Video Lottery Terminals

[Linda_Boyd]

Harry Porter wrote:

Linda, could you address a couple of things and share what information
you have ...

Linda Boyd wrote:

Furthermore, the definition they use for RNG is different from the
one I use. They define a random number generator as a mechanism
inside the computer of Class III games that ensures that each pull
has an equal chance of hitting the jackpot

Harry Porter wrote:

This suggests a potential loophole in which generated hands aren't
fully random -- a mechanism that would produce a RF once in every
100,000 hands would see consistent twith that definition.

The RNGs found on the EPROM--Erasable, Programmable Read Only Memory--chip in commercial casinos are just like those found in Nevada casinos.

AGA's State of the States report makes it clear that VLT games are authorized through the state lottery and considered by law to be lotteries and not commercial gaming. Does this mean that the RNG isn't just as random when required on a VLT? No. It may be just as random, even just like the ones on video poker devices in commercial casinos. The problem is that the definitions are vague and may vary from state to state. The reason New York was able to have the type of device they chose for their racinos is the decision is up to the state. More states may decide to have devices like those in New York--it wouldn't be illegal.

Harry Porter wrote:

Assuming that it's known that a vp machine in play is a VLT (i.e.,
the dealing mechanism is centrally based), is there some means by
which to know with confidence that the machine is "fair" and operates
analogously to a standard LV-certified machine when it comes to the
randomness of the deal?

Linda Boyd wrote:

I don't think anybody is trying to cheat or be unfair--the problem is that sometimes officials just can't answer the question with any confidence. They know there's a RNG required on their VLTs, but can't define random number generator. If you read legislation from several states you'll see the definitions/requirements vary when it comes to VLTs.

Harry Porter wrote:

Bill has indicated that while DE/WV equipment may be VLT based,
it's reasonable to assume that the machines deal randomly. Do you
hsve information to support this?

Finally, is a player at any risk at a DE/WV in playing a non-random
game, or can they rely upon a paytable as providing their actual
playing odds.

Linda Boyd wrote:

I spoke to Jim Logue from the Delaware lottery office--he went out of his way to be helpful. Jim told me Spielo, IGT, Bally and Williams were all approved manufacturers for their VLTs. He also said that all of their VLTs had RNGs. The approved payback, however, is 87-95%--to me, that's honest, but not fair. I believe that you can determine the ER by using the pay table information on the VLT with tutorial software.

I spoke to several officials in West Virginia--Diedra, Wilma and finally Deputy Director Arnold. Once more, all officials working with their lottery office were polite and accommodating. I was told by Deputy Director Arnold that there was a RNG on all their VLTs, including those in their non-casinos. He said that bars, and restaurants could have up to 5 devices, while fraternal organizations were allowed up to10. Their machines are manufactured by IGT and Konami--their central control system was by GTECH. However, the payback is 80-95% for their VLTs. Personally, I find the payback to be unacceptable. Once more, I believe that the game's ER can be determined by using the pay table information with tutorial software.

My biggest frustration with VLTs is that there is no uniformity of definitions from state to state.

[Non-text portions of this message have been removed]

[Harry_Porter]

Harry Porter wrote:
> Assuming that it's known that a vp machine in play is a VLT (i.e.,
> the dealing mechanism is centrally based), is there some means by
> which to know with confidence that the machine is "fair" and
> operates analogously to a standard LV-certified machine when it
> comes to the randomness of the deal?

Linda Boyd replied:
I don't think anybody is trying to cheat or be unfair--the problem is
that sometimes officials just can't answer the question with any
confidence. They know there's a RNG required on their VLTs, but can't
define random number generator. If you read legislation from several
states you'll see the definitions/requirements vary when it comes to
VLTs.

My concern stems from a state regulation that may be poorly written.
Frankly, I could see where a state might permit hands to be dealt in a
fashion that would yield probabilities other than that of a random
deal (akin to a card deck). One can argue that a patron should have
no fixed idea of probabilities and expect the same degree of
incalculable odds of any other slot machine.

Setting that aside, if the state regulations are poorly defined in a
manner that would permit a non-random deal and there aren't sufficient
testing measures to ensure one, I have no doubt that a casino would
opt for equipment that would bias odds in their favor compared to the
paytable on the machines.

Linda Boyd wrote:
I spoke to Jim Logue from the Delaware lottery office--he went out of
his way to be helpful. Jim told me Spielo, IGT, Bally and Williams
were all approved manufacturers for their VLTs. He also said that all
of their VLTs had RNGs. The approved payback, however, is 87-95%--to
me, that's honest, but not fair. I believe that you can determine the
ER by using the pay table information on the VLT with tutorial
software.

If it's the case that the deal is random (i.e. not biased toward a
given outcome), then certainly you can determine the ER based upon
expected hand distribution. I consider that quite "fair", inasmuch as
a player can determine the odds of the game.

In this case, I wouldn't expect a "full pay" machine to appear on the
floor with very weak payout restrictions such as this.

I don't have first-hand experience with either DE or WV casinos (on
which I focus since a question concerning casinos in those
jurisdictions gave rise to this thread). I will be VERY interested to
see what shows it's head in PA when the newly approved "slot parlors"
being to open. If a "9/6" Jacks game is on the floor, I'll be dubious
of it's fairness. Even if played particularly poorly, it's difficult
to imagine that the game would be profitable given the harsh PA taxes
that have been imposed on casinos.

