29 JUNE 2004
NEW ZEALAND GAZETTE, No. 79
1943

a) the same sequence of random numbers is never used in more than one device at the
same time (i.e. there is to be a method whereby each gaming machine can have a unique
seed generation technique or RNG startup values).

b) the “next” game outcome is not able to be predicted.

1.27.9 Seeding and re-seeding must be kept to an absolute minimum. Both the method of re-seeding
and the instances when it may occur must be approved. Re-seeding should not in general be
under operator control. Re-seeding should not be a routine or regular practice.

1.27.10 If for any reason the background RNG activity is interrupted (e.g. gaming machine power
down), the next input variable(s) for the RNG must be a function of the value(s) produced by the
RNG immediately prior to interruption.

Minimum Period for Prize Determination using One RNG Value.

1.27.11 Following a low probability game outcome (e.g. a jackpot win, major prize win, or a particular
graphic game result presentation), where that game outcome is represented by only one RNG
value or a small number of RNG values, it is important that subsequent game play on that
machine is unpredictable. The intent here is to ensure that the machine does not subsequently
go through one defined sequence of game outcomes, or one of only a few possible sequences
of game outcomes, and that game outcomes are consistent with the mathematical model of the
game. Thus in such implementations, the ‘Period’ of the RNG must be a factor greater than its
‘Range’.

Minimum Range Requirement

1.27.12 The range of values produced by the RNG must be adequate to provide sufficient precision and
flexibility when setting event outcome probabilities, (i.e. so as to accurately achieve a desired
expected return to player).

Mapping

1.27.13 Mapping of random numbers into symbols (or reel stop positions) should observe the following
principles:

a) the output resulting from the mapping of an RNG to symbols (or reel stop positions) must
not be predictable;

b) any outcome derived from the random number generator are uniformly distributed;

c) any mappings to convert random numbers into game symbols are linear, and the
distribution of the mapped symbols is identical to the distribution of the unmapped random
number from which they were derived;

d) the mapped random number sequence must demonstrate that they are statistically random
when subject to the same statistical tests for randomness specified for the base random
number generator; and

e) the game outcomes which are derived from either a combination of mapped symbols or
directly from the unmapped random numbers must have the same distribution and
probability of occurrence as the game that the machine implements. In particular, poker
games must have the same first hand distribution and probability as hands dealt from a
randomly shuffled deck of cards; spinning reel games must have the same outcome
probabilities and outcome distribution as the physical model upon which the game is
based, and so on.

Scaling Algorithms

1.27.14 If a random number with a range shorter than that provided by the RNG is required for some
purpose within the gaming machine, the method of re-scaling, (i.e. converting the number to the
lower range), is to be designed such that all numbers within the lower range are equally
probable.

1.27.15 If a particular random number selected is outside the range of equal distribution of re-scaling
values, it is permissible to discard that random number and select the next in sequence for the
purpose of re-scaling.