CasinoCityTimes.com

Home
Gaming Strategy
Featured Stories
News
Newsletter
Legal News Financial News Casino Opening and Remodeling News Gaming Industry Executives Author Home Author Archives Search Articles Subscribe
Newsletter Signup
Stay informed with the
NEW Casino City Times newsletter!
Recent Articles
Best of Alan Krigman
author's picture
 

Pick a Number, any Number (Well, almost any)

17 April 2002

About 10 years ago, word got out that slot machine payouts don't just "happen." Results are determined, ostensibly unpredictably, by internal electronic engines known as random number generators or RNGs. Ever since RNGs got unmasked, solid citizens wanting to understand the systems in which they're placing their hopes for comfortable futures, have been asking questions about them.

Unfortunately, they usually ask the wrong questions. This isn't surprising. Randomness - what it is, and whether it can be created by the mind and hand of man -? has baffled mathematicians for generations. Not that the likes of Andrei Kolmogorov, John von Neumann, or Donald Knuth cared about the slots. Randomness and use of RNGs affect many critical computer-oriented tasks, from encoding messages for security to simulating physical structures to study safety under the forces of natural phenomena. Gaming machine designers, however, have been quick to jump on developments prompted and financed by such applications.

The trouble with asking the wrong question is that you're often worse off after you have the answer than you were before. This, because you think you know something consequential, or at least relevant, and you don't. So you have a false sense of confidence.

Here's the "wrong" question seemingly posed most often about RNGs and the slots. "When during a round does the RNG get triggered?"

One thing wrong is the implication that the RNG is dormant until tweaked, when it produces a value. Were this true, an RNG would step through a sequence which looks random, yet someone who knew the software could predict slot results. In fact, the RNG chugs along continuously, at a rapid clip. The trigger just picks the most recently synthesized value. Tripping the device at one instant, or a split second later, yield vastly different results.

Another problem with this question is the connotation it carries that a device may be programmed to account for factors such as number of lines or coins in play, or whether bets are made using credits or cash. A related idea is that the outcome of a spin can be influenced using skill or secret information about when or how to bet, push a button, or pull a lever. Neither notion is valid. The chances associated with the possible outcomes of a round are independent of the means of picking the value from the RNG.

Three questions about RNGs do turn out to be important to good gamblers. Not from the perspective of acquiring skill, since these are games of chance. Competence as a slot player has to do with establishing goals, betting strategies, and exit criteria as opposed to operating the machines. Rather, the importance involves understanding the activity and trusting its integrity.

The first key question is whether all numbers over the specified range are equally likely. For instance, say a machine has three reels, each with 10 virtual stops. A thousand numbers are needed to reach all combinations. Slot makers must prove: 1) that their RNGs can produce each number, and 2) that occurrences equalize, not in every 1,000-step sequence, but averaged over the long run.

The second meaningful question is whether the numbers appear randomly. There is no single universally-accepted statistical test of randomness. Machines, however, are certified by applying various criteria, including "autocorrelations" and failure to match patterns or to be described with simple rules.

The third question involves the "period" of the RNG, the series produced before the sequence repeats itself. Is it long enough to preclude using an earlier period as a pattern? Ideally, the period should exceed the expected machine lifetime. This turns out to be readily achieved - a patent granted this year to one manufacturer includes an RNG which could generate a million new numbers per second, and not cycle in 18,000 years.

Players and casino bosses have complementary stakes in the integrity of RNGs in gaming machines. Neither to be shorted nor to overpay. It's as the poet, Sumner A Ingmark, coyly cautioned:

Dependence on phenomena that no one understands,
Puts too much of our destiny in someone else's hands.

Alan Krigman

Alan Krigman was a weekly syndicated newspaper gaming columnist and Editor & Publisher of Winning Ways, a monthly newsletter for casino aficionados. His columns focused on gambling probability and statistics. He passed away in October, 2013.
Alan Krigman
Alan Krigman was a weekly syndicated newspaper gaming columnist and Editor & Publisher of Winning Ways, a monthly newsletter for casino aficionados. His columns focused on gambling probability and statistics. He passed away in October, 2013.