Dice Roller · Practical guide
Dice Roller for Probability Experiments
Generate real random data fast and watch the law of large numbers even out your results.
At a glance
Fast Random Data Without Physical Dice
A probability experiment needs lots of independent, equally likely outcomes — exactly what a dice roller produces. Set the dice count and sides above, roll repeatedly, and the history panel logs your last 10 rolls with individual values and totals. The roller uses your browser's cryptographically secure random generator, so every face is equally likely.
Interactive tool
Dice Roller
Enter your own values, then compare the result with the examples below.
Roll Result
History (Last 10)
A Simple Classroom Experiment
Roll a single d6 thirty times and tally each face. The expected count per face is 30 ÷ 6 = 5, but actual counts will wander — getting a 4 or a 7 for some face is normal. Roll 300 times and each face should land within a few of 50. That convergence toward the expected 16.7% per face is the law of large numbers in action.
Expected Values to Compare Against
| Setup | Expected Average | Range |
|---|---|---|
| 1d6 | 3.5 | 1-6 |
| 2d6 | 7.0 | 2-12 |
| 1d20 | 10.5 | 1-20 |
| 3d6 | 10.5 | 3-18 |
Tips for a Clean Experiment
- Decide your sample size before you start — 50, 100, or 500 rolls — and stick to it.
- Record every roll, not just the interesting ones; cherry-picking destroys the statistics.
- Compare the mean of your sample to the expected average, and note how the gap shrinks as you add rolls.
- For sums like 2d6, compare your distribution to the theoretical table: 7 should appear about 6 times in every 36 rolls.
Frequently Asked Questions
How many rolls do I need for a probability experiment?
At least 30 rolls show the rough shape of a distribution, but 100 or more gives results that sit close to the theoretical probabilities. More rolls always means tighter agreement.
Why do my results not match the expected percentages?
Short samples are noisy. In 30 d6 rolls, a face appearing 2 or 8 times instead of 5 is ordinary variation, not a broken roller. Keep rolling and the percentages converge.
Is this roller random enough for a school experiment?
Yes. It uses the browser's cryptographically secure random number generator, which passes the same statistical randomness tests used for security applications.
Can I roll several dice at once to speed things up?
Yes. Set the count up to 20 dice per roll. Each die is independent, so a 20-dice roll gives you 20 data points at once; the result lists every die individually.