A perfect rigged slot - this is how you make it!

steinhaug

Senior Member
Joined
Feb 21, 2006
Location
Norway
To make my words a little easier to understand, let me write a logical example. This could be programming in any language, however I hope you have a basic understanding on how programming works - and this example would only be the low-end logic oversight.

Proove my logic wrong - how would you detect that this system was rigged?

Basic assumption of a slot:

PHP:
function slot(){
  get_numbers();
  compute_spin();
  update_balance();
  update_graphics_with_numbers();
}

function get_numbers(){
  // Return random numbers
}
function compute_spin(){
  // Calculate if the spin is a win or not
}
function update_balance(){
  // new bank balance
}
function update_graphics_with_numbers(){
  // Updates the GUI, what we persieve as the game infront of us.
}
Also - note that the get_numbers(); function is most likely proprietary, and inside a company like MG this function
is most likely only known by a few. We will never know the story here, we will only know the outcome...

However we already know that all your history is stored by the playcheck data. If you ever asked CS for a dumpt of your data
you will see thay have a nice overview on lots of % payouts, deposits, manager bonus +++ at their disposal. So infact - the
system already have all the calculated stuff freely available at their need, so It is fairly easy to tweak the system
and still it is impossible to detect the rigged system.

Let's assume we tweak the get_numbers() to something like this:

PHP:
function get_numbers(){
  //This function is really extended by the really_get_numers()
  // Remember this function is proprietary, so nobody knows what is happening
  // here. It might aswell be something as this.
  // New logic in place is:
  var userstate = compute_play_mode(); // Three states: default, loser and winner
  if(userstate == winner)
    var maxwin = compute_winner_max_win();

  if(userstate==default)
    return really_get_numers();
    
  var rnd = really_get_numers();
  var win = calculate_winX_compared_to_bet_size();

  if(userstate==winner AND win >= maxwin){
    return really_get_numers();    
  } else if(userstate==loser  AND !win){
    return really_get_numers(); 
  } else {
    return rnd;
  }
}
function compute_play_mode(){
  // Poll earlier playing data and descide if we have a winner or a loser
  // at the moment. This would be a complicated part of the system with
  // alot of weighted triggers depending on this and that. However the
  // general idea is here.
}
function compute_winner_max_win(){
  // Calculate, depending on previous function what wins we will ignore if
  // they would occure.
}

function really_get_numers(){
  // Return random numbers
}
This would be perfect for audit, peferct for all since all numbers are 100% random in the players end.
However you clearly see that the system is rigged, and the casino assures steady players getting more luck when
losing and harder time when winning! In the end a quicker way to meet expected 95%.
 
I think everyone understands that it would be incredibly easy for software companies to incorporate this type of rigging into their codes and be totally undetectable to anyone playing the games.


Speaking of rigging; I spotted something VERY worrying last night when I played MGs new Wealth Spa slot;
Previously most here accepted that MG bonus slots produce totally random spins - both in normal play, and in free-spin mode. But this one actually states that this is not the case.
So is it a slot, or is it an AWP? :confused:
I can't see anywhere where is says it is an AWP...

Old Attachment (Invalid)
Old Attachment (Invalid)

By the way, for anyone thinking of trying this slot, it took me 543 spins, 222.20 in wagers & nearly 70 down to get this bonus. It gave me 16 free-spins which returned 97.20 on a 1 spin.
Oh, and it's possibly the most BORING slot MG have ever made!

KK
 
Speaking of rigging; I spotted something VERY worrying last night when I played MGs new Wealth Spa slot;
Previously most here accepted that MG bonus slots produce totally random spins - both in normal play, and in free-spin mode. But this one actually states that this is not the case.

Derailing in first reply... Typical!

Anyway, this slot obviously take advantage of excactly what I am assuming above! When and where the coins appear are abviously connected to your previous play. Just like picking all the first coins at $0.01 and getting the freeplay at max lines... Go figure! I am not even referring to the possible outcome, but the accual outcome from the spins at winning combinations from a "cheap bonus" compared to an "expensive" bonus.

I have my own statistical graphed outcome from quite a few runs on this machine, just as with tomb raider. It's weighted all the way to the bank. I wonder why you are even confused to be honest. (It's no news either, the MG system itself have weighting already in place on several slots, so why not take advantage of the existing system...)
 
I'm not a slots player on a regular basis, but I've tried some in the past. It would be safe to say that with programming everything is possible, no secret there.

From what I've played, the inconsistency that I could see was between betsize. I'm not surprised or anything by rigged slots, they did that with the real slot machines too right? Paying out whenever the balance was at a certain positive amount.

