Random Numbers, Histograms, and Distributions in Desmos

TL;DR

Desmos’s normal distribution uses mean and standard deviation to set the bell curve’s center and spread.

Briefing Cornell Notes

Briefing

Desmos can generate random numbers from a chosen statistical distribution, then use those samples to build histograms that gradually “converge” toward the theoretical curve—especially for the normal distribution. The core workflow is: pick a distribution (mean and standard deviation for normal), generate many random observations from it, plot the results, and then compare a histogram of the sample to the distribution’s bell curve. As the sample size grows, the histogram’s shape becomes increasingly faithful to the underlying model, turning randomness into a measurable pattern.

The lesson starts with Desmos’s built-in normal distribution and how its parameters control the curve. Setting a mean of 10 and a standard deviation of 3 centers the bell curve at 10, where the highest density occurs. The standard deviation then quantifies spread: using the familiar empirical rule, about 68% of values fall within 1 standard deviation of the mean (7 to 13), about 95% within 2 standard deviations (4 to 16), and about 99% within 3 standard deviations (1 to 19). Cumulative probability is also used to interpret tail behavior—for example, the chance of getting a value below 4 is around 2.2%, making values outside 4 to 16 relatively uncommon.

With that distribution defined, Desmos’s random function becomes the bridge from theory to simulation. Using random with the normal distribution, a single draw produces one observation (such as 11.6, plausibly near 10). To generate a sample, the transcript uses a count parameter n (an integer) to create a list of random observations, then plots those points along the x-axis. A key practical feature is seeding: changing a seed value (like 1, 2, or 3) forces a different but reproducible set of random draws, effectively “rolling the dice” again for the entire sample.

The next step is histograms, which translate the raw random points into frequency bins. A histogram takes the dataset and a bin width, then counts how many observations land in each interval (e.g., between 3.5 and 4.5, 4.5 and 5.5, and so on). With only 10 observations, the histogram looks jagged and far from the smooth bell curve. Increasing the sample size to 100 makes the pattern clearer but still imperfect: even with a normal distribution, finite samples won’t match the theoretical shape exactly.

To make the comparison more meaningful, the histogram can switch from counts to relative frequency (so bin heights represent proportions, such as 0.15 for 15 out of 100 in a bin). As the number of observations rises further—into the thousands and beyond—the histogram aligns more tightly with the normal distribution curve. Adjusting bin width also matters: smaller bins produce a histogram that more closely traces the underlying density function. The result is a practical “reverse engineering” loop: generate data from a known distribution, build a histogram from that data, and watch the sample distribution converge toward the theoretical normal curve as randomness averages out.

Cornell Notes

Desmos can simulate random data from a chosen distribution and then visualize it with histograms. For a normal distribution, setting a mean (like 10) and standard deviation (like 3) determines where the bell curve peaks and how wide it spreads. Using the random function, Desmos generates n observations from that distribution; changing the seed produces a different reproducible sample. A histogram bins those observations and shows how often values fall into each interval. With small samples the histogram is noisy, but as the number of observations increases (and bin width is adjusted), the histogram’s shape converges toward the theoretical normal distribution curve.

How do mean and standard deviation change the normal distribution in Desmos?

Setting a mean M shifts the bell curve horizontally; the highest point (peak) occurs around M. Setting a standard deviation s controls spread: larger s makes the curve wider and flatter, while smaller s makes it narrower and taller. In the example with mean 10 and standard deviation 3, most values cluster around 10, and the spread around that center reflects how much variation is expected.

What does the “68–95–99” style rule mean for values generated from N(10, 3)?

With mean 10 and standard deviation 3, 1 standard deviation corresponds to 10 ± 3, i.e., 7 to 13. That interval is expected to contain about 68% of values. Two standard deviations gives 10 ± 6, i.e., 4 to 16 (about 95%). Three standard deviations gives 10 ± 9, i.e., 1 to 19 (about 99%). These are theoretical expectations that guide what ranges should appear frequently in simulated samples.

How does Desmos generate random numbers from a distribution, and what role does the seed play?

Desmos’s random function can be used to draw observations from a specified distribution (here, the normal distribution with chosen parameters). A single draw returns one value; using n generates a list of n values. Adding a seed (e.g., a seed slider) makes the random sample reproducible: changing the seed regenerates a different sample of the same size from the same distribution.

Why does a histogram look “wrong” with only 10 random observations?

A histogram bins data into intervals, and with only 10 points, each bin count is heavily affected by chance. The resulting bar heights jump around, so the histogram doesn’t yet form a smooth bell shape. Even when the underlying distribution is normal, finite samples won’t perfectly match the theoretical curve.

What changes make a histogram converge toward the normal distribution?

Two main levers are sample size and bin width. Increasing the number of observations (from 10 to 100 to 1,000 and beyond) reduces random fluctuation, making the histogram’s shape more stable and closer to the bell curve. Using relative frequency (proportions per bin) helps compare shapes across different sample sizes. Shrinking bin width also improves the match because the histogram better approximates the continuous density function.

Review Questions

If the mean is changed from 10 to 7 while keeping the standard deviation at 3, what happens to the peak location and the expected ranges for 68%, 95%, and 99% of values?
How would increasing bin width likely affect the histogram’s ability to match the smooth normal density curve?
What effect should changing the seed have on the sample, and what should remain the same?

Key Points

1
Desmos’s normal distribution uses mean and standard deviation to set the bell curve’s center and spread.
2
The empirical rule provides expected ranges: about 68% within ±1 standard deviation, 95% within ±2, and 99% within ±3.
3
The random function can generate lists of observations from a specified distribution, producing simulated data for analysis.
4
A seed value makes random samples reproducible; changing the seed regenerates a different sample of the same distribution.
5
Histograms convert random samples into binned frequencies; small samples produce noisy, jagged histograms.
6
Switching to relative frequency and increasing sample size makes histogram shapes more comparable to theoretical distributions.
7
Smaller bin widths improve the histogram’s approximation of the continuous density curve.

Highlights

A normal distribution with mean 10 and standard deviation 3 predicts most values fall between 7 and 13, with much smaller chances outside 4 to 16.

Seeding in Desmos lets users regenerate different but repeatable random samples from the same distribution.

A histogram built from only 10 draws looks irregular, but it becomes bell-shaped as the number of observations grows.

Relative frequency (proportions per bin) and bin width both strongly influence how closely a histogram matches the theoretical curve.