Computers are very systemized and none of the procedures conducted by them are random. But computers are seldom required to generate a random number for many practical applications like gaming, accounting, encryption/decryption and many more. The number generated by the computer relies on the time or the CPU clock. A given computer can be programmed to return random number (or character) arrays from a number (or character) data set. The returned dataset can have repeated values. Even though the repeated values are not related with its degree of randomness (in fact, it may be a sign of higher randomization), but still to humans, it appears as biased or not random. We propose an algorithm to minimize repetition of values in the returned data set so as to make it appear more random. The concept proposes returning a data set by using biased or nonrandom procedure in order to make it more random in “appearance”. © 2018, Springer Nature Singapore Pte Ltd.