Anyway I'm no genius on this subject, but I'm sure its controled in some way or another.
 
KimSS,

With all due respect, but what you suggest in your post would be detected by even the most naive test, and there is no way on earth that it would pass the die-hard tests (by Marsaglia), that are industry standard today for randomness testing.

In fact the algorithm you suggest would fail nearly _every_ test in the die-hard test battery. And if I'd have to think of a specific test to disprove randomness in the scenario I think you suggest it would be to plot out the distribution of the distance between wins of a certain size - which should be normally distributed with characteristic mean and sigma. (and it most certainly wont.)

Furthermore, to pass any type of certification, it are exactly the pieces of code that draw the random numbers _AND_ their calls from the game logic that are being certified and fingerprinted. No self respecting auditing company would certify any RNG without reviewing exactly those parts.



KK,

If I'm not mistaking, this concept was introduced by MG before - don't remember which slot, maybe it was TR2? Either way, since this feature will actually be triggered over multiple spins, it is an obvious solution to players stacking up 'triggers' on lowest stake and switching to higher stakes when they have only one trigger left to go. I would assume it takes the average bet of the actual spins that get you the triggers, not all spins inbetween, and rewards freespins based on that. Since it were in fact e.g. 5 bets that started that feature, it seems like a fair solution to me to say you get free spins at (total bet over those 5 triggering spins)/5.

Cheers,

Enzo
 
To make my words a little easier to understand, let me write a logical example. This could be programming in any language, however I hope you have a basic understanding on how programming works - and this example would only be the low-end logic oversight.

This would be perfect for audit, peferct for all since all numbers are 100% random in the players end.
However you clearly see that the system is rigged, and the casino assures steady players getting more luck when
losing and harder time when winning! In the end a quicker way to meet expected 95%.


So basically what you are saying is that when the player is losing, the machine will toss them some wins, and when they are winning, the machine will win it back?

What's the point?

The casino doesn't want to meet expected 95%. Why would it? Some players will win $50,000, great, they flash it up next time you login 'XY from Z-Town won $50,000 playing MegaSlot - why not try your luck'?

Over an extended period of time the casino will achieve an actual % win that's more and more likely to be very close to the theoretical win. There will be a little noise and random variation along the way of course, but it's nothing to worry about.
 
@Enzo

Lovely to see an answer by someone in the business! Thanks for this.

I would assume it takes the average bet of the actual spins that get you the triggers, not all spins inbetween, and rewards freespins based on that. Since it were in fact e.g. 5 bets that started that feature, it seems like a fair solution to me to say you get free spins at (total bet over those 5 triggering spins)/5.

I bet you do. While this sounds like a grest explernation and a logical one to, acctual gameplay contradicts it. What you are saying is that getting the 5 triggers on a $0.60 bet would normally be a good win, atleast could be. I would say that usually if you lose a lot inbetween you complete the feature! If you manage to get tham luckily fast (ctalking about the triggers not the spins in between) you get LOSE in first two ones! If you keep playing this slot, and keep track of notes (or even my sweet graphs) you cant help feeling you always get beat on the finnish line... Sopme magical unluck, always in favour of the computer.

Regarding your comments on my quick system, are you saying that when Marsaglia are doing tests they are infact reviewing the RNG engine? Do they get access to the accuall code? Are 3Dice code reviewed?

My above example is failry an easy example, you could easy add skewing of luck - and you could also add it only sometimes - randomly. Are you telling me that this is impossible to do, it sounds very easy to me to fool a die hard test with some fuzzy logic accompanied with some random triggers.

To be honest, I am sure I can fool this test myself, die hard or not! I am sure I can produce a million spins for you and you will not be able to tell me that this is rigged, however I will be able to skew the end result being either lucky or unlucky! Cause that is the problem here, it is possible to be lucky and unlucky within the realm of random.
 
So basically what you are saying is that when the player is losing, the machine will toss them some wins, and when they are winning, the machine will win it back?

Not quite...

The system doesnt toss wins really, the system could forget that spin and give you another try without you knowing - thus doubling your chance of infact getting a win, that doesnt mean you get one. A lucky streak. It's still 100% random ironically.

(The system checks if you win, and if you didnt discard the result and give you a new random spin instead. If you win or not is beside the point at this stage, the system has just randomly doubled your chance of winning).

Surely there cannot be a pattern on how this is applied, but this is only dynamics really, how you apply your code and is beside the point and when it triggers and with what weight it triggers. The point is it is extremely easy to throw some luck at the player, or on the other hand - take away some luck.
 
