Why 'Random' Results Can Feel Suspiciously Repetitive Over a Small Sample
Genuine randomness doesn't guarantee even distribution over a small number of trials — streaks, repeats, and apparent "clusters" are a normal, expected feature of true randomness, not evidence that a generator is broken; this feeling of suspicious repetition is a well-documented perceptual bias, since humans tend to expect random sequences to look more evenly spread out than true randomness actually produces.
This is one of the most counterintuitive aspects of randomness, and it explains a lot of unwarranted suspicion directed at genuinely working random number generators.
Why streaks are actually expected in true randomness
Over a small number of random draws, getting the same number twice in a row, or several results clustered in a similar range, is entirely consistent with genuine randomness — a truly random process has no "memory" preventing repeats, and over a small sample, some clustering is statistically normal rather than exceptional.
Why humans perceive this as suspicious
People tend to expect a random sequence to look evenly spread out with no repeats or streaks — a well-documented perceptual bias — but a sequence deliberately arranged to avoid all repeats and clustering is actually less random, not more, since it imposes an artificial pattern (avoidance of repeats) that true randomness doesn't have.
When repetition would actually be a red flag
Over a genuinely large number of trials, a truly random process should show roughly even distribution across all possible outcomes — persistent bias toward specific numbers over a large sample size, not just an occasional short streak, would be the actual signal of a flawed or non-random generator.