What Exactly Does Clearing The RAM DO, And Does It Affect The RNG?

[reho33]

someone needs to contact an actual gaming commission slot lab because I think they have already done the work on this stuff.

[Neonkiss]

Well. I don't buy the reel motor worn theory you used, because if it is a RNG it doesn't matter how long it takes to spin or stop since the number has already been picked at the time of pressing the button.

BUT THATS NOT YOUR QUESTION YOU ASKED IF THEY WOULD STOP AT SAME TIME

No, My questions was in reference to two completely Ram cleared machines that have power applied at exactly the same time. (thru a common power strip) Will the machines be exactly the same after the first spin?
If the RAM clear, establishes a zero point of the machine, (as it's been suggested) Won't the RNG come up with the same reel stops on both machines.

[stayouttadabunker]

someone needs to contact an actual gaming commission slot lab because I think they have already done the work on this stuff.

Chances are...they won't spill the beans...

[ROCKET]

Well. I don't buy the reel motor worn theory you used, because if it is a RNG it doesn't matter how long it takes to spin or stop since the number has already been picked at the time of pressing the button.

BUT THATS NOT YOUR QUESTION YOU ASKED IF THEY WOULD STOP AT SAME TIME

No, My questions was in reference to two completely Ram cleared machines that have power applied at exactly the same time. (thru a common power strip) Will the machines be exactly the same after the first spin?
If the RAM clear, establishes a zero point of the machine, (as it's been suggested) Won't the RNG come up with the same reel stops on both machines.

ok you clarified the question etc so i understand what your saying and asking neonkiss ..

LETS SAY HYPOTHETICALLY THE   RNG ON BOTH MACHINES DOES START OFF AT ZERO --FOR HYPOTHETICAL PURPOSES HERE ..

THE ODDS WOULD STILL BE IN THE BILLIONS THAT ONE RANDOM NUMMBER GENERATOR IS GOING TO FOLLOW THE SAME NUMBER PATH OF THE OTHER RANDOM NUMBER GENERATOR ..

IF THAT WAS POSSIBLE TO PREDICT ?? YOU COULD PREDICT MANY OTHER THINGS IN LIFE ?? LIKE 2012 ?? LOL

[PWRSTROKE]

I am going to try the rng t'mrow at the casino's.  I have not been in a while now.  I will try and find some worn out ones as they seem to pay the best-- .  Happy New years to all while it is still "New Year" Day---

[5 ACES]

Okay!! Okay!! Can I at least get some K+'s for starting this thread!! LOL!! Seriously though, I often wonder how "Random" the RNG really is. I know just from my experiences that while playing our S+, we have hit the 2nd award jackpot twice in one day, four times in one week. Now, if you look at the PAR sheet for that particular game, the odds were roughly 1 in 200,000 for hitting that prize. Now, if you really believe the odds on that, you feel pretty damn good after hitting it that many times, in that short amount of time. Personally, I have often wondered if you have let's say a 90% chip in it, then how "Random" can it really be, if it's going to hold 10% of the total money put into that machine, over the 10,000,000 pulls? In my eyes, there is nothing random about that. That machine is supposed to hold 10% for the casino, so if it's random, then how in the hell can they predict that? The RNG then must be programmed to work out to keep 10% for the casino. That has always bothered me and my faith in how random the RNG really is, has always been questioned. I too have always felt like there is a starting point to the RNG and it goes from there. Just my feeling though. I'm sure many a slot players, can attest to the fact that these machines seem to pay like mad out of nowhere and then take your money just as fast, as though something starts the winning mode and then something ends it as well.

P.S. The Subject now includes the RNG!! LOL!! So we don't feel guilty about talking about it!! LOL!!

[Thor777]

Okay!! Okay!! Can I at least get some K+'s for starting this thread!!
P.S. The Subject now includes the RNG!! LOL!! So we don't feel guilty about talking about it!! LOL!!

K+ to you for starting this...     