About Wealth Spa - the "average bet" during free spins is there to prevent abuse by getting the first four coins at minimal bet then the fifth one at a more higher bet - like Tomb Raider II and the five passports to collect - quite fair to make this , i'd say. (It seems that the "average bet" is (sum of ALL bets placed from first spin to last coin earned)/(number of spins) )

For the second mine , it's there to avoid the case of a disappointing zilch win after collecting all the five coins (quite an upgrade from TR II) that would be very likely to make the player angry and leaving the casino.
 
KimSS,

With all due respect, but what you suggest in your post would be detected by even the most naive test, and there is no way on earth that it would pass the die-hard tests (by Marsaglia), that are industry standard today for randomness testing.

In fact the algorithm you suggest would fail nearly _every_ test in the die-hard test battery. And if I'd have to think of a specific test to disprove randomness in the scenario I think you suggest it would be to plot out the distribution of the distance between wins of a certain size - which should be normally distributed with characteristic mean and sigma. (and it most certainly wont.)

Furthermore, to pass any type of certification, it are exactly the pieces of code that draw the random numbers _AND_ their calls from the game logic that are being certified and fingerprinted. No self respecting auditing company would certify any RNG without reviewing exactly those parts.

Hmm, I've never seen any credible claims by a casino that their code is sufficently well audited to guarantee against cheating. Not that I'm alleging that they are cheating, just that I don't think the 'we are random' audits prove much.

I had a look at your site, and it doesn't say much

"How can I be sure the games are fair?
All software has met strict criteria to guarantee fairness for our users."

'Strict criteria' is a completely empty claim that means nothing.

Intercasino makes more specific claims, which are among the most specific I've seen:

Old / Expired Link


"2.How can you prove that the numbers generated on InterCasino are completely random?

A random number generator is used to determine the outcome of all our games.

The algorithms in the random number generator have been proved to conform to results of similar games found in Atlantic City (with respect to randomness and payout frequency), in line with our license requirements.

In some games the rules differ slightly from the Atlantic City rules. For example, in some card games, the deck is shuffled after every hand. Any exceptions from Atlantic City rules and regulations are noted in the game rules section of the website.

A typical test of the current version of the random number generator involves a sequence of 300,000,000 (three hundred million) numbers.

The random number generator is tested on an ongoing basis to ensure its continued fairness.

A typical test of each game involves a sample size of between 1,000,000 (one million) and 10,000,000 (ten million) bets. If the frequency of winning combinations and general payouts are within expectations, the game is passed to a Quality Assurance team of engineers who play the game (with the real gaming server) until they are sure that the game is flawless.

Based on these tests, the algorithm for each game has been found to be fair.

The software has also been independently tested and certified by an international gaming-testing house to ensure it is fair and completely random.
"

Exactly what the international gaming-testing house has tested isn't made clear, but I'm assuming it's the RNG only.

A later statement appears to confirm that:

"Each game at InterCasino undergoes extensive testing. First, we imitate all possible situations and determine whether the rules are obeyed, payouts are properly paid, etc.

The random number generator is then added and tests are run with automated agents. For games which involve strategy (blackjack, poker, etc.), the optimal strategy is implemented, and for games of chance (red dog, baccarat etc.) a random bet choice is implemented. Depending on the game, between 1 and 10 million bets are run.

If the frequency of winning combinations and general payouts are within expectations, the game is passed to a Quality Assurance team of engineers who play that game to insure its flawless play.
"
In other words, while the RNG is assured to be fair, the game testing is strictly in-house; similarly, I seriously do not believe that Microgaming send each of the slots they spew out each month off to be externally audited before release.

And the payout audit would not appear to be much of a guarantee either:

"10.Are InterCasino payouts reviewed by a third party?

Yes. With one of the world's leading global accountancy firms, provides a monthly review of InterCasino's payout percentage. They examine the data log files for each month and certify the average payout percentage by game.
"

They are certifying payout percentage. Not really rocket science. Intercasino's bj is paying out 98% - is that through bad player strategy or because the casino is cheating (e.g., through a subtle weighting on certain cards, which could be done, despite having a certified RNG)? I'm sure it's just bad strategy, but a simple payout audit does nothing to prove or disprove it - you would need to conduct analysis of card distributions, including analysing different bet sizes, and many other things, and as Intercasino are not claiming that their auditors are doing that, it is very unlikely that they are.

The reality is that if a casino wants to cheat, no third-party company can stop them. Logs can be falsified, code can be switched.

The only true way to catch cheating is logging your own results with real money.
 
