Scary thoughts about Casinos

Sorry Nifty! I just forgot to add my special ironic guy to my last post:rolleyes:

Well, I would still encourage him to continue with or without that guy I suppose.

I have to add to this thread something else that I believe you guys are forgetting. We are so many reading this forum that not know English so well. It's going through hell trying to understand some of the posts. No googletranslation will help out either.
If I put it another way. Who are you/we trying to reach?
To explain all of the terms about RTP and RNG and everything else I understand that it's complicated to do with an easier form of the language, but please understand why some of us can't get a grip of what you are trying to say.
I'm not stupid, just Swedish..
...and blond :D

Posters like Busted Flush will use a few industry terms, combined with some fallacies, that may sound good, but are based on wrong or incomplete premises. I think his understanding of how RNGs work is flawed.
 
Posters like Busted Flush will use a few industry terms, combined with some fallacies, that may sound good, but are based on wrong or incomplete premises. I think his understanding of how RNGs work is flawed.

Since I and many others, english speakers or not, probably didn't understand what BustedFlush were saying at all, (I couldn't read half of his first post in this thread), I'm so happy that there are so many members here that really have true knowledge.
I have to trust that what you're saying is true, or I'm questioning how the whole gambling businis is set up. I would never gamble if I didn't believe it was fair and honest.
 
I've PM'd JStrike to ask for his opinion on what Bustedflush has stated. JStrike is always straight to the point and tells it like it is with no bullshit!

Figured I'd love to hear his opinion on the posts made by Bustedflush and get some counter arguments and valid explanations.


Cheers
Gremmy
 
Gremmy asked me to weigh in on this, so here goes. Usual disclaimer: I can only say for sure based on how my own software works. What I know about other companies' software is gathered from reading audit handbooks from different licensing jurisdictions, and the technical requirements for becoming a licensed game manufacturer in Nevada (which, btw, I'm not -- the cost of that license is in the mid-six-figures). What I do know for sure is that if you were going to rig software, the method described by BustedFlush would not only be illegal in most jurisdictions, it would also be a really long, roundabout and completely pointless way to get right back to the same drop/hold (and same RTP) the house gets anyway from completely random machines (or online games).

BustedFlush's theory of ladders is an interesting way of visualizing streaks of wins and losses, but it's actually just a way of trying to infuse personal meaning into randomness, and in the end that's just another twist on the gambler's fallacy. And it rests on a few false assumptions.

1. No casino I know of keeps a separate set of random numbers, or a separate RNG, for each player. It's not feasible. There's also no point.

2. There's a misunderstanding about RTP. RTP is just an average for the whole casino over time. It's not a guarantee that any one player is going to average that. Yes, the more spins you have, the closer you will get to approaching it. If you're the only player in the casino, and you spin 10 million times, then you'd better well be getting close to the advertised RTP. But it would be completely freakish if every player in a casino ended up at exactly the same RTP, even over 100,000 spins. When I simulate new games, I need to test at least a million spins to even get within 1% of knowing what a game's real RTP will be. And slots are not designed for the purpose of keeping each individual's payouts approaching the expected RTP. They don't need to be, because that's just a natural side effect of the fact that over time, they're going to average out to that for the casino as a whole. The casino doesn't care if it's one guy pulling the slot a million times, or a million players pulling it once. The casino doesn't care if you win the jackpot twice, and someone else wins zero times. And it doesn't care who you are, or who he is. It doesn't need to put you on a personalized program. All the house cares about is the percentage it keeps.

3. If a house wanted to limit its volatility - to prevent the freak occurrence of, say, 20 large jackpots paying out in a row - the correct way to do that is by leveraging progressive jackpots. Money sitting in progressives, to the house, is already a loss on the books, and therefore not a risk. If a house wanted to cheat to do it, by forcing machines to not pay jackpots for some period of time after the last one was paid, there would be no reason to do that on a player-by-player basis. That would actually be less effective than just stopping all jackpots for some length of time, and far more costly and difficult to do from a software perspective. Either way, with thousands of players, the chances of the last jackpot winner being the one to hit it next are very remote. But again, why would the casino care if you hit it twice? That's actually good PR.