[stayouttadabunker]

I would like to go out on the limb and see what happens with this statement.
It's just for fun and to make one think!>>>

Take for example, you have a 90% chip installed in an S+.
This is the statement...rip it apart to shreds if you'd like...
But I'll guarantee you - It will make you think

[ "The RNG is programmed to read the reel chip paytable for 90% of the winning combinations" ]

True or False?
ok...now, after you've selected an answer, ask yourself - is it truly "random"?

[brichter]

IIRC, the RNG runs the same no matter what the payout percentage is, the payout table and virtual stops are manipulated to create the proper percentage based on random spins from what I've heard. I'm pretty sure StatFreak can set me straight on that point...

BTW, I don't believe the RNG program is in the eprom, I'm pretty sure it's implemented at a lower level. The eproms hold the game characteristics and the payout tables, and the rest is on the processor board, along with the sound chip, etc.

  Note also that software implementations are pseudo random number generators, which start with a seed or key, so they're not truly random. That's the downfall of a software RNG program, it holds true for both slot machines and crypotgraphy applications (which is where I learned the theory behind it  ). The way they increase the randomness is by increasing the length of the number, because f the number is shorter, it MUST repeat sooner than a longer string would. If the string is only 1 bit (0 or 1) then it will repeat a lot sooner than a string that is, say, 128 bits.

Maybe we can get Stolistic in here to drop the hammer on the S+ implementation, I'm sure he knows more than I do about that specific application.

[brichter]

I would like to go out on the limb and see what happens with this statement.
It's just for fun and to make one think!>>>

Take for example, you have a 90% chip installed in an S+.
This is the statement...rip it apart to shreds if you'd like...
But I'll guarantee you - It will make you think

[ "The RNG is programmed to read the reel chip paytable for 90% of the winning combinations" ]

True or False?
ok...now, after you've selected an answer, ask yourself - is it truly "random"?

See my post, I say it's false. However, according to what I know about software RNGs, they can't be truly random because of the "seed" which is the starting point. I'd be interested to know the string length the RNG in an S+ uses.

[jay]

A RNG works from a base seed. When I went to school we did similar tests where we had a series of XT PCS and used the same seed value to step through the ring and each of the computers generated the exact same numbers over and over again. To get a divergence from the results we used the current TIME as our base seed. We ran through about 20,000 numbers before we started to get divergence and the explanation was that the clock frequency between the PCs were flawed.

To get a truly random number we needed to introduce a human element to the equation. In our case we still used time as our seed however the number was only picked when we hit the enter key. So in practice the time seed was different in each case as no two people were able to hit the enter key at the exact same 100'th of a second.

It would be my belief that if we had two slots and had control over the hardware we could generate the exact same random numbers for some period of time until we got divergance based on the clock frequency. I do not believe that anyone without this control could ever generate the same set of results. Certainly no one would be able to do something as simple as a Ram Clear, power on and spin to generate the identical number set.

[brichter]

Your understanding of the RNG is the same as mine, if you can control the time the output will be identical until a difference between the platforms is evident.

Now, where's Jim to answer my question about the string length?

<edit>


Had to edit my other posts, spelling was horrendous!

[theDotster]

It would be my belief that if we had two slots and had control over the hardware we could generate the exact same random numbers for some period of time until we got divergance based on the clock frequency.

To the next extreme, would the length of any cabling between the circuits/buttons have any effect on this, i.e. amount of time taken for electricity to travel along cabling if you were running the same machines?

[brichter]

In a copper conductor, the signal is approximately .95C. So, the cabling length difference would have to be huge, because the time division in an internal clock on the slot machine just isn't that granular.