I'll note that in the case of Indian casinos that have IGT games, Bob
Dancer has been emphatic in the fairness of those games. I accept
that statement. However, I hold out reservations in general (no play
on words intended ;).

The often cited safeguard is that no game will be approved in NV
unless the same game placed in other jurisdictions meets NV
requirements. How much, or how little, would you need to change a
game appearing in another jurisdiction for it to be considered
different and not fall under the NV requirement? You'll forgive my
skepticism in suggesting that it's corporate nature to take advantage
of any profit opportunity presented, even if misleading.

I spoke to several officials in West Virginia--Diedra, Wilma and
finally Deputy Director Arnold. Once more, all officials working with
their lottery office were polite and accommodating. I was told by
Deputy Director Arnold that there was a RNG on all their VLTs,
including those in their non-casinos ... Their machines are
manufactured by IGT and Konami--their central control system was by
GTECH. However, the payback is 80-95% for their VLTs. Personally, I
find the payback to be unacceptable. Once more, I believe that the
game's ER can be determined by using the pay table information with
tutorial software.

I've discussed in a separate post this morning that there's little
question that there is a RNG installed in WA state VLT's. That, by
itself, doesn't assure a random deal and that's definitely not the
case there.

Depending on the nature of the IGT machines (in particular), I don't
have confidence that these are randomly dealt machines. For that
matter, IGT has gone into the central determinant VLT and Class II
bingo business, by virtue of their acquisition of SGT. They now have
developed Class II based machines based upon LV IGT standards ... and
in most respects you can't tell the difference from a cursory
examination (I don't know if the machines carry the IGT nameplate or
if they're marked as other than Class III).

Again, given my concerns, I wouldn't play in these states. Of course,
that may well be a moot point if the paytables are consistent with a
<95% game. You should find a 7/5 Jacks game, much less a 9/6 Jacks one.

Of course, my paranoia kicks in once again. I can't help to wonder if
a machine that has a stronger than 95% paytable, but that is expected
to hold more than 5%, would be permitted under a poorly written
statute ;).

Tell you what, I'll try to restrain myself and give this topic a rest.
You've thoroughly responded to my questions and I suspect no one will
complete assuage my concerns. (You'll hear back from me if 99%+ ER
games started popping up in PA

- Harry

[Bill_Coleman1]

Harry,

Let me try a few more quick responses:

My concern stems from a state regulation that may be poorly written.
Frankly, I could see where a state might permit hands to be dealt in a
fashion that would yield probabilities other than that of a random
deal (akin to a card deck). One can argue that a patron should have
no fixed idea of probabilities and expect the same degree of
incalculable odds of any other slot machine.

Setting that aside, if the state regulations are poorly defined in a
manner that would permit a non-random deal and there aren't sufficient
testing measures to ensure one, I have no doubt that a casino would
opt for equipment that would bias odds in their favor compared to the
paytable on the machines.

Regardless of how the regulations are defined other than scratch-card
or Class II machines they all deal like Nevada. Remember, NJ has no
specific regulation but will not approve a machine that does not
behave this way.

I'll note that in the case of Indian casinos that have IGT games, Bob
Dancer has been emphatic in the fairness of those games. I accept
that statement. However, I hold out reservations in general (no play
on words intended ;).

The often cited safeguard is that no game will be approved in NV
unless the same game placed in other jurisdictions meets NV
requirements. How much, or how little, would you need to change a
game appearing in another jurisdiction for it to be considered
different and not fall under the NV requirement? You'll forgive my
skepticism in suggesting that it's corporate nature to take advantage
of any profit opportunity presented, even if misleading.

I've addressed this before, but briefly -- the oft-cited safeguard
doesn't actually exist. However, NV requires machines shipped to
other jurisdictions meet the requirements of THAT jurisdiction. Since
no jurisdiction permits "gaffed" machines you can rely on that protection.

Companies that only ship to specific jurisdictions (they are not
Nevada-based) will also guarantee that their machines meet all
regulatory requirements.

I've discussed in a separate post this morning that there's little

question that there is a RNG installed in WA state VLT's. That, by
itself, doesn't assure a random deal and that's definitely not the
case there.

WA machines do not have a game-controlling RNG. The central server
sends them a scratch-off ticket result and that controls game play.
This includes IGT machines (hence the Genie in the VP games to ensure
the "correct" result). It is still random and fair according to the
rules in WA.

Depending on the nature of the IGT machines (in particular), I don't
have confidence that these are randomly dealt machines. For that
matter, IGT has gone into the central determinant VLT and Class II
bingo business, by virtue of their acquisition of SGT. They now have
developed Class II based machines based upon LV IGT standards ... and
in most respects you can't tell the difference from a cursory
examination (I don't know if the machines carry the IGT nameplate or
if they're marked as other than Class III).

All IGT class II machines display the Bingo card (as of last Sept)
and in WA and NY Racinos you can look at the help screen. When it
speaks about the Genie you know what's happening. A cursory
examination won't help, but you should never perform a cursory
examination when in a new jurisdiction.

[Non-text portions of this message have been removed]

···

At 04:31 AM 5/18/2006, you wrote: