Microgaming slots - are the odds adjustable?

[lojo]

It turns out they might be

 

[Simmo!]

It turns out they might be

Be the world's best kept secret for 11 years if they do

So who made you think there was lojo? There are forever people trying to start rumours about things like this just to try and damage a competitor's rep. I mean you believe who you want to believe I guess, but an MG guy and a top operator told me it can't be done, and naive or otherwise, I'm inclined to believe them myself.

Not only would it have come out in the last 11 years if you could, I'm sure, but also in an area where we're talking about public listed companies and transparency, a factor such as this that would have a major impact on profitability would be in the public domain I think.

 

[aka23]

ALL of them!
I don't know how many there are - but I would guess in the region of 23-30...?
Someone will need a lot of spare time to review all those!

I'd guess more than that. In alphabetical order some are:

3Dice, Absolute, AST, B3W, Betsoft, Boss Media, BTD, Cantor, Casinova, Chartwell, COA, Cryptologic, DGS, ETSI, Futurebet, G-fed, Gamesys, Global Player, Globotech, Grand Virtual, iGaming, iGlobalMedia, iXent, Kismet, Live (various SWs), Mansion, Microgaming (viper, classic, and poker/casino), Net Ent, Odds On, OGS, Ongame, Parlay, Player Preferred, Playtech, Progressive Gaming, Random Logic, Real Time Gaming, Rival, Slotland, St. Miniver, VueTec, Wager21, Wagerworks, Wintech, World Gaming, Whatever SW US players are directed to at some of the Playtechs (forget the name)

That's about 50, and I know there are many more. I consider Microgaming Viper far above the rest, although I agree that others are catching up.

 

[lojo]

Be the world's best kept secret for 11 years if they do

So who made you think there was lojo? There are forever people trying to start rumours about things like this just to try and damage a competitor's rep. I mean you believe who you want to believe I guess, but an MG guy and a top operator told me it can't be done, and naive or otherwise, I'm inclined to believe them myself.

Not only would it have come out in the last 11 years if you could, I'm sure, but also in an area where we're talking about public listed companies and transparency, a factor such as this that would have a major impact on profitability would be in the public domain I think.

Simmo, I have total respect for webzcas and others who have worked for them, he wouldn't lie. Yesterday I had total respect for the gurus at casinocitytimes, but I found a piece of Nevada Gaming Commision fodder that told me that the paytables are adjustable between coins in a video slot. That is adjustable and not fair in my book. I've run tens of thousands of dollars through MG slots and I've yet to see an actual spokesperson from MG come on here, with all the questions better men than me have asked, and tell us it ain't so.
If Microgaming holds itself to a higher standard than the NGC holds their subjects, I would like to hear them say so. Until then, I won't spend another penny there. I'll take anecdotal evidence if it's from the horses mouth but if they won't come on and proclaim, I'll stand by "It turns out they might be."

 

[Simmo!]

We could probably work this out ourselves with enough people targetting one or two specific slots at different denominations for a set amount of spins. Whether we could get a big enough sample to be definitive would be questionable but if say 20 players did 1,000 spins each at a set denomination, it may give a rough indication of a major discepancy?

A low/medium varaince slot like Avalon where you can bet in $1 increments and still get some decent size wins might be a good target. You could run an initial test in Fun mode as I'm lead to believe that the same RNG is used for Fun and Real. Although that in itself is another whole debate ( that we've had before )

 

[lojo]

That's a beautiful and elegant solution, Simmo, but it won't cut it for me.

When I walk into a B&M the first thing i do is ask for the manager, if he's busy I go to the tables and ask for a pitboss, if he's busy I'll entertain a host for a while, but I get my answer and I get my treatment.

If these putz can't come to the finest site online and answer a simple question, duff em I say.

Respect, Simmo, I totally respect you. You never have and never will lie to me.

 

[AudiManinBoro]

I believe that MG might very well be launchinh a brand new type of lobby in the near future. I have received emails from the Spin Palace group offering me 100 if i take part in trials for new lobby implementation. Other people might have received the same. Alas for me i didnt have the time to take advantage but it could have been interesting..

 

[footdr]

I PERSONALLY DISAGREE ON SAME OUTCOME REGARDLESS OF BET SIZE

I do not believe that the you get the same outcome regardless of the bet size. My personal experience over the last 4 years playing MG software prevents me from believing that.

I have no hard proof other than playing 20 spins at max bet and never hitting say, free spins or bonus and then finally reducing the bet and quickly hitting the bonus.

It happens to to often in my opinion. I figure that there is a different payout odds on each coin size and number or coins played. i.e. coin size 10 cents number of coins l-5.

I would need HARD proof for me to believe otherwise.

Also, if each casino has their own server does that mean they have their own random number generator and if so, how does that affect Progressive Jackpots? I would like to know if it matters where you play a Progressive. Are the odds of hitting it equal no matter which casino you play at. Seems some casinos constantly have winners and others hardly ever do. WHY?, if it is all totally random? and the spins are all being controlled by the same RNG.

 

[chuchu59]

Footdr,

I am not in disagreement with you but it can be argued that when you have spun a number of spins without any bonus rounds then the free spins will come in sooner or later. Have you tried the reverse ie betting low first and increasing your bets afterwards. If you get roughly the same results then it would seem that bet size does not matter. Otherwise, hmmm.

 

[footdr]

I WILL GIVE THAT A TRY BUT........

I will give that a try, but my gut feeling is that there is really no way to prove or disprove this because obviously there is no way to know if I had not changed to lower bet if I would have got the same result. Also, this doesn't just happen when lowering a bet or change to lower coin denomination. It can happen by increasing the denomination or bet amount.

I think of it this way, I never say to myself "geez I shouldn't have lowered my bet, I could have won X"; because I believe that I wouldn't have got the same result if I hadn't changed the bet amount. At lease thinking this way I never get upset!!!!

 

[bridgehajen]

Yes, MG slots are the same everywhere - making it a very strong brand.

Agreed

I have said too many times now that it is a shame this does not apply to RTG casinos. It is up to each casino manager to configure the slots and they could be set to a ridiculous low payout% and you will never now what happened to you money before it is too late. And worse of all, whenever I have a bad session I will always suspect I got robbed (in terms of low payout%). While at MG I know it was a fair game.

Well, yes ... I suppose that for "regular" slots, which have weighted reels anyway, it's possible to adjust the weights without the player being able to tell. I'm sure that you are right about RTG doing this. Technically, as long as they don't advertise a specific payout, they are cheating no one.

However, video slots do not have weighted reels, or at least that is my assumption. If that is true, I can see only two ways of adjusting the payouts:
1) Change the actual reels, which is not a trivial thing. And detectable by players, given enough time and patience.
2) Apply weights to the reels. IMHO this is clearly overstpping the line. It's cheating in much the same way as if "weights" were applied to the cards at the blackjack table, making the player more likely to draw a bust card. Also, this method would be detectable, given enough spins.
So are RTG operators able to cheat the players in either of these ways? Or in ways that I didn't consider? It's an interesting discussion ..

 

[steinhaug]

As Ed Ware from 32Red has once said before... and I believe it is true! MG software does not make any distinctions or alterations in its outcome depending on the bet size, nor does it check to see if you're playing for .1 cent/spin or $10,000.00/spin. The outcome is the same! .... and I do not believe that individual casinos can vary the odds as far as MG is concerned.

Yeah right! Sure they will say this, however a hot machine on $1 isn't hot on $2 and will be hot again going back for $1 - experienced that many times. If this was true, you could "manipulate" the payout ods with doing all the features on high wagering and all the others on low - making the machine payout far above 100%.

Sure the casinoes say this, it's the correct thing to say! Just as casino managers say that 1 coinsize and 5 coinsize (eg. 20lines x 1 versus 20 lines x 5) has same variance.... Lower wagering gives lots of more low prizes, making your worth while... Higher wagering gives fewer small wins and more big wins.