I think everyone understands that it would be incredibly easy for software companies to incorporate this type of rigging into their codes and be totally undetectable to anyone playing the games.


Speaking of rigging; I spotted something VERY worrying last night when I played MGs new Wealth Spa slot;
Previously most here accepted that MG bonus slots produce totally random spins - both in normal play, and in free-spin mode. But this one actually states that this is not the case.
So is it a slot, or is it an AWP? :confused:
I can't see anywhere where is says it is an AWP...


It's a slot. The chance of hitting the free spins is random. But because you need to get 5 wins to get the bonus, they take the average bet of all 5.

Which is no difference to me playing a normal game at 9 lines, 1 coin and getting 3 scatters to enter a bonus game which pays 500 coins. If I had played 5 coins, then I would have won 500*5 = 2500 coins.

The payout of bonus games are always multiplied by your coin size and number of coins played. Nothing unusual.

The payout of the game is still random - it's just a different method of calculating what to multiply it by.
 
The only true way to catch cheating is logging your own results with real money.

I am, and which is the basis of my thesis/theory. Studying this for some years now analyzing playcheck datalogs, is why I am quite consistent in my views. Most players in here know I have analyzed reels, randomness on how they land compared to the last wheel when a possible win could occure and such. (example, when the first 4 reels could give a good win if the 5th reel stops at the right place - is the last reel position weighted?) There are a million ways to give you bad luck, almost none detectable. However, being a gambling addict I have lost my head quite some times, many people wonder why I keep playing - so I wanted to add that.

Thanks for a great reply accually doing some study to finalize your thoughts.
 
As an experienced Software Engineer and online casino forum moderator I find myself wondering why you would post such a thing?

In particular I don't think this is appropriate material for the 'Online Casinos' forum whose stated purpose is:
Information, Experiences, questions and such. ....

Now you might say 'I was just asking a question' but I don't think it's as innocent and well-intentioned as that.

From what I know of you from your posting history here you are an intelligent person who has a bit of an axe to grind regarding online casinos. I can only assume that you are attempting to unsettle the readers here with what might best be described as a 'conspiracy theory'.

If I'm seeing this clearly and fairly I'd have to say that a 'conspiracy theory' intended to incite and unsettle the readership is more appropriate for 'The Attic' than anywhere else. Hence the move.
 
Discrediting my post by moving it into the jokes and crazy shit section only strengthens it.

Casinomeister is no longer a place for discussing rigged game topic from a programmers perspective - this was just proved.

Regards,
Kim Steinhaug
 
In particular I don't think this is appropriate material for the 'Online Casinos' forum whose stated purpose is:

That I would agree, however in past theese discussions have been popular. I myself is an experienced programmer, for almost 25 years. So yes, conspiracy or not I should have a firm basis of my views.

However, casinomeister is a forum for players which should embrase wins and loses and not the mechanics of a slot. I get it!

I wont be adding my thoughts more in your fora!
 
Discrediting my post by moving it into the jokes and crazy shit section only strengthens it.

Casinomeister is no longer a place for discussing rigged game topic from a programmers perspective - this was just proved.

Regards,
Kim Steinhaug

I think is always hard to prove negatives...

Prove casino do not cheat, online and land based.

I challenge your to prove any of the top software providers cheat.
.
Then, also define cheat. Might sound silly but i have heard many definitions of this this so I am curious.

PS, your code would not stand up to any serious testing I am afraid...
 
I wont be adding my thoughts more in your fora!

As in the past you are reacting with hostility and anger rather than a reasoned discussion.

If you think this belongs in the mainstream forums say so, justify it, explain why it's not just a conspiracy theory.

Simply saying "I'm pissed and I won't play in your sandbox anymore!" seems unnecessarily petulant.
 
@maxd

Let me rephrase, I won't be adding my grind on this fora! Obviously I agree that casinomeister has exhausted the talk about rigged system and is not the appropriate place of discussing this. I on the other hand have offered statistics, graphs, datamaterial - which is far more than other partisipants have done.

However, I would be much better off doing this with other developers and statistics experts another place - so I agree and respect your choice - however a little offended, it's only human.

So not being hostile, just ending myself taking part in talks about rigged software in this forum which I agree is not the correct place.

@casinojack

My point with the logic is a simplified overview. Anyway, I do not need to throw away my time explaining this for someone that is not willing to accually try to see my point of view. Either you get or you dont, or wont for that matter.