[ROCKET]

 

THIS ONE REALLY GOT SOME POST LAST NIGHT

BEWTWEEN;
 ROCKET
 BRITCHER
 NEONKISS
 5ACES
 BUNKER
ETC ETC WE WERE ON FIRE LOL

somewhere i have the complete technical explanation of the RNG in true scientific reading .. its lengthy to say the least .
it still boils down to all the questions posted --- a ram clear only clears the memory of said slot for the purpose of solving error codes that wont disapear or to game change at times that wont except a different game and most important to just clean the memory out 100% . thats it

NOW RNG -- the random number generator runs as soon as power is given to the machines main board .. it starts there counting counting etc .. in the perfect world talk .. ya no B/S LOL

it counts numbers combinations etc when you hit the buttonto play the electrical slotmachine !! it locks in the numbers it has picked which produces the outcome of the reel postions
on the given game in the slot cabinet !!.. it does this and the percentage is based on the life of the slot !!now going from the memory of the tech paper i have in the file somewhere ?? i have no clue ??lol the life of machine is based on a time of millions of spins of the game . to give what we know as par sheets to say ok this one does this this one does that etc etc .

as far as ram clearing affecting the random number generator it does not !!

I never said it did i dont think any previous post member did !!

WE DID OBSERVE AFTER A FULL RAM CLEAR & INSTALATION OF THE NEW GAME& REEL CHIP GLASS STRIPS ETC .. THAT THE RESPONSE TIME OF THE SLOT IN QUESTION !!
DRASTICALLY SPEED UP IN ALL WAYS !!

 meaning if you read the early part of this thread my statement i think was >>ALL ITEMS << speed up as in reel speed start/ stop credits in credits out --this was noted by me on the post of having wittnessed this myself on a 5-line double diamond with 88% payout chip on nickels for I know over 5 yrs that had enough nickel grindings on the floor of slot to fill a 2-oz. container .after i ram cleared that slot it was like it turned from turtle to a rabbit with speed of operation .

so to finalize this for a bit

WHAT DOES A RAM CLEAR DO ?
THE RAM CLEAR IS TO CLEAR MEMORY OF THE MAIN & MOTHER BOARD ONLY !


DOES THE RAM CLEAR AFFECT THE RNG REEL CHIP ?? ANSWER COULD ONLY BE NO !!your not clearing the random number generator -but you are giving it a clean plate to go from !!!


LAST -DOES A RAM CLEAR SPEED UP A MACHINE THAT HAS NOT BEEN RAM CLEARED??>>>>> ANSWER IS NO !!<<<<<<

IT DOES HOWEVER PUT THE MACHINE BACK TO ITS ORIGINAL SPEED WHICH WAS FASTER THAN IT WAS BEFORE THE RAM CLEAR !!

THIS MAY CLEAR UP ALL THE TALK OF RAM CLEARING MAKING MACHINE RUN FASTER ETC --

if i find the file of the scientific explanation of the RNG i will post it . but it is lengthy !!
i hope there are no hard feelings to any of the members that i banntered with back and fourth last night  this is how we all learn tossing out plausible ideas ??
and some non plausible ideas ?? and some ideas just off the wall 

rocket

[cfh]

Wow i read this whole thread and i have to absolutely agree with
Stat Freak. You guys just don't understand how the S+ board
system is designed or the components that it uses.

First off, STOP comparing the EEPROM or the RAM chip to Cache
in your computer!  That is just, well, silly. It's not the same.
It's apples and oranges.

The cache in your computer is not a pre-determined size like
an EEPROM ST24C04. This chip is very small in capacity. It "fills up"
the first time it's written to. Only 2 blocks of 256 x8 bits of data.
That's less than 1k of data!

Look at the datasheet. It explains it A LOT better than i can!
http://www.datasheetcatalog.org/datasheets/134/80442_DS.pdf

The other RAM (volatile RAM, powered by the battery) is
also frankly quite small. That's a 6264 RAM, 8192 words by 8 bits.
That's not a big RAM either, about 8k of data. That's K not Meg!
Data sheet here:
http://web.mit.edu/6.115/www/datasheets/6264.pdf

These RAMs don't "fill up". THEY ARE ALWAYS FULL!  From
the first time the game is power up to the last time, there
is EXACTLY the same amount of data in these chips.
The same memory locations as addressed by the hardware
are just overwritten.

Stat Freak was exactly correct in his explanation.