 

[AudiManinBoro]

It does appear sometimes that what your success rate changes when ever a bet value is changed. Im sure weve all been playing on slots at some point where theyve been doing very well and you risk upping your best stake and than it goes all pete tong lol I know I have but it could be coincidence

 

[steinhaug]

The perception would be that all slots are totally random (however, Vegas has had several incidents where slots are busted for having programmed loosing streaks, and only winning the usual payout when tested). Online systems have also been busted, software aswell.

However that aside, the mechanics of slots are still the mechanics given we talk about a truly RNG - let's assume no faul play is in hand for the case of the discussion.

As mentioned, the perception is that the mahine is totally random and the outcome is random whatever coinsize you are betting. This is totally false.

Casino slots use different variance on different coinsize, this in turn means that doing 150 spins on lowest bet then increasing the bet to get the feature on maximum wager simply doesnt work. All random systems need some sort of seed to create the random, and given the same seed you get the same predicted random (unless knowing the seed it will still be totally random). Most likely the different coinsizes have different seeds, which in turn will result in a random machine - however changing coinsizes is like switching machines.

After playing slots for 15+ years I would say that I have the experience to know what I am talking about, whatever a casino representative (which is the one earning my money) would say to hypen up the traditional perception.

An easy experiment can be done:

Go to a landbased casino (where I have the most experience, and where the overall payout usually is a bit lower than online slots).

Start off with $1.000 and wager $1 at a time, you will most likely play all night, not win anything big, but most likely go home with the same $1.000 you started with.

Do the same, but start with $10.000 and wager $10 at a time, you will most likely go bust in 1-2 hours. (here and there you could get lucky and do a $30.000 win, sure, but they are faaaar apart).

This falls neatly together with alot of the slot sites I've been reading up on, and with a company I know who created slots some years ago. Variance differs from coinsize.

The discussion however, is not of this, but if the random sequence is influenced by other things (like deposits, if you won on the other game and such). The overall payout is surely 95%, no problem with that - but depending on how you distribute the wins 95% can force you broke in a snap!

 

[lojo]

Yeah right! Sure they will say this, however a hot machine on $1 isn't hot on $2 and will be hot again going back for $1 - experienced that many times. If this was true, you could "manipulate" the payout ods with doing all the features on high wagering and all the others on low - making the machine payout far above 100%.

Sure the casinoes say this, it's the correct thing to say! Just as casino managers say that 1 coinsize and 5 coinsize (eg. 20lines x 1 versus 20 lines x 5) has same variance.... Lower wagering gives lots of more low prizes, making your worth while... Higher wagering gives fewer small wins and more big wins.

I've decided to assume that coin size does not alter the paytable/program on MG machines, as all we have to go on is their word.
I've also decided to assume that it does in Vegas, because the nevada gaming commision allows it...

 

[steinhaug]

I've decided to assume that coin size does not alter the paytable/program on MG machines, as all we have to go on is their word.
I've also decided to assume that it does in Vegas, because the nevada gaming commision allows it...

Nice of you to aknowledge the fact that coinsize infact does influence the variance. However, we all agree that the typical feature appears around every 130 spins in example, Thunderstruck. If what you say that MG is true, this means doing 130 spins without feature on 0.01 wager, then moving for say $9 wager in the long run would probably ruin the casino, ending up with far above 100% payout. If we assume that the random variance gives 95% in the long run, skipping 130 spins here and there with much lower payout would skew the overall payout - dont you agree?

Maybe Zoozie could jump in to verify this.

If you look at the stats for a particular game, you will quickly see that a big payout rarely is followed by another big payout. Do 10.000 spins on whatever game to verify this. This would in the long run mean do a few 100 spins on low wager, play high til you get a big win, then go down again. If the system is truly random, this doesnt work. Do you agree? The system has to have different variance on the different coinsizes.

 

[lojo]

Hopefully Zoozie can bring more knowledge and clarity to this, but for now...

The 1/130 is a probability, not a 'possibility'. Even if it averages out to 1/130ish they aren't grouped in exact batches of 130's, and you don't know where in the probability 'cycle' you are starting.
........x..............
x......................
.......................x
x..x...................
etc.

