About the Random Number Generator
This random number generator draws values between any minimum and maximum you set, up to two hundred numbers in a single run. Ask for whole numbers or set the decimal places as high as six for continuous values, tick No repeats when you need a draw without replacement, and sort the result ascending when order does not matter. A summary panel reports the count, sum, average, lowest and highest, so you do not have to paste the output into a spreadsheet to see the shape of what you drew. Every draw is reproducible from its seed.
How to use the Random Number Generator
- 1
Enter the Minimum and Maximum bounds; the maximum must be the larger of the two.
- 2
Set How many values to draw and Decimal places — leave decimals at 0 for whole numbers.
- 3
Tick No repeats to draw without replacement, and Sort ascending if you want the list ordered.
- 4
Press Generate and read the Summary panel for the sum, average and extremes of the draw.
What people use it for
Rolling dice without dice
Set the minimum to 1, the maximum to 6 and the count to the number of dice in play. The summary total gives you the roll instantly for board games and tabletop sessions.
Picking a sample from a numbered list
To audit 20 invoices out of 850, set the range 1 to 850, the count to 20 and tick No repeats — you get twenty distinct row numbers to pull.
Generating test values for a numeric field
Two decimal places between 0.01 and 9999.99 gives realistic monetary amounts for exercising validation, rounding and currency formatting in a form.
Assigning people to groups
Draw one number per participant in the range 1 to the number of teams, leaving No repeats off, and you have a quick random allocation.
Inclusive bounds and how the draw works
For whole numbers both endpoints can come up. The tool takes the ceiling of your minimum and the floor of your maximum, then maps a random fraction across that inclusive span, so a range of 1 to 6 really can produce a 1 or a 6. Decimal output behaves differently: the underlying random value lies in the half-open interval from 0 up to but never including 1, so the result approaches the maximum without ever landing on it. At six decimal places the practical difference is one part in a million and rarely matters, but it is worth knowing when you are testing boundary conditions. No repeats is implemented by rejection — a value already seen is discarded and another drawn — which is why the tool refuses a no-repeat draw when the integer range does not contain at least as many distinct values as you asked for.
Pseudo-random, and what that rules out
The numbers come from mulberry32, a compact 32-bit state generator that multiplies and xor-shifts an incrementing counter into a well-distributed output. It passes the usual statistical tests for uniformity and independence, which is all that matters for simulations, sampling and games. What it is not is unpredictable: the entire sequence is determined by the seed, so anyone holding the seed can reproduce every number you drew. That property is a feature for reproducible research and stable test fixtures, and a disqualification for anything where a participant benefits from predicting the outcome. Prize draws with real value, lotteries and anything subject to gaming regulation need a certified hardware or cryptographic source with an audit trail, because a seeded generator lets whoever learns the seed recompute the winner in advance.
Tips
- Widen the range before increasing the count when a no-repeat draw is rejected.
- Record the seed alongside your results if you may need to demonstrate how a sample was selected.
- For a weighted draw, generate a decimal between 0 and 1 and compare it against your cumulative probabilities.