Question 1

Is Math.random() truly random?

Accepted Answer

No. Math.random() is a pseudo-random number generator (PRNG) — it uses a deterministic algorithm seeded internally by the browser. The output appears random and passes statistical tests, but it is not cryptographically secure. For passwords, tokens, or security-sensitive values, use crypto.getRandomValues() instead. For games, simulations, and lotteries, Math.random() is perfectly adequate.

Question 2

What is the difference between random and pseudo-random?

Accepted Answer

True random numbers derive from physical phenomena — radioactive decay, thermal noise, atmospheric static — and are fundamentally unpredictable. Pseudo-random numbers are produced by algorithms: given the same starting seed, the sequence repeats identically. Modern PRNGs like xorshift128+ (used in V8 for Math.random()) produce sequences that are statistically indistinguishable from true random for most practical purposes.

Question 3

How to pick random lottery numbers?

Accepted Answer

Set Min to 1 and Max to the highest number in your lottery (e.g., 49 for most 6/49 lotteries), then set Count to how many numbers you need (e.g., 6). Each generated number is independently chosen from the full range. Note: since the lottery draw is also random, using a random generator gives you the same statistical odds as any other selection method.

Question 4

What is a fair random number generator?

Accepted Answer

A fair RNG produces each number in the range with equal probability — a uniform distribution. This generator uses the formula: floor(Math.random() × (max - min + 1)) + min, which gives every integer in [min, max] an equal 1/(max-min+1) chance. Fairness can be verified by generating a large sample and confirming each value appears roughly equally often.

Question 5

What are common uses for random number generators?

Accepted Answer

RNGs are used in games (dice rolls, card shuffles, procedural generation), statistical sampling (picking survey respondents, A/B test assignment), simulations (Monte Carlo methods, physics engines), lotteries and raffles, cryptographic key seeding, load balancing, and scientific research requiring randomized controlled trials.

Question 6

Why do repeated numbers show up when I generate six values from 1 to 49?

Accepted Answer

Each value here is drawn independently, which statisticians call sampling with replacement: the previous draw does not remove that number from the pool. A physical lottery draws without replacement, so its six balls are always distinct. Drawing six independent values from 1 to 49 leaves them all distinct only about 73 percent of the time, so roughly one run in four contains at least one repeat. Generate a few extra values and discard duplicates if you need a distinct set.

Question 7

How do you generate a set of distinct random numbers or shuffle a list?

Accepted Answer

Use the Fisher-Yates shuffle rather than repeated independent draws. Build the array of candidates, then walk from the last index down to the first, each time picking a random index between 0 and the current position inclusive and swapping the two entries. Every permutation becomes equally likely, and taking the first k entries of the shuffled array gives an unbiased sample without replacement. Sorting with a random comparator, such as sort(() => Math.random() - 0.5), is a common shortcut but does not give a uniform shuffle, because the comparator is inconsistent and the outcome depends on the sort algorithm.

Question 8

Why does the formula add 1 to (max - min) before scaling?

Accepted Answer

Math.random returns a float in the half-open interval [0, 1), so it can equal 0 but never reaches 1. Multiplying by (max - min) alone would let the floor reach max only if the float hit exactly 1, which it cannot, so the top of the range would never appear. Using (max - min + 1) creates one bucket per integer, making the range inclusive at both ends and giving each value probability 1/(max - min + 1).

Question 9

What does seeding mean, and can I reproduce the same sequence twice?

Accepted Answer

A seed is the starting internal state of a pseudo-random generator; the same seed replays the same sequence. Math.random deliberately exposes no seed, so its output cannot be reproduced from run to run. When a simulation, test fixture, or procedural level needs to be repeatable, implement a small seeded PRNG such as mulberry32 or xorshift and drive it from a fixed integer, keeping Math.random for cases where repeatability does not matter.

Question 10

How can I check that the generated values really are uniformly distributed?

Accepted Answer

Generate a large batch over a small range, tally how often each value appears, and compare the tallies with the expected count of n divided by k, where k is the number of possible values. A chi-squared goodness-of-fit test formalises this by summing (observed minus expected) squared divided by expected across the k buckets and comparing against k minus 1 degrees of freedom. Keep the expected count per bucket at five or more for the test to behave well.

Random Number Generator