I had always assumed that the paytable does not change for coin size in Vegas, because the gurus at casinocity told me it doesn't. However a poster here claimed it could, so I checked the NVC rules, and yippers.

MG asserts (by proxy) that they don't. And indeed it isn't necessary over the long run.
just my 2 bits

edit: oops missed the 'do you agree' part, will get back to it later, gotta run.

 

[steinhaug]

I owuld love some feedback from some of the other statistics people.

My point is this. I know 130 is only probability, but that is not the point. Let me rephrase myself.

We agree that by spinning the wheels for 1.000.000 times, or 10.000.000 doesnt matter, at $90 wagering - the outcome should be 95%.

So, if we - in the long run - manage to squeeze out 50 spins after a major win (say 100x+ multiplier) without any deasent win, and with far below 50% payout, the $90 wagering would pay far above 100% making the casino go bust.

Why, because the random system needs all the non paying spins to end up width the 95% payout. If this isn't true, the machine would need a calculator which kicks in and really distorts the payout - and then we are really fubar.

Sure, sometimes two larger wins, or three, even 4 could happend 4 spins in a row - but in the long run - it's pretty easy to atleast squeeze out 50 spins on 1 cent, and then jump up to $90, or $200 or whatever it should be. (And surely, should we miss a large win, we sould do the extra spins to compensate this win). Point is, system has to have different variance, or atleast a payout stabilator (which is very bad).

This doesnt only sound logical to me, it would have to be a minimum if the casinoes would want the machines in the first place. If the system would work the way people say it works, it sounds really strange that noone has figured them out.

 

[spearmaster]

The first problem is that I do not believe there are any slot machines in Vegas that have variable betting units. Only some video poker machines have this. This is a feature which is unique to online casinos.

Secondly, the particular payout for a machine must be set in the presence of a Gaming Control representative - no changes can be made without a representative present.

And thirdly, machines in Vegas must be random.

This is *not* the case in Europe, however. Your experience there cannot be compared to Las Vegas or any gaming jurisdiction in North America.

Anyhow, what you are talking about with regards to variable payback is interesting and may be relevant in many places - just not the US.

Here is a bit on Nevada technical standards for slot machines - read Standard 2, the first clause says that payback percentage may not be capable of being changed without having to change the hardware *or* the software.

Old / Expired Link

 

[Zoozie]

Nice of you to aknowledge the fact that coinsize infact does influence the variance. However, we all agree that the typical feature appears around every 130 spins in example, Thunderstruck. If what you say that MG is true, this means doing 130 spins without feature on 0.01 wager, then moving for say $9 wager in the long run would probably ruin the casino, ending up with far above 100% payout. If we assume that the random variance gives 95% in the long run, skipping 130 spins here and there with much lower payout would skew the overall payout - dont you agree?

Maybe Zoozie could jump in to verify this.

If you look at the stats for a particular game, you will quickly see that a big payout rarely is followed by another big payout. Do 10.000 spins on whatever game to verify this. This would in the long run mean do a few 100 spins on low wager, play high til you get a big win, then go down again. If the system is truly random, this doesnt work. Do you agree? The system has to have different variance on the different coinsizes.

This is a betting system and no betting system ever beat the house edge for any game. Besides getting the feature will only pay 35*betsize(or is it 32?) in average! (this is for Thunderstruck). And trying this aggressive increasing betting system is the worst you can do really. High variance + gigantic house egde. If you want to risk a fortune for a small win very often, it is much better to do in blackjack and even though it is STILL a bad idea. (ask Cypher...)

 

[steinhaug]

Im not sure if you are getting my drift here. The point of the machine is a payback of 95% - this we agree on - and the playcheck logs clearly shows this being the case - no argument here.

Let's say the machine has 2 betsizes only, $0.9 and $90. Lets say the variance was extremely predictable (for this example only). Say the win cycle was:

0 0 0 x 0 0 0 x 0 0 0 x 0 0 x

My point is, aslong as I wager $90 all the time, the payout would be 95%. If I were to, after every win, take the next spin on $0.9 and then back to $90 - I would "cheat" the payout for $89.10 in the 95% payout scheme after every win. This would in the long run give my a much higher payout - since the setup is "rewuired" to get all theese looser spins aswell.