Memory cache in your computer's browser is basically unlimited,
only dictated by the size of your hard drive. Well that's a slight
exaggeration, but it's close to the mark.

This thing with 'performance' being better after a RAM clear,
that just doesn't make a lot of sense.

If you really want to prove your point, i guess you would need
to set up benchmarks and measurable things before and after
a clear. That's the only way i'm gonna believe the goofy
position that these games' memory "fills up".

[ROCKET]

wel im sorry CFH
I RESPECTFULLY DISAGREE WITH YOU &  Stat again for the question that was asked on >>THE ORIGINAL POST <<

the original post was >>WHAT DOES A RAM CLEAR DO <<<<

after Statfreaks lengthy explanation & comment its used way to often i posted to him I respectfully disagree !!

perhaps your missing the ORIGINAL QUESTION CFH ??

WHAT DOES A RAM CLEAR DO ?
THE RAM CLEAR IS TO CLEAR MEMORY OF THE MAIN & MOTHER BOARD ONLY !  THATS THE PURPOSE OF THAT CHIP MADE BY IGT IN THIS CASE !!

DOES THE RAM CLEAR AFFECT THE RNG OR REEL CHIP ?? ANSWER COULD ONLY BE NO !!>>>>>>>your not clearing the random number generator >>>>

but you are giving it a clean plate to go from !!!

NOW I HAVE SERVICED IGT S+ MACHINES FOR 15+ YRS MYSELF
DOES A RAM CLEAR SPEED UP A MACHINE THAT HAS NOT BEEN RAM CLEARED??>>>>> ANSWER IS NO !!<<<<<<

IT DOES HOWEVER PUT THE MACHINE BACK TO ITS ORIGINAL SPEED WHICH WAS FASTER THAN IT WAS BEFORE THE RAM CLEAR !!

THIS MAY CLEAR UP ALL THE TALK OF RAM CLEARING MAKING MACHINE RUN FASTER ETC --

as far as your saying the main board or mother board of a S+ not getting clogged up with info ??  I hate to say it does !!

if the main board & mother board never got clogged with stored information -->> the RAM CLEAR CHIP WOULD NEVER OF BEEN INVENTED !!

I am sorry ..i stick by my findings & my observations from reading most every manual on the s+ and the best training possible HANDS ON TRAINING 15+ YRS ON THOSE ALONE

[stayouttadabunker]

It seems that no one actually knows where the RNG is actually located on the MPU board anyways.
I never thought it was in the reel chip...
I only stated that the RNG read data "FROM" the reel chips' paytable arrangement.
Where the actual RNG program is located is anyones guess...
it could be located in the AYxxx sound chip for all I know...LOL
I'm suspecting that it's located in microprocessor chip though.

Okay , off the RNG for a minute and back to the original question...LOL

I always thought the Clear chip re-routed electricity to the two memory chips
( the motherboards' 24C04 and MPU's CMOS ) via certain socket legs
to energize all the internal chips' gates to a "1"?

[StatFreak]

A RNG works from a base seed. When I went to school we did similar tests where we had a series of XT PCS and used the same seed value to step through the ring and each of the computers generated the exact same numbers over and over again. To get a divergence from the results we used the current TIME as our base seed. We ran through about 20,000 numbers before we started to get divergence and the explanation was that the clock frequency between the PCs were flawed.

To get a truly random number we needed to introduce a human element to the equation. In our case we still used time as our seed however the number was only picked when we hit the enter key. So in practice the time seed was different in each case as no two people were able to hit the enter key at the exact same 100'th of a second.

It would be my belief that if we had two slots and had control over the hardware we could generate the exact same random numbers for some period of time until we got divergance based on the clock frequency. I do not believe that anyone without this control could ever generate the same set of results. Certainly no one would be able to do something as simple as a Ram Clear, power on and spin to generate the identical number set.

 

As Jay said, if two identical RNGs are initially seeded with the same vale, they will run the same series of outcomes, but as Jay also said, even if one used computer precision to check the current values at specific intervals, the two systems would slowly drift out of sync due to the slight differences in the clock speed of the computers. These differences are caused by slight physical differences in the clock crystals, in the manufacture of the hardware and components, and operating temperature differences.

