How to Play
🎲 First Rolls: See Randomness in Action
Start with the default — two dice and 100 rolls. Press Play to send the dice tumbling. Each roll flashes the result in the tray and raises a bar in the histogram. Watch how the bars dance — uneven, jumpy, sometimes surprising. This is what small samples look like: randomness on full display.
Try pausing after ten or twenty rolls to make a quick prediction: which total will end up tallest? Then continue and see whether the trend holds. Even in short runs, you’ll notice that middle numbers like 6, 7, or 8 start winning more often — and that pattern only gets stronger over time.
⚙️ Experiment with the Controls
Use the Dice per roll slider (1–10) to change how many dice are rolled at once. With one die, results are spread evenly from 1 to 6. With two dice, totals range from 2 to 12 and the chart forms a small mountain around 7. Add more dice and that mountain becomes wider and smoother.
Next, use the Roll count slider to decide how many times to throw. Try short bursts of 10 or 20 rolls to see quick wiggles, or long runs of 500 or 1000 to watch the bars stabilize. Notice how increasing the number of trials reduces the “wiggle” and brings the shape closer to a gentle curve.
📈 Compare Data to the Expected Curve
Turn on the comparison overlay to see the predicted shape for your chosen number of dice. The game draws a smooth theoretical curve — a preview of what would happen if you rolled forever. As you gather more data, the live bars begin to match the curve almost perfectly. This side-by-side view helps you understand the power of sample size and why averages matter.
🔁 Reset, Predict, Repeat
When a run finishes, press Reset to clear the board and set up a new scenario. Try doubling the dice, halving the roll count, or switching back to a single die to see how the distribution changes. Before each run, make a prediction about which totals will appear most often — then test your guess. Each run feels like a mini experiment where randomness gradually gives way to predictability.