4. Runs of good or bad luck don't need to be explained by personal predestined patterns. I see it all the time where one player on a slot is having an amazing run going up, while another one at the exact same time is going down. BustedFlush's example is about one person's variance, but the reality any floor manager can see is that the luck of players is just a reflection of RTP working itself out on a larger scale. As a basic example, on a slot with 95% RTP, two players can put in $50 at the same time, and one guy will often walk away with $95, and the other guy leaves empty. This isn't because the casino has rigged it so one's on a winning streak and the other one's on a losing streak, it's just a feature of randomness, plus the fact that some people walk away while they're up and others don't. In fact, the best way to see it from the casino's point of view is that we're just redistributing one player's money to another, and keeping 5% for ourselves. If everyone just ended up with 95% of what they started with, there'd be no reason for anyone to play. It's the variance between players that actually makes gambling interesting, not the variance for one player between spins. You think it's your own variance that keeps you interested, but your own chance of winning only exists because (1) you're not playing for an infinite length of time, and (2) other people are losing more than the house is keeping. Again, there's absolutely no reason to have pre-programmed sequences for all of a player's lifetime spins for that to happen, it just happens naturally.

One of the reasons I invented the game Mayan Gold was to disprove exactly this kind of theory. Mayan Gold is basically a 3-reel, 6-line slot, where the reels are turned sideways so you can see all the icons around them, and put on a table where they spin in opposite directions like a triple wheel of fortune. Ten people can sit around the table and share the same reels, and every symbol ends up on someone's pay lines. Some of the symbols are repeated. When you have 10 people sitting around the same table, they're all sharing their own part of the same three reels, and at least one player wins on almost any given spin. But even though the game has extremely high variance, it keeps a 96.6% RTP for the table over time, and that's completely achieved by choosing the payout amounts carefully, and ultimately by the percentage kept in the progressive jackpot at each table. 20% of each bet goes into the progressive pot. If we wanted to change the game's RTP, we'd only have to change what percentage goes into the progressives. It's proof of the principle that in a casino, one guy's loss is another guy's luck, and it's not personal or preprogrammed. The law of large numbers and randomness take care of everything.
 
New VWM:what:

Don't be in a hurry to throw out the old one just yet:D

BustedFlush has described one method of programming a compensated slot or "fruit machine", and is probably very close to the way the simple games like Bar-X were programmed. These simple or "Lo-Tech" fruit machine games tended to run on a long cycle, and this would repeat itself after a large number of spins (tens of thousands). When better processors became available, a more dynamic method of compensation became possible that relied less on cycles, and more on controlling future outcomes depending on whether the machine was paying over, or under, it's set RTP. It also allowed for a "streak pot" to be used that could be triggered by a predetermined set of criteria (and which could often be "forced out" by the experienced player;) ). The base game on such Fruities would pay a lower RTP, and put the surplus into the streak pot. "Forcing" meant deliberately refusing ALL "base" wins, thus forcing ALL this pent up RTP into the streak pot, thus bringing forward it's trigger point. This relied on one triggering parameter being the short term RTP of past play, and the belief that if this was forced too low, the program would compensate in the only manner available to it by triggering the payment of the streak pot. This worked very well on some games, but was quite a struggle to pull of on others.

The problem with the old Bar-X machines was that when new, the current RTP for past spins was zero, thus the first spins paid over 100% RTP in an effort to bring the compensation RTP parameter up from zero to a point somewhat above the set RTP, at which point the compensation code began to function normally. Running off the first 10,000 spins before installing the machine just meant that the compensator had "settled", and the machine would stick closely to the set RTP thereafter.

In most online games, there are none of these rungs, ladders, compensators, etc. Many don't even have weighting. This is because unlike physical slots, they can have as many symbols on a reel as needed to allow operation on simple random principles. Each reel would also run from it's own separate call to the RNG. Having each reel determined this way serves to counter the disadvantages of using pseudo-random numbers, which have the convenience of being cheaper to use, and can be part of the machine itself, even run as a separate task on the same processor chip.

The rules for the UK variant stipulate that a separate "unit" must generate the RNG, rather than it being done on the same processor that runs the game. These units were proposed to be "external to the machine", which could be interpreted as having to be elsewhere and connected by a cable, but this was impractical, so most manufacturers placed the RNG unit in the same cabinet, but used a separate processing "unit".

