Randomness is the foundation of gambling games – what people really do when they gamble is bet on which of several possible randomly-determined outcomes will occur. For those who enjoy gambling, the uncertainty that randomness brings makes it exciting to find out what the outcome happens to be. What number does the roulette wheel land on? What cards do your opponents have? Where will the slot wheels stop? And of course, there’s the fact that you win money when your guess matches up to reality

Eneba Many GEOs

True randomness – theory and practice

It is actually very hard, if not impossible, to have “true” randomness. But equally, in physical roulette, the number which the ball lands on is the product of a number of interacting parameters. There is the speed at which the wheel is spun, how and when the ball is thrown in, the elasticity of the ball and the material of the wheel, as well as the shape of the surface of the wheel and the depth of the pockets. Arguably something is not truly random if so many outside factors affect it; but it would still be nigh on impossible for a player to observe these parameters for a given spin, predict where the ball will land, and place their bet before the croupier places the dolly on the table.

As such, roulette ends up being entirely random in practice: the players usually cannot know or measure these parameters to anything like the required degree of accuracy. That’s a lot of number crunching. Indeed, a study has been done on whether observing the behavior of a roulette wheel over the long term and detecting whether its accumulation of defects and wear-and-tear leads certain numbers to come up more often than others. This was the successful approach of the famous gambler Richard Jarecki, though it took a team of 10 to collect and analyze a mountain of data from across Europe before Jarecki could make this strategy pay off.

Pseudorandomness – seeds and algorithms

Nowadays many gamblers prefer to play online, where it’s possible to play casino slots for real money. At sites like Slots of Vegas, there is an abundance of games such as online slots, video poker, table games, and a variety of roulette options. So how does randomness in the world of online gambling compare to that in land-based games? When it comes to computer systems, again it is very difficult to generate “true” or total randomness: note that most games use what are called “pseudorandom number generators” – the name says it all. This is because it is very technically difficult to get a computer to generate a random outcome because they operate deterministically – that is, they apply logical rules in sequences, making them predictable if you know the rules and the sequence. That’s true technically speaking, but not in practice.

For most purposes, random number generators are perfectly adequate. In the same way that predicting a roulette wheel’s outcome is technically possible but virtually impossible in practice, random number generators are essentially random from the player’s perspective. They use highly complex algorithms that take one number as an input, known as the seed, and spit out another number. That output number then becomes the new input for the next game, which generates a new output, and so on.

To “hack” such a machine, you would have to know both the seed and the algorithm to predict the sequence of output numbers. But, unlike with the roulette wheel, you can’t even observe these parameters and try to measure them. Online casinos keep their algorithms and seeds secret, plus they tweak and change them all the time. The algorithms are so complex it is also practically impossible to reverse-engineer them. And with the intense data gathering that would be required to determine the pattern and the time and cost this would involve, it does not seem likely doing this would be worth anyone’s while!

Can it be done on a computer?

Computers use the flow of electricity to perform computations and electricity is the flow of sub-atomic particles, which are subject to the semi-random rules of quantum mechanics. There is an argument that, as computations can be affected by random fluctuations in this way, if harnessed properly, that could be used to help generate random sequences of numbers. But it would take a lot of programming sophistication: unnecessary when pseudorandom random number generators are just as good to the naked eye.
So hacking a pseudorandom number generator is much like the prospect of ‘hacking’ a traditional gambling game – possible in theory but virtually impossible in practice.

%d bloggers like this: