AxiomOfChoice
AxiomOfChoice
  • Threads: 32
  • Posts: 5761
Joined: Sep 12, 2012
March 12th, 2014 at 5:47:57 PM permalink
I was listening to some old GWAE shows on youtube, and in one of the shows that had the Wizard as a guest, near the end of the show, Munchkin asked about figuring out returns on slot machines.

The Wizard mentioned (almost in passing) that for 5-reel video slots (unlike 3-reel physical slots) each possible reel stop was equally weighted, so, if you could map an entire reel, you would be done.

This surprised me. I'm wondering, is that a regulation, or just the standard way that it's done? Or, did I misunderstand the statement? Wizard? Anyone else?
tringlomane
tringlomane
  • Threads: 8
  • Posts: 6281
Joined: Aug 25, 2012
March 12th, 2014 at 5:53:05 PM permalink
Quote: AxiomOfChoice


This surprised me. I'm wondering, is that a regulation, or just the standard way that it's done? Or, did I misunderstand the statement? Wizard? Anyone else?



Standard way. With video slots, there is no physical limitations to the reel strip size. 256 stops with a symbol on each stop, no problem for a video slot.
AxiomOfChoice
AxiomOfChoice
  • Threads: 32
  • Posts: 5761
Joined: Sep 12, 2012
March 12th, 2014 at 6:11:06 PM permalink
Quote: tringlomane

Standard way. With video slots, there is no physical limitations to the reel strip size. 256 stops with a symbol on each stop, no problem for a video slot.



Interesting. 2 more questions:

1. Is it a requirement that the symbols that you see going by are actually a part of the reel (and are possible stops)?

2. How do they change the payout of the machine if they can't re-weight the reels? It seems like there are 3 options:

a. Change the reels by making them longer or shorter.
b. Change the reels by replacing some sections of symbols with others.
c. Change the value of the free spins / bonus game, and leave the value of the base game alone.
CrystalMath
CrystalMath
  • Threads: 8
  • Posts: 1911
Joined: May 10, 2011
March 12th, 2014 at 6:17:30 PM permalink
Quote: AxiomOfChoice


The Wizard mentioned (almost in passing) that for 5-reel video slots (unlike 3-reel physical slots) each possible reel stop was equally weighted, so, if you could map an entire reel, you would be done.



There is no regulation, but every video game I've designed is like this, as well as most that I tested at GLI. The exception was for video games made to look like physical reel slots.
I heart Crystal Math.
CrystalMath
CrystalMath
  • Threads: 8
  • Posts: 1911
Joined: May 10, 2011
March 12th, 2014 at 6:20:48 PM permalink
Quote: AxiomOfChoice


1. Is it a requirement that the symbols that you see going by are actually a part of the reel (and are possible stops)?


I believe that GLI does enforce this. Also, a reel animation cannot give a false representation of the frequency of symbols, such as displaying extra wild symbols in the animation.

Quote:


2. How do they change the payout of the machine if they can't re-weight the reels? It seems like there are 3 options:

a. Change the reels by making them longer or shorter.
b. Change the reels by replacing some sections of symbols with others.
c. Change the value of the free spins / bonus game, and leave the value of the base game alone.



a. yes
b. yes
c. yes
I heart Crystal Math.
AxiomOfChoice
AxiomOfChoice
  • Threads: 32
  • Posts: 5761
Joined: Sep 12, 2012
March 12th, 2014 at 6:30:05 PM permalink
Quote: CrystalMath

I believe that GLI does enforce this. Also, a reel animation cannot give a false representation of the frequency of symbols, such as displaying extra wild symbols in the animation.



Thanks. Are all slot machines in Nevada casinos compliant with GLI standards? (I am reading through them now)
CrystalMath
CrystalMath
  • Threads: 8
  • Posts: 1911
Joined: May 10, 2011
March 12th, 2014 at 8:26:19 PM permalink
Quote: AxiomOfChoice

Thanks. Are all slot machines in Nevada casinos compliant with GLI standards? (I am reading through them now)


Probably. GLI standards are very similar to Nevada. You can also find the NGCB rules online.

Edit: you're welcome :)
I heart Crystal Math.
AxiomOfChoice
AxiomOfChoice
  • Threads: 32
  • Posts: 5761
Joined: Sep 12, 2012
March 14th, 2014 at 7:05:44 PM permalink
Ok, so, this is where I am going with all this...

I like quick hits, a lot. Small jackpots that can get really high because short-coiners add to the progressive but can't win it. And the smaller ones don't take long to hit. At the right casino at the right time (when all the broke people are gambling for minimum bets) they can get big.

So, Mission posted this link a while ago. With this information combined with the assumption that all reel stops are equally likely, it should not be too hard to deconstruct this game, which is great -- I would know exactly when a particular machine or bank of machines had gone +EV, if I could work out its base return.

The problem is, something is not right. If each individual reel has all stops equally weighted, then, clearly, the frequency of a particular result must divide the product of the number stops on all the reels. Therefore any prime factor of any frequency must divide the number of stop on at least one of the reels. However, some of these frequencies have extremely large prime factors (in the thousands).

There is no indication that these numbers are rounded, however (my understanding is that they were copied from information that was accidentally posted on Bally's web site, briefly). Do video slot games really have reels that long (several thousand stops?). Or can we conclude that, if this information is accurate, the 5-reel video QHP games do not have equally weighted reels after all?
tringlomane
tringlomane
  • Threads: 8
  • Posts: 6281
Joined: Aug 25, 2012
March 14th, 2014 at 7:41:22 PM permalink
Quote: AxiomOfChoice

Ok, so, this is where I am going with all this...

I like quick hits, a lot. Small jackpots that can get really high because short-coiners add to the progressive but can't win it. And the smaller ones don't take long to hit. At the right casino at the right time (when all the broke people are gambling for minimum bets) they can get big.

So, Mission posted this link a while ago. With this information combined with the assumption that all reel stops are equally likely, it should not be too hard to deconstruct this game, which is great -- I would know exactly when a particular machine or bank of machines had gone +EV, if I could work out its base return.

The problem is, something is not right. If each individual reel has all stops equally weighted, then, clearly, the frequency of a particular result must divide the product of the number stops on all the reels. Therefore any prime factor of any frequency must divide the number of stop on at least one of the reels. However, some of these frequencies have extremely large prime factors (in the thousands).

There is no indication that these numbers are rounded, however (my understanding is that they were copied from information that was accidentally posted on Bally's web site, briefly). Do video slot games really have reels that long (several thousand stops?). Or can we conclude that, if this information is accurate, the 5-reel video QHP games do not have equally weighted reels after all?



Something I left out of my initial post out of laziness.

Video slots are allowed to have sets of stops that repeat.

For example, on a sample Quick Hits reel:
If stops 25, 26, and 27 have 7, 1-Bar, and Blank, respectively, and stops 44, 45, and 46 also has 7, 1-Bar, and Blank, respectively. Then this particular result becomes twice as likely to appear than a unique set of 3 stops. Finding these duplicates is a pain.
AxiomOfChoice
AxiomOfChoice
  • Threads: 32
  • Posts: 5761
Joined: Sep 12, 2012
March 14th, 2014 at 8:00:00 PM permalink
Quote: tringlomane

Something I left out of my initial post out of laziness.

Video slots are allowed to have sets of stops that repeat.

For example, on a sample Quick Hits reel:
If stops 25, 26, and 27 have 7, 1-Bar, and Blank, respectively, and stops 44, 45, and 46 also has 7, 1-Bar, and Blank, respectively. Then this particular result becomes twice as likely to appear than a unique set of 3 stops. Finding these duplicates is a pain.



Sure, I assumed that. That doesn't really address my concern about the frequencies having large prime factors though. I don't understand how any 5-reel video slot in which all stops are equally weighted could have any result that appears exactly 1 time in 87,617. 87,617 = 41 * 2137 -- does one of the reels really have some multiple of 2137 stops? (2137 is prime) How long are these virtual reels?
tringlomane
tringlomane
  • Threads: 8
  • Posts: 6281
Joined: Aug 25, 2012
March 14th, 2014 at 9:12:20 PM permalink
Quote: AxiomOfChoice

Sure, I assumed that. That doesn't really address my concern about the frequencies having large prime factors though. I don't understand how any 5-reel video slot in which all stops are equally weighted could have any result that appears exactly 1 time in 87,617. 87,617 = 41 * 2137 -- does one of the reels really have some multiple of 2137 stops? (2137 is prime) How long are these virtual reels?



Oh, I assume that's rounding.

I would be surprised if QH had more than 256 stops per reel. Most older slots have well under 100 per reel, or at least the ones I have deconstructed for practice from PC software...yes, I have no life...lol

But games with things like tiled wilds can easily have >100 stops per reel.

Nice post at exactly 8pm PDT, btw.
Mission146
Mission146
  • Threads: 142
  • Posts: 16832
Joined: May 15, 2012
March 15th, 2014 at 10:00:15 AM permalink
The game would be virtually impossible to deconstruct pursuant to the fact that Free Games operates on a different reel assignment than the main game.
https://wizardofvegas.com/forum/off-topic/gripes/11182-pet-peeves/120/#post815219
  • Jump to: