Recently I embarked on a project in which I required a list of completely random entries. A simple thing to do yes, well no, and that is because ‘random’ is not always as random as we are told. Here is an explanation.
Algorithmic
Algorithmic randomness refers to the generation of seemingly random numbers or sequences using algorithms or computational methods. While these algorithms are designed to produce results that appear random, it’s important to note that they are ultimately determined by the programming and logic implemented by the developers.
In theory, if there are biases or patterns within the algorithm or programming code, it could potentially introduce biases or non-randomness into the generated sequences. The inherent limitations and biases of the algorithms themselves or the programming choices made by the developers can influence the randomness or lack thereof in the results.
To mitigate this, developers of random number generators (RNGs) aim to create algorithms that pass various statistical tests for randomness. These tests analyse the generated sequences for patterns, biases, or other statistical anomalies. Reputable RNGs go through rigorous testing and evaluation processes to ensure they meet certain standards of randomness and fairness.
However, it’s important to recognize that achieving true randomness is a complex challenge, and even well-designed RNGs may have limitations or subtle biases that could be identified under specific circumstances or through advanced statistical analysis.
A remarkable solution
I came across a website that appears to have a solution for the ‘bias’ problem in random generation.
Random.org is a website that provides random numbers and randomization services for various purposes. One of the methods it uses to generate random numbers is through ‘atmospheric noise’.
Atmospheric noise refers to the random fluctuations or variations in the electromagnetic signals present in the Earth’s atmosphere. These fluctuations can come from various natural sources, such as cosmic radiation, lightning, and other atmospheric disturbances. The randomness of atmospheric noise makes it suitable for generating random numbers.
Random.org captures and analyses atmospheric noise to generate random data. The process involves collecting samples of atmospheric noise and subjecting them to statistical analysis to extract random numbers. The specific details of the analysis techniques employed by Random.org are proprietary and not publicly disclosed.
Atmospheric noise
By utilizing atmospheric noise, Random.org aims to provide a source of randomness that is independent of human influence or predictable patterns. This makes it useful for applications where true randomness is required, such as cryptography, simulations, and other randomization processes.
Note that Random.org is just one option for generating random numbers, and there are other methods and sources available as well depending on one’s needs and preferences. It can be useful to explore different randomisation techniques or sources of randomness to suit specific requirements.