The external unit and cable could also be used as an exploit by disconnecting the machine from it's own RNG and adding a "gaffed" unit designed to trick the machine into paying out. Pulling this off would require some inside knowledge of what data the machine expected from it's own RNG, and how it would be interpreted. Knowing this meant you could send it data that it would interpret as a few big wins close together, which would empty the machine.

These theories gain credibilty because some CASINOS believe them, and we hear tales of players having winnings confiscated because a machine has "paid too much" to be the result of chance. The most recent case is that of Sky Vegas, who claim that the big wins from the new "Treasure Ireland" game were a "malfunction", rather than an unusual run of luck. The code for a simple random slot game just CANNOT "malfunction" like this unless something very obvious is wrong, such as a paytable error. For the fault to be more subtle, code that shouldn't even be there has to have malfunctioned.

All the doubters see Sky Vegas and others using this argument when they are faced with players winning too much, so why is it so far fetched an argument when a player has LOST far more than they believe is possible for a "fair" and "random" slot game.

For this to stay firmly with the "foil hat brigade", online casinos need to purge those member of the "foil hat brigade" that have somehow managed to get jobs with them.
 
Thanks jstrike:thumbsup: for explaining in a language that even I could understand;)
I wish as many as possible will read it...and understand what you are saying.

And a big thank you to you VWM for being who you are. Nope, I don't want to replace you:D
(but that please please please I asked you about still stands):)
 
VWM - that's an amazing technical explanation...

The base game on such Fruities would pay a lower RTP, and put the surplus into the streak pot.

I actually didn't know they did that, but it makes sense in principle. For people who don't get the huge significance of what VWM's saying here...when you're designing a slot machine, imagine you're starting with three reels and you're putting symbols on them, and there are more instances of some symbols than there are others. So for example, there are three cherries on the third reel, but only one bar. You can also weight the reels so they're more likely to land on a cherry, which multiplies that effect. But eventually you end up with some fixed number of cherries times their weight, and you have to make a paytable and figure out what that pays.

So you write a pay table. And you're aiming to make it so the payouts add up to the RTP you want on the game. You could set a 95% payback for any combination, but that's no fun. So what you do is start paying back 0% on the most likely combinations, and then paying back more than 100% on the least likely ones, until you find a balance where there's enough variability to keep the game interesting but not so much that players have to wait more than a few spins without getting paid.

But by convention, slot payouts are always done in multiples of the bet amount. So inevitably, your pay table is going to pay a little bit over or under the RTP you actually want. Let's say your chances of hitting two cherries should pay out 2.2:1. So you remove that extra .2 from the payout, and make it pay 2:1, and then you weight the extra onto the next payout up, say three cherries. You keep doing that all the way up the paytable until the end, when you're faced with something like 3 bars that pays out 2500.87 to 1. And you have that 87¢ left hanging over, and nowhere to go with it if you want to make a perfect RTP. One way of dealing with that is to round the top payout down to the nearest whole number, and put the leftovers percentage into a progressive pot. Another way, that VWM's alluding to, would be to put it into a secret pot inside the machine itself and pay it out in streaks.

Online casinos don't have to worry about it, because we can just start by designing a game around a 95% RTP, and then test it and see yeah, if we have to round three bells down to pay $100 instead of $100.14, then the RTP just went down to 94.876% or something, so that's the RTP we're going to announce. But a game that was required to have an exact whole-number RTP could never have all whole-number payouts, or vice versa. And so it could never rely completely on randomness...there would have to be an extra "kick" that paid players a little more once in awhile, to make up for the shortfall that was caused by rounding the payout amounts.

My ignorance, that VWM pointed out, was that I didn't know they actually did that in Vegas. But it does make sense if the law says you have to pay that percentage, no matter what. The only other way to do it, besides a progressive pot, is you'd have to have a reel on your slot that was almost infinitely huge. So tip of the hat to VWM. Still, none of that means they're tracking one player's behavior in any way.
 
VWM - that's an amazing technical explanation...



I actually didn't know they did that, but it makes sense in principle. For people who don't get the huge significance of what VWM's saying here...when you're designing a slot machine, imagine you're starting with three reels and you're putting symbols on them, and there are more instances of some symbols than there are others. So for example, there are three cherries on the third reel, but only one bar. You can also weight the reels so they're more likely to land on a cherry, which multiplies that effect. But eventually you end up with some fixed number of cherries times their weight, and you have to make a paytable and figure out what that pays.

So you write a pay table. And you're aiming to make it so the payouts add up to the RTP you want on the game. You could set a 95% payback for any combination, but that's no fun. So what you do is start paying back 0% on the most likely combinations, and then paying back more than 100% on the least likely ones, until you find a balance where there's enough variability to keep the game interesting but not so much that players have to wait more than a few spins without getting paid.

But by convention, slot payouts are always done in multiples of the bet amount. So inevitably, your pay table is going to pay a little bit over or under the RTP you actually want. Let's say your chances of hitting two cherries should pay out 2.2:1. So you remove that extra .2 from the payout, and make it pay 2:1, and then you weight the extra onto the next payout up, say three cherries. You keep doing that all the way up the paytable until the end, when you're faced with something like 3 bars that pays out 2500.87 to 1. And you have that 87¢ left hanging over, and nowhere to go with it if you want to make a perfect RTP. One way of dealing with that is to round the top payout down to the nearest whole number, and put the leftovers percentage into a progressive pot. Another way, that VWM's alluding to, would be to put it into a secret pot inside the machine itself and pay it out in streaks.

Online casinos don't have to worry about it, because we can just start by designing a game around a 95% RTP, and then test it and see yeah, if we have to round three bells down to pay $100 instead of $100.14, then the RTP just went down to 94.876% or something, so that's the RTP we're going to announce. But a game that was required to have an exact whole-number RTP could never have all whole-number payouts, or vice versa. And so it could never rely completely on randomness...there would have to be an extra "kick" that paid players a little more once in awhile, to make up for the shortfall that was caused by rounding the payout amounts.

My ignorance, that VWM pointed out, was that I didn't know they actually did that in Vegas. But it does make sense if the law says you have to pay that percentage, no matter what. The only other way to do it, besides a progressive pot, is you'd have to have a reel on your slot that was almost infinitely huge. So tip of the hat to VWM. Still, none of that means they're tracking one player's behavior in any way.

They don't. This is a peculiaralty of the UK "Fruit Machine", and was more common when the jackpot was limited to around £4.80 on a 20p stake. The "streak pot" was a way to make the game more interesting, yet stay just inside the law. The "streak" could pay over £20 in a single session of winning spins, but 20p was deducted as a fresh stake at every £4.80 interval.

Streaks are less common now that the top jackpot is £70 for stakes between 25p and £1. The £70 is only allowed to repeat ONCE to £140 under the new rules, and rarely happens in any case. Rather than a streak pot, players now get the £70 jackpot given to them, after which the machine usually goes pretty cold.

The video slots that pay up to £500 on stakes up to £2 have no need for a streak pot, as £500 is exciting enough on it's own, even for a £2 stake.

UK Fruit Machines are set from 72% to 78%, and the £500 jackpot games are set between 88% and 92%.
 
VWM - that's an amazing technical explanation...



I actually didn't know they did that, but it makes sense in principle. For people who don't get the huge significance of what VWM's saying here...when you're designing a slot machine, imagine you're starting with three reels and you're putting symbols on them, and there are more instances of some symbols than there are others. So for example, there are three cherries on the third reel, but only one bar. You can also weight the reels so they're more likely to land on a cherry, which multiplies that effect. But eventually you end up with some fixed number of cherries times their weight, and you have to make a paytable and figure out what that pays.

So you write a pay table. And you're aiming to make it so the payouts add up to the RTP you want on the game. You could set a 95% payback for any combination, but that's no fun. So what you do is start paying back 0% on the most likely combinations, and then paying back more than 100% on the least likely ones, until you find a balance where there's enough variability to keep the game interesting but not so much that players have to wait more than a few spins without getting paid.

But by convention, slot payouts are always done in multiples of the bet amount. So inevitably, your pay table is going to pay a little bit over or under the RTP you actually want. Let's say your chances of hitting two cherries should pay out 2.2:1. So you remove that extra .2 from the payout, and make it pay 2:1, and then you weight the extra onto the next payout up, say three cherries. You keep doing that all the way up the paytable until the end, when you're faced with something like 3 bars that pays out 2500.87 to 1. And you have that 87¢ left hanging over, and nowhere to go with it if you want to make a perfect RTP. One way of dealing with that is to round the top payout down to the nearest whole number, and put the leftovers percentage into a progressive pot. Another way, that VWM's alluding to, would be to put it into a secret pot inside the machine itself and pay it out in streaks.

Online casinos don't have to worry about it, because we can just start by designing a game around a 95% RTP, and then test it and see yeah, if we have to round three bells down to pay $100 instead of $100.14, then the RTP just went down to 94.876% or something, so that's the RTP we're going to announce. But a game that was required to have an exact whole-number RTP could never have all whole-number payouts, or vice versa. And so it could never rely completely on randomness...there would have to be an extra "kick" that paid players a little more once in awhile, to make up for the shortfall that was caused by rounding the payout amounts.

My ignorance, that VWM pointed out, was that I didn't know they actually did that in Vegas. But it does make sense if the law says you have to pay that percentage, no matter what. The only other way to do it, besides a progressive pot, is you'd have to have a reel on your slot that was almost infinitely huge. So tip of the hat to VWM. Still, none of that means they're tracking one player's behavior in any way.

What VWM said is standard knowledge for UK fruitie players (the serious ones anyway).

I don't think fruities were ever used in Las Vegas.....only SKY Vegas.
 
What VWM said is standard knowledge for UK fruitie players (the serious ones anyway).

I don't think fruities were ever used in Las Vegas.....only SKY Vegas.

If I read it right, it's different from parimutuel or Class II slots, or what they call "fruit machines" in the UK, though...what he's saying would apply to anything that has a pay-table a little lower than the reported RTP. I don't know if it's legal to re-invest that to the player in Vegas, but that was how I took his comments...

Edit: I just read VWM's post -- for some reason I missed it before. So now I understand this has to do with parimutuel slots, which is a whole different animal, never seen in Vegas. I still don't know whether what he's describing would be legal in vegas or not... but, the main point is that this is a contortion of the game to fit a certain set of laws, and there's no reason you'd build something like that unless it was to get around a government.
 
Thanks for responses and apologies for convoluted explanations, up in 6 hours for 4 day installation away from base, so will look to respond in turn to points in this thread hopefully Thursday evening or Friday.
 
I've been up almost 24 hrs now, so this may add to my confusion (probably not though)...what is BustedFlush talking about?:confused: I thought I was a semi-intelligent person...have understood VWM's responses, JStrike's responses, skiny's responses, Nifty's responses...going to go look for my "I'm stoopid cap" and my "I may be smarter tomorrow pills":eek2:
 
I've been up almost 24 hrs now, so this may add to my confusion (probably not though)...what is BustedFlush talking about?:confused: I thought I was a semi-intelligent person...have understood VWM's responses, JStrike's responses, skiny's responses, Nifty's responses...going to go look for my "I'm stoopid cap" and my "I may be smarter tomorrow pills":eek2:

If his wall of text doesn't seem to make sense it is because it doesn't. He uses a lot of words to say nothing that has any resemblance to how it actually works and is basically just a bunch of conspiracy theories presented as facts.
 
The difference will be when you overlay your peaks and troughs, over mine for the 100000 stakes, the pattern will be different. If you are fortunate enough to get your peaks early in the events after joining the casino for example are at 500% or 2000% you w/d and quit the casino otherwise attrition in the gameplay will ensure you end up at 97%and down overall.

If in any doubt, ask members on here to paste (when they've joined a casino and turned over hundreds of thousands of pounds) their aggregate gaming report and see the figures. If you think online slots are truly random and not compensated or contained then you are in fantasy land. Show me one genuine report where (discounting outside pooled jackpots/awards) of a substantial turnover for several hundred thousand spins/pounds where you have a figure of say 110%.........
I think you're getting arcade type Fruit Machines & Slots confused with Online Slots.

Of course you will gradually get closer to the published RTP the more you play - it's simple mathematics. This will still happen with a 100% completely random game.
Take Euro Roulette - a player could win big on his first few spins - or start with a horrible losing streak. But the longer he keeps playing the more it would even-out until he gets very close to a loss of 2.7027% of all the bets he placed.
Again, simple mathematics; there are 37 spots on the wheel and every win for a bet on a single number returns bet x36.

As has been said ad infinitum, EVERY casino game has a house edge which 100% guarantees the house WILL win in the long run.
They don't NEED to cheat - and they would be very foolish to risk it because if they got caught cheating it would be disastrous for them.

KK
 
I think you're getting arcade type Fruit Machines & Slots confused with Online Slots.

Of course you will gradually get closer to the published RTP the more you play - it's simple mathematics. This will still happen with a 100% completely random game.
Take Euro Roulette - a player could win big on his first few spins - or start with a horrible losing streak. But the longer he keeps playing the more it would even-out until he gets very close to a loss of 2.7027% of all the bets he placed.
Again, simple mathematics; there are 37 spots on the wheel and every win for a bet on a single number returns bet x36.

As has been said ad infinitum, EVERY casino game has a house edge which 100% guarantees the house WILL win in the long run.
They don't NEED to cheat - and they would be very foolish to risk it because if they got caught cheating it would be disastrous for them.

KK

Or the casinos could just cheat and win more in a shorter time.
 
Or the casinos could just cheat and win more in a shorter time.

They could, but it would be really stupid since they would go out of business if they got caught.
Most casinos will win in the long run anyway. One thing to be scared about is new casinos trying to make fast money. Always do your homework before you sign up to any casino.
 
Last edited:
Or the casinos could just cheat and win more in a shorter time.


Yes they could.

In fact, some dodgy operators have written software to do just that. However, none of the major suppliers have ever done so, as there are too many very smart people around who can spot that kind of nonsense from a mile away.

If you think this is what MGS and RTG and wagerworks etc do, then I assume you don't play online any more. If you do still play, then it doesn't sat much for your theory does it? How could you expect anyone to take you seriously when you don't believe it yourself? Just an example....not saying this is you..its just that I've seen so many people come here and accuse good operators of cheating only to come back later and tell us how they lost again.
 


In fact, some dodgy operators have written software to do just that. However, none of the major suppliers have ever done so,.

Nifty, I agree that all the major software suppliers are fair, and I share your loathing of conspiracy theories. However your statement above is wrong, and I urge you to check your facts before making such sweeping statements.

The history of partygaming shows rigged software sadly, from their wikipedia page
You do not have permission to view link Log in or register now.


"In the 1990s, Las Vegas consultant and actuary Michael Shackleford ran a computer trial of the first blackjack and roulette games offered by the company. Shackleford stated that the "results clearly showed they (the games) weren't fair". Ruth Parasol's spokesman Jon Mendelsohn acknowledged that the chances had "tipped too much toward the house", but attributed the problems to "software flaws", not rigging. It led to the development of their own proprietary software rather than using external platforms.[16]"

This partygaming rigged software story appeared on shacklefords blacklist page itself for a long time, before disappearing without explanation. Nonetheless they had live, rigged, software at the start. I have no doubt they are fair now tho, and have been for many years.
 
This partygaming rigged software story appeared on shacklefords blacklist page itself for a long time, before disappearing without explanation. Nonetheless they had live, rigged, software at the start. I have no doubt they are fair now tho, and have been for many years.
Party Casino did not (and still do not AFAIK) use any of the "Big Name" software brands that Nifty was talking about.
I believe (I know I will be corrected if wrong!) that it was/is their "Own Brand" made specifically for their own casinos.

KK
 
Party Casino did not (and still do not AFAIK) use any of the "Big Name" software brands that Nifty was talking about.
I believe (I know I will be corrected if wrong!) that it was/is their "Own Brand" made specifically for their own casinos.

KK

Yeah ok. I took PartyGamings casinos to fall under the "big brand" category.

I guess this is subjective and entirely up to each individual to decide what is a big brand in online gambling.
 
It is very hard to trust that online gaming is 100% random. Software created by man will have some sort of "randomness algorithm" built in - they are at best, pseudo-random. Even things such as PRBS (pseudo random bit sequence) is "pseudo".

As such, you can win if you "play the percentages". Patience is key.
 

Users who are viewing this thread

Meister Ratings

Back
Top