On our perpetual journey through chances and choices that govern online gambling realms, we sometimes stumble into simple questions with complicated answers. One such inquiry, perhaps the most often one in the iGaming industry, is indeed straightforward — is the random number generator (RNG) truly random?
In short: no, not entirely, although it might be as much as possible in the first place.
Since this is kinda no answer (but it’s the only honest one, backed by science), to truly understand all the whys, hows, and whats of the RNGs — which we intend to provide here in the simplest possible way — we have to take one step back.
Firstly, we need to understand what randomness is.
According to Merriam Webster dictionary, it is a haphazard occurrence that happens “without definite aim, direction, rule, or method,” and something that lacks a specific “plan, purpose, or pattern.”
By this definition, true randomness which results in unpredictable, honest, and reliable sequences that happen with no pattern whatsoever exists only in nature. It cannot be found anywhere else.
To have such randomness in any artificially-generated RNG would equal using entirely accidental quantum phenomena, recording them, and then feeding a computer with such data which would, in turn, produce a truly random RNG.
More often than not, this is not quite possible (though we’ll get back to this later).
…by the criterion of definition, which is also used in scientific elaborations, not a single RNG is random in online gambling or anywhere, for that matter.
To substitute and compensate, computer engineering uses two emulating programming methods — quasi-random and pseudo-random — and one hardware solution. For the time being, that’s the closest we’ll ever get to the randomness in random numbers creation.
So, what are they?
Whenever any computer programmer tries to create an RNG, he or she essentially uses textbook algorithms to do so. In other words, an engineer takes the set of logical instructions that can be given to a computer which then produces a quasi-random RNG as it tries to emulate true randomness.
The challenge in this approach is that results show a pattern since the machines are governed by the programming.
Dr. John von Neumann (1903-1957), considered to be “the last representative of the great mathematicians,” that also worked on the Manhattan Project, famously explained randomness in mathematics in 1951: “Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin.”
Surely, it might be demanding to identify the pattern, but the numbers which come as the result of quasi-random RNGs are not truly random. Instead, they only appear to be random.
By measuring and testing the outcome of quasi-randomness in the long-run — hundreds of thousands, even millions of repetitions, that is, rolls or hits — the programming mold does show.
Indeed, when it comes to craps, roulette, or backgammon, the quasi-randomness in RNG can be noticed after years of observation. The most advanced players use it occasionally, just as some online casinos do, which puts them in an advantageous position.
Luckily, when it comes to online hubs that use this method, they’re rare nowadays. Usually, these are underdeveloped operators often residing in murky online jurisdictions, which is to say, beginners or ones looking for a quick hit-and-run.
With the advent of computer programming, a statistically independent method to recreate true randomness emerged as the RNG standard in current interactive games — pseudo-random method.
Dr. Steve Ward, Professor of Computer Science and Engineering at MIT, sheds some light into pseudo-randomness and software engineering:
“One thing that traditional computer systems aren’t good at is coin-flipping. They’re deterministic, which means that if you ask the same question you’ll get the same answer every time. In fact, such machines are specifically and carefully programmed to eliminate randomness in results. They do this by following rules and relying on algorithms when they compute. On a completely deterministic machine, you can’t generate anything you could really call a random sequence of numbers because the machine is following the same algorithm to generate them. Typically, that means it starts with a common ‘seed’ number and then follows a pattern. They are what we call ‘pseudo-random’ numbers.”
For the majority of practical applications, though, this approach is more than enough.
One can sample huge data, run a CD player, or conduct lottery with an absolute sense of randomness since there’s “there’s no quantitative advantage in the degree of randomness.”
…this is still not the true randomness by definition.
Consequently, when it comes to gambling — and answering the question from the beginning — even the pseudo-random RNG is not truly random.
Furthermore, as Dr. Ward also points out, “if you go to an online poker site, for example, and you know the algorithm and seed, you can write a program that will predict the cards that are going to be dealt.”
The most devoted eSports players of MMOGs (massively multiplayer online games) are fully aware of this and strive to master pseudo-random patterns to secure the best gear or loots in the opportune moments. Their online forums elaborate on pseudo-randomness quite often.
So, is the real randomness a unicorn?
Not quite. The only completely unpredictable random number generator is a hardware device that creates numbers from physical processes — changes that affect the form of a chemical substance but not its composition — instead of a software algorithm.
These devices are based on microscopic phenomena that generate statistically random signals such as thermal noise (agitation of the electrons inside of an electrical conductor which happens regardless of applied voltage, present in any electrical circuit), the photoelectric effect (emission of electrons when light hits any material which, in turn, create photoelectrons), or any quantum phenomena (like superfluidity, superconductivity, or the quantum Hall effect).
In other words, hardware RNGs are based on randomness that exists in nature.
This method is today used in data encryption to create cryptographic keys or in security protocols (TLS/SSL, among others), since these devices create sequences that are, at least in theory, unpredictable, and more resistant to cryptanalysis.
As Dr. Wards notes, the use of hardware RNGs makes reverse engineering of a poker algorithm impossible, because they rely on unpredictable processes instead of human-defined patterns.
Of course, as he also notes, “the results might still be slightly biased towards higher numbers or even numbers, but [at least] they’re not generated by a deterministic algorithm.”
Why this bias?
Because hardware RNGs can produce only a limited number of random information per second. To increase the output, devices are only used to create the ‘seed’ — a number which is used to initialize pseudo-randomness — and afterward, the software takes over and boost up the whole sequence.
Be it as it is, only such well-designed processes can assure that the outcome of the roulette wheel will be the most random possible.
One Extra Not So Random Layer
Thus, at the end of the day, the answer to a simple question from the beginning has to be ‘no’ although it might sometimes be ‘yes’.
In terms of the fundamental randomness, not a single human-created RNG can be random and therein lays the obvious ‘no’. It’s simply impossible to create truly random numbers by any arithmetical method.
The most dedicated members of our community may further journey through the 62-page study by IBM Haifa Research Laboratories on all types of randomness, which explains that even the number π (3.14) is prone to patterns.
…in terms of the closest possible resemblance to the true unpredictability, only the well-tested and certified RNGs can at least give some notion of randomness, which is why there’s also ‘yes’ in the answer.
In the iGaming industry, the latter is of paramount importance.
All RNGs are created with some form of certainty that players will win, which is the essence of RTP. The potential issue, of course, is that none of us know what algorithm, let alone device, casinos use when it comes to creating RNGs.
And, that’s where each one of us comes into the picture.
It’s up to us to choose certified online casinos that perform regular RNG tests, carry out software integrity check-ups, and conduct fairness audits of games — which the best interactive operators do. Other than being well informed, that’s the only way to at least somehow suppress a lack of randomness.
Also, that’s the only thing any prudent and responsible player can do, other than choosing games wisely, using the skills to the best of its knowledge, and staying well within the limits of gaming budget.
Such determination should not be random, just as any RNG is not random at all.