The most important statement that Jay made is the addition of the HUMAN FACTOR. In gaming design, it is called Player Initiated Action (PIA). PIA is what ultimately makes slot machines truly random: no one can time their actions down to milliseconds. PIA would cause the two hypothetical machines' RNGs to become significantly divergent after only a few plays.

Wow i read this whole thread and i have to absolutely agree with
Stat Freak. You guys just don't understand how the S+ board
system is designed or the components that it uses.

First off, STOP comparing the EEPROM or the RAM chip to Cache
in your computer!  That is just, well, silly. It's not the same.
It's apples and oranges.

The cache in your computer is not a pre-determined size like
an EEPROM ST24C04. This chip is very small in capacity. It "fills up"
the first time it's written to. Only 2 blocks of 256 x8 bits of data.
That's less than 1k of data!

Look at the datasheet. It explains it A LOT better than i can!
http://www.datasheetcatalog.org/datasheets/134/80442_DS.pdf

The other RAM (volatile RAM, powered by the battery) is
also frankly quite small. That's a 6264 RAM, 8192 words by 8 bits.
That's not a big RAM either, about 8k of data. That's K not Meg!
Data sheet here:
http://web.mit.edu/6.115/www/datasheets/6264.pdf

These RAMs don't "fill up". THEY ARE ALWAY FULL!  From
the first time the game is power up to the last time, there
is EXACTLY the same amount of data in these chips.
The same memory locations as addressed by the hardware
are just overwritten.

Stat Freak was exactly correct in his explaination.

Memory cache in your computer's browser is basically unlimited,
only dictated by the size of your hard drive. Well that's a slight
exaggeration, but it's close to the mark.

This thing with 'performance' being better after a RAM clear,
that just doesn't make a lot of sense.

If you really want to prove you point, i guess you would need
to set up benchmarks and measurable things before and after
a clear. That's the only way i'm gonna believe the goofy
position that these games' memory "fills up".

 

Now that nails it on the head cfh.

If you want to satisfy me that there is a difference in performance, you will need to gather empirical data, not inferential observation. Set up independent optical sensors at the outermost diameter of each of the reel baskets and report their RPMs to a PC which records them.

Use an automatic play system (Bunker's car alarm autoplay, for example) to consistently spin the machines as quickly as the slot makes a new spin available and use another separate monitor to time the delay between the end of the previous spin and the beginning of the next spin. Connect this monitor to your PC for historical data recording. To do this will require you to have an electronic monitor on the MPU board to determine when the machine finishes its spin cycle (pun intended) since some spins will result in winning payouts. The delay between spins must not include payout time. Keep in mind that the system is designed to have inconsistent spin times (from start of spin until the last reel stops).

Use an electronic timer to determine how fast credits are being added to the meter (credits per second) during winning payouts and tie that to the PC for historical data recording.

Record several thousand spins worth of data before the clear, and then again after the clear. Compare the results and report back.


Regarding the assertion made by many techs that these machines are more likely to hit a top prize after a clear, or within some specified time frame following a clear (I have heard stories ranging from 24 hours to 168 hours), there is a possibility  that the RNG might  be more likely to choose certain number combinations early in its cycle following a reseed.

However, consider that every game theme (and even some different chips within a single game) puts the jackpot symbols at different virtual stops. In plain English, the specific set of three numbers required from the RNG to get a jackpot on one game are completely different from the three numbers needed to get a jackpot on another game.

Let me repeat that for those in the nosebleed section: The specific set of three numbers required from the RNG to get a jackpot on one game are completely different from the three numbers needed to get a jackpot on another game.

Explain to my satisfaction how the RNG will produce more jackpots on various games immediately following a clear, given the above statement.