The machine needs you to loose alot of $90 spins for the 95% to become a 95%.

This is what I mean by "squeezing" alot of spins out on $0.01 and then bump back to $90.

Zoozie, this is not a playing style of me - I am not crazy! It's an example on why - unless my above theory is totally wrong - MG machines must have differenet random variance on every bet size. Atleast I can't get my head around it some other way.

Mind you aswell - if what I say is true - this is not cheating in any way. This system is not fraud or false or anything regarding the RNG, this only determines that there is one RNG sequence for every coinzise - which ensures that the casino in the long run always will have it's edge (mathematical that is, in one game it could start with a jackpot ofcourse).

If they didnt, since all players are moving up and down in betsizez, the payout couldn't be 95% but probably 250% or 30% depending on what prizes people got or didnt when jerking up the coinsize...

- - -
Edit: somy typoes

 

[aka23]

Let's say the machine has 2 betsizes only, $0.9 and $90. Lets say the variance was extremely predictable (for this example only). Say the win cycle was:
...

If the "variance was extremely predictable", then it is not a random game.

If the game is random, then you have the same chance of getting a bonus round on any spin... the same chance if you have gone 1000 spins without a bonus round or if you got one on the last spin. Random games don't attempt to make up for losses or wins to achieve the expected payout. The actual payout might be 100% one month and 90% another.

 

[lojo]

The first problem is that I do not believe there are any slot machines in Vegas that have variable betting units. ...

I'm not sure I understand... variable betting units. Almost all slots have multidenominational input - coin size. I may have misunderstood.

Here is a bit on Nevada technical standards for slot machines - read Standard 2, the first clause says that payback percentage may not be capable of being changed without having to change the hardware *or* the software.

Old / Expired Link

Thanks I'll read that.



Here's something else, I think it says they can change the paytable per coin size, but I could be daft and i would love to be corrected if I'm reading it wrong.

100. For licensees that have not installed an On-Line Slot Metering System approved by the Board pursuant to
Regulation 14 Technical Standard 3, when multi-game or multi-game/multi-denomination machines are initially
placed on the casino floor and when the active paytables within the slot machine are changed, the theoretical hold
percentage used in the slot analysis report is a simple average of the theoretical holds, as set by the manufacturer,
of all the active paytables of the slot machine. The slot analysis report is revised to indicate the new simple
average theoretical hold percentage whenever a change is made to the active paytables within the slot machine.
Note 1: For multi-game and multi-game/multi-denomination machines, a new machine number is not assigned
when paytables are changed within the same library of paytables.
Note 2: The theoretical hold percentage needs to be obtained for each active paytable when multi-game/multidenominational
machines have different paytables for each denomination within a game that are activated for
play.
101. For licensees that have installed an On-Line Slot Metering System approved by the Board pursuant to
Regulation 14 Technical Standard 3, that is connected and communicating with the slot machines to read and
record the coin-in amount by paytable or by wager type of the slot machine, the system is utilized to complete the
following procedures that applies only to multi-game and multi-denomination/multi-game machines and for slot
machines which have a difference in theoretical hold percentage which exceeds 4 percent for a single-coin play
versus maximum-bet play:

 

[lojo]

I owuld love some feedback from some of the other statistics people.

My point is this. I know 130 is only probability, but that is not the point. Let me rephrase myself.

We agree that by spinning the wheels for 1.000.000 times, or 10.000.000 doesnt matter, at $90 wagering - the outcome should be 95%.

So, if we - in the long run - manage to squeeze out 50 spins after a major win (say 100x+ multiplier) without any deasent win, and with far below 50% payout, the $90 wagering would pay far above 100% making the casino go bust.

Why, because the random system needs all the non paying spins to end up width the 95% payout. If this isn't true, the machine would need a calculator which kicks in and really distorts the payout - and then we are really fubar.