Some people obviously translate the example as a production example. It's like explaining a browser rendering engine in 20 lines of code, surely those 20 lines wont' work! However you get the general idea...0

My work in here covering this field is obviously done! People are obviously not looking into this at the depth I am doing, so you really have nothing to offer me! However I am tired of it myself, since proving or disproving anything wouldnt make a difference. People will gamble anyway, just a new software... Nothing changes - this is life! Hurray!

I was infact hoping for more intelligent replies on this topic. maby things have changed in here lately!
 
Let me add a few words to clarify where I'm coming from.

I could say "at Casino X every 999th blackjack player triggers 'take down' mode ... they clean him out".

Now on the surface of it this is entirely possible. The software could do it. The management could be complicit in it. Prove it doesn't happen!

However 'possible' it might be it is not a reasonable hypothesis, it's basically just something I pulled out of my arse to try and pin on Casino X.

To me there's not a lot of difference between my theory and yours, except that you've provided some code for show and tell.

So this is why I said it smells like 'conspiracy' to me and that's my reasoning for moving it. I could be completely wrong here, reading this totally the wrong way. I tried to explain why I was doing it, not trying to discredit you personally. Besides, 'The Attic' is the place for 'off topic' stuff not 'jokes and shit'.

And last but not least saying things like 'Casinomeister is no longer the place for discussing' and suchlike is not going to win you any friends or forward your topics here. Such things are meant to damage the forums because you are pissed off. They're not a legitimate contribution to what we and the membership do here.
 
So this is why I said it smells like 'conspiracy' to me and that's my reasoning for moving it. I could be completely wrong here, reading this totally the wrong way. Discuss.

Well everyone agrees that this is a conspiracy, since it's not proovable. If it were all hell would break lose...

On the other hand, what I find interesting here from a programming perspective would be this:

How hard would it be to create a system that one could enable a win switch or a losing switch, and still the outcome would be classified as random?

Am I the only one that finds this idea intreaging?
 
Let me add a few words to clarify where I'm coming from.

I could say "at Casino X every 999th blackjack player triggers 'take down' mode ... they clean him out".

Now on the surface of it this is entirely possible. The software could do it. The management could be complicit in it. Prove it doesn't happen!

However 'possible' it might be it is not a reasonable hypothesis, it's basically just something I pulled out of my arse to try and pin on Casino X.

To me there's not a lot of difference between my theory and yours, except that you've provided some code for show and tell.

So this is why I said it smells like 'conspiracy' to me and that's my reasoning for moving it. I could be completely wrong here, reading this totally the wrong way. Discuss.

So if I understand this correctly, there are no rig-threads allowed in the mainstream fora unless proven? Not even speculation?

There's definitely a line being drawn from my point of view.
 
Theory or not, I don't have a clue, it's way over my head, but the thread is interesting, so hang with it. :thumbsup: It's a good read.
 
PARS

Kimiss,

I don't know if you've seen a PARS ("paytable and reel strip") spread sheet. This is the mathematical model on which modern slots are built, in both land and i-casinos. These models are done in Excel, and the design and programming of the slot follows the specifications of the PARS.

If you haven't seen a PARS sheet, I suggest you contact one of the major slot manufacturers and see if you can get a legacy copy of one. This will give you a good foundation on which to proceed with your arguments.

If you've seen such a sheet, and still wish to pursue your arguments, then please produce complete code written in C++ that I can compile from the command line in Linux using the g++ compiler that you believe passes all the statistical tests conducted on such programs, but is not fair for some reason that only you know. I will run it through my suite of tests, and we'll see what we get. In particular, assume we have a slot with 48 reel stops per reel, 3 reels, and that the reels are equally weighted so that each stop occurs with equal probability, that is, each reel satisfies the most basic of statistical tests independent of the other reels, a chi-squared test on the distribution of each stop. Your program should produce a single line of data in CSV format: X,Y,Z per spin, with each of X,Y,Z between 0 and 47, representing the stop for that reel. I will run it and produce some number of millions of rounds to generate data to test, but I will not give you any more specifics in advance. However, the data is the only thing I will use to determine fairness, I will not inspect your code. I will provide my ftp site to you so you can upload your files. It can also be written in C or Java, but PHP won't do. No manufacturer that I know of programs their slots in PHP.

Best regards,

Eliot
 
Thanks, that is a good start, but it is a trivial example of a PARS sheet. That slot doesn't have free spins, bonus rounds, scatter symbols, or other things that players claim can be fixed. More typically, PARS sheets have several sheets in them with the calculations for the various features.

By the way, I am very good friends with the mathematician who did that sheet.

--Eliot
 

Users who are viewing this thread

Meister Ratings

Back
Top