you have no chance of winning the Mega progressive, but I'm sure you have one hell of a chance of hitting the mini. :rolleyes:
You are stating that as if it is an actual fact (rather than just your opinion), but I think you're wrong.

As I understand it (according to MG's rules) you have an equal chance (no matter how tiny) of hitting any of the 4 jackpots regardless of how much you are betting.
However triggering the bonus wheel in the first place most definitely IS based on your bets.
i.e. you will get the bonus wheel 10 times more often at $5.00 per spin than at $0.50.

With the increased lesser jackpots, I just wish I could get the wheel once! :(
 
You are stating that as if it is an actual fact (rather than just your opinion), but I think you're wrong.

As I understand it (according to MG's rules) you have an equal chance (no matter how tiny) of hitting any of the 4 jackpots regardless of how much you are betting.
However triggering the bonus wheel in the first place most definitely IS based on your bets.
i.e. you will get the bonus wheel 10 times more often at $5.00 per spin than at $0.50.

With the increased lesser jackpots, I just wish I could get the wheel once! :(

Thanks for the clarification. I knew that your bet size had something to do with it, but I thought you never had a chance to hit the top progressive if you weren't playing the max amount per spin. Now I know better. :)
 
We recreate Thunderstruck and put it to the test with a real RNG!

Zoozie could easily give us a random alrorythm, we already have the reels and we know what the win combinations are. We do not need to do anything, since the reels are already 95%. This would be perfect, and so easy for us. All we need is just a playable environment with a bankroll and different denominations.

I am not quite certain what tests you want to perform.

The java 'shuffle' method is pseudo-random which basically means it is good enough for any long term tests!

I can already now tell that the slot-analyser program will behave 100% identical no matter what denominations you are playing. Simply because the denomination/betsize are not used for determining the outcome. It is just used afterwards to correct the bankroll by substracting bet and adding wins.

So what you might are looking for is how the behaviour is when you start with 100 units (1000$ deposit, 10$ spins) or 1000 units. (1000$ units, 1$ spins). Because in the first case you will go bust very fast, very often.

I can easy make simulations using the slot-analyzer program and I already did a few that you think you have seen. But I need you to tell me exactly what is the start condition, end condition and what data should be extracted.

It would be take too much time to improve the slot-analyzer GUI to support these simulations as there simple are too many cases and often the simulation needs a little more programming. Instead it is much faster to just make the change and run the program without the GUI.

But you already extracted thousands of spin-data with various bet-sizes in fun mode. And I believe you saw the same 'pattern' here that talk about. So what you simply need is some statistical analysis of these data that you already have and can make more of. I did a few simple statistical tets and I saw nothing unusual. Each reel position (from the non-weigthed slots...) was within expected range and even more the different reels was also not correlated in any way. (independant)
 
Did you take the time to see the slot I was referring to? Mega Moolah has different progressives, and each goes by how much you are betting. IE: Major, Mini, etc. If you're betting .50 a spin, you have no chance of winning the Mega progressive, but I'm sure you have one hell of a chance of hitting the mini. :rolleyes:

:confused: The last $5 Million Mega Moolah jackpot was won from a 0.50$ bet...

kimss said:
Well, some of us in here are software developers and we sure as hell have people with math skills. Why do we not do the following:

We recreate Thunderstruck and put it to the test with a real RNG!

Zoozie could easily give us a random alrorythm, we already have the reels and we know what the win combinations are. We do not need to do anything, since the reels are already 95%. This would be perfect, and so easy for us. All we need is just a playable environment with a bankroll and different denominations.

Some months ago I programmed MG Tomb Raider I slot which uses ASCII graphics and behaves 100% identical to the real version. It's not too difficult to use this program code to simulate any other slot by just changing the reel layout string which is read from a text file. Of course special bonus features such as the idol picking game have to be programmed separately.

Here is a screenshot from the game. Nice graphics, eh :p

You do not have permission to view link Log in or register now.
 
I can easy make simulations using the slot-analyzer program and I already did a few that you think you have seen. But I need you to tell me exactly what is the start condition, end condition and what data should be extracted.

Well, the betting scheme would be something like this:

You start off with 1 line 0.01 size, and spin away. You keep track of the other lines, so that you are aware of what you would have won if you were playing them. Now say that you get 100 spins without any real win on any lines, this would be an indicator to switch on all lines - or only 5 for higher variance.

If you get another say 25 spins without any paricular win, up the betsize to 9x2x0.01, after another 25 up again. As soon as you hit 4 alike, down to one line again.

Do not care for the freespins really. 5 lines would probably be better since it costs less. However if you managed to get say 500 spins without a good win on line 6,7,8,9 you would enable all 9 lines until hit here (or in the 1-5 lines which would make you go down again).

This system works pretty well locally on my machine in a random slot, however a slot with reelbands of 100 would need some computation to know what should be the amount of spins before statisically have skewed the house edge so that you would up the ante.

The reason why this shold skew the payout in a random - but please correct me if I am wrong.

Let's simplify the statistics, and lets say a cycle is 1000 spins, and we have a 1 liner slot.

Given 95% payout, wagering $1 a thousand times should pay $950 in wins.

Since I can up and down the denomination, If I start of by doing a 100 spins on 0.01 without any special wins at all - say 20% payout - upping and doing the rest on $1 would absolutely give me the edge.

If I were to miss a win, say I got a win on 100 units on the 0.01 bet, that means I would have to spin this away whit doing 100 spins atleast, also adding up all wins in theese spins and also spins theese away. After that I would need to build a 40-50 atleast loose spins in the house favour, before incrementaly building again.

A very tedious job, but done carefully and not greedy you should be able to skew thie in your favour.

As noted - it seems this works bery nicely locally with a local slotversion using the CPU random functions, however in MG it doesn't seem to work at all since the system seems far to streaky. If the system has decided to give you bad spell - you get it and for 500 spins easily!

How off am I in my calculations here Zoozie?
 

Users who are viewing this thread

Meister Ratings

Back
Top