Sure, sometimes two larger wins, or three, even 4 could happend 4 spins in a row - but in the long run - it's pretty easy to atleast squeeze out 50 spins on 1 cent, and then jump up to $90, or $200 or whatever it should be. (And surely, should we miss a large win, we sould do the extra spins to compensate this win). Point is, system has to have different variance, or atleast a payout stabilator (which is very bad).

This doesnt only sound logical to me, it would have to be a minimum if the casinoes would want the machines in the first place. If the system would work the way people say it works, it sounds really strange that noone has figured them out.

You might not like this idea and I'm sure the statisticians won't because they can't factor chaos.
First we have to assume that the paytable does not change per bet size, that the program and the rng are not querying each other. Without that premise it is not a truley random machine! If they are asking each other questions we have to have, say 4% varience here for vegas slots under a certain system. (At least it may be a know commodity but any variance is acceptable if it is known) I'd rather work from the premise that it is a true random system - the rng and program do not communicate based on coin size.

Simple, you must introduce random increases and decreases in bet size.
If you come onto a machine and hit the 'feature' within x (say 35) spins you should move to another machine,because though the odds are the same, the probability that you will hit the feature again is lacking.

Before doing this you have to manage your wagers for the possibility of not hitting a feature in 130 spins. I won't bother to do the math, i play by gut, but a 1k bankroll shouldn't be betting any more than 75 per and probably less.

My ol - lady is bitchin to me to get to work so have to leave it at that for now.

Good luck

 

[vinylweatherman]

With these schemes, every time a player gets one over on the house by hitting the big features and wins on big bets, there will be one who gets the opposite, with big wins on small bets given back with bad spells when betting large.

For there to be potential for any system to work, there would need to be a degree of predictability with future outcomes.
It has often been argued that MG slots seem to have long spells that favour high payouts, and long spells where the payout is very low. These spells seem to run for many thousands of spins.
A case in point are the Casino Action tournaments. I have played these recently by making very many spins at very small bets (I can do this with ease since January on my new "science machine", which can easily run 4 or 5 MG casinos simultaneously). Last Thursday was a case in point (Thunderstruck).

Music Hall 100 starting balance and spinning at 0.09p per spin - busted out and added a further 38 in comps. Busted out again. 138 gone for score of 1967 (1967 wagered, approx 21,856 spins, Payout 93%)

Challenge 300 start, and added around 20 comps. Busted out after 5862 wagered at 0.27p per spin. 21,711 spins, payout 94.5% - close to expectations.

Golden Reef 150 start and set to 0.18p per spin.
CASHED OUT 200! AND scored 3454 over 19,188 spins payout 101.5%

Also, these were all playing SIMULTANEOUSLY, same game, same software & RNG setup, only different casinos. It seems that Golden Reef consistently got more than it's fair share, while Music Hall missed out. Challenge held pretty much to expectations.

20,000 is still a small amount compared to all the possible outcomes on Thunderstruck, but doing this week in and week out has been building "intuitive" data, and this is telling me that, in general, low stakes pay a little better than high ones.

One striking feature of this experiment, run since Febrary/March, was the sequence of 5 scatter hits that I posted, and which got a couple of members here thinking these was "something going on", as they were hitting at more than 5x expectations.
The balance has been redressed a little, as I did not hit the 5 Rams either this last Thursday, nor the one before.


One other slot worth looking at is Cabin Fever. I have noticed that it has spells where the sun on the first reel hits pretty much every 2 to 4 spins, but at other times it remains absent for many spins at a time. This seems to have an effect on the number of bonus rounds granted, as well as the likelihood of multiple retriggers.
It is possible to do an experiment based on the sequential outcomes of this first reel alone, thus reducing the problem to one of under 40 possible outcomes, from one of several million.
To analyse the bonus round chances fully, only the first and last reels need to be looked out, making the full analysis still manageable, perhaps one of 40 squared possible outcomes - this would analyse the bonus round itself, but not the slot payout. It should be possible to work out if the occurence of the bonus round is truly random, or whether it is "streaked" artificially through variable weightings (my suspicion about some aspects of MG in general).

If it turns out there is some kind of "streak" weighting (probably not what will be found in the end, but.....), then MG casinos better run for cover