To prove your point, set up the following experiment:
Get 100 S pluses that have not been cleared and have at least 1,000,000 games of "memory" and install the same game and reel chips on all of them.
Set up automatic play and record the outcomes of 150,000 successive pulls* on each of these machines.
Clear all 100 machines.
Set up automatic play and record the outcomes of the first 150,000 successive pulls* on each of these machines following the clear.
Demonstrate that the payouts on the newly cleared machines are statistically higher (>1 SD) than the pre-clear payouts and
Demonstrate that the payouts on the newly cleared machines are outside the expected 90% confidence rate return for the reel chip used (Okay, for you purists, calculate and use a 95% confidence value, which is an established standard.)

*150,000 based on 900/hour at 168 hours.

[ROCKET]

thats a nice reply stat ,
but to answer the original question that started this madness was addressed by me !!
the

THE RAM CLEAR CHIP --SIMPLY CLEARS THE MEMORY STORED ON THE MAIN BOARD & MOTHER BOARD -NO LENGTHY EXPLANATION NEEDED !!

as far as the statement you said its used to often I STILL DISAGREE WITH YOU only from a tech hands on working on the S+ FOR 15 + YRS

when it went to the>>> RNG does a ram clear GIVE THE RANDON NUMBER GENERATOR A NEW START POINT ?

I answered that before Jay last night in my post along with today . THAT NO TWO RNG'S  are going to track the same number sequence regardless of the human factor !!if you retrace my post .

as far as the speed of machine increasing after a ram clear -- I WENT THE EXTRA TYPE ..posted answer is no !! BUT IT WILL RETURN THE MACHINE TO OPTIMOM SPEED --THAT WAS SLOWER BEFORE THE RAM CLEAR .

as far as the main board or mother board not getting full of info or clogged .. thats totally rediculous not to think it does not happen !! with coin in coin out bill in re-sets power offs power on etc it does dont act like it does not .. yes i concede its not as like the home pc cache being cleared that was mentioned for understanding purposes . but its a clear of information on a smaller basis non the less ..

i hope this was not as lengthy and easy to understand as my brother members post above !

[stayouttadabunker]

I'd like to perform the 100 machine test...
but I'd need a government grant to get that many machines...

[ROCKET]

I'd like to perform the 100 machine test...
but I'd need a government grant to get that many machines...

STOP OVER BUNKER  ,
I WILL LET YOU SET UP 100 IGT S+ MACHINES

you break any glass you become a DEXTER VICTIM LOL

I will sit in my directors chair & even do a live video feed to NLG as you mess up my game room or burn my house down lol

DEAL OR NO DEAL ??

[stayouttadabunker]

NO DEAL!
That's a major undertaking...LOL
I'm NOT getting into that!
I'll be constructing "Auto-Play" relay switches until the next millennium!
I have enough snuff on my plate!

But I'll still take on chance on some venison stew? LOL

[uniman]

This is a great post guys!  
You need Myth Busters help.

Listen to Stat, he has it right.
And let me add one more thing. In my Uni programs, and I would think the same for IGT, you have a virtual stop table for each reel. The RNG is really selecting three address locations. (sort of like throwing a dart at a map or grid) One in each table. What is at that address is the stop location on the reel. Only after the program reads that selected address does it determine what symbol it has chosen.    
Most Uni games have the jackpot at the same address, hmmm, but I see IGT has jackpot symbols at various locations. Therefore, they would be at various addresses in the virtual stop table for each game

I can't think of a reason why a machine would run "crisper", "faster", etc, after a ram clear.  

And older Uni's use a chip N82S123AN to help generate a psuedo-random number. - I believe

What you need to do is have four machines. Clear the ram on machines #1 & #2. Do not clear the ram on #3 & #4. Then tell four players that machines #2 and #4 have had their ram cleared. Have the four players play each machine and record the outcomes and the players opinions of how each game played.
I would wager they would select machines #2 and #4 as playing better, faster.

Again, great post!

[stayouttadabunker]

Nice post UNI!
However, there's only question burning in my mind as I read and re-read these posts.
Where's the RNG program located at?
Could it be a mini Vanna White in a chip device turning 3 tiles?