
A scatter plot maker is a tool that plots one dot per row of data, using two numeric columns as the x and y positions, so you can see whether the two variables are related. For most people the best free option is Google Sheets or Excel, which also add a line of best fit with its equation and R². Use Desmos for math class regressions, RapidTables for a quick plot with no account, and Datawrapper when the scatter plot goes on a web page. This guide compares those tools, shows how to read a scatter plot, and works through one example with real regression numbers.
We checked every tool in this guide on its own website or help center on November 1, 2026. The example data is fictional; the statistics were calculated from it.
Use a scatter plot when every item has two measurements and you want to know whether they move together: hours studied and exam score, ad spend and sales, height and weight.
| You want to show | Better chart |
|---|---|
| Whether two numeric variables are related | Scatter plot |
| How one value changes over time | Line graph |
| How categories compare | Bar graph |
| A third variable, such as company size | Bubble chart (a scatter plot where dot size is the third value) |
Google describes the use case simply: a scatter chart helps you "find out how much one variable is affected by another" (Google Docs Editors Help). For other chart types, read our graph maker guide to choosing a chart and free tool.
Checked on November 1, 2026.
| Tool | Account needed | Line of best fit | Shows equation and R² | Getting the chart out | Best for |
|---|---|---|---|---|---|
| Google Sheets | Google account | Linear, exponential, polynomial, logarithmic, power, moving average | Yes (R² needs the legend turned on) | Download as PNG, PDF or SVG | Shared data and class projects |
| Microsoft Excel | Free for Excel for the web; desktop needs Microsoft 365 | Linear, logarithmic, polynomial, power, exponential, moving average | Yes, as chart options | Copy into Office files; Save as Picture on desktop | Analysts and lab reports |
| Desmos | None | Regression templates or your own model | Shows r for linear fits, R² for others, plus residuals | Share link or image | Math and statistics classes |
| RapidTables | None | Linear, exponential or polynomial per series | Draws the line | Print or save from the browser | A quick plot with no sign-up |
| Datawrapper | Free account | Linear, quadratic, cubic, exponential, logarithmic, power or custom (Refine tab > Trend line) | Draws the line | Responsive embed; free plan exports PNG with attribution | Scatter plots on websites |
| Python (Matplotlib + NumPy) | None | Any model you code | Yes, you calculate it | Any format | Repeatable analysis |
Quick picks:
Every scatter plot maker expects the same shape:
For several groups on one plot, Google Sheets reads each extra column of y values as a new series (Google). In Excel, add each group with Select Data > Add.
Twelve fictional students:
| Hours studied (x) | Exam score (y) |
|---|---|
| 1 | 52 |
| 2 | 58 |
| 2.5 | 55 |
| 3 | 63 |
| 4 | 66 |
| 4.5 | 71 |
| 5 | 68 |
| 6 | 75 |
| 6.5 | 79 |
| 7 | 77 |
| 8 | 85 |
| 9 | 88 |

What the numbers say:
And what they do not say: correlation is not causation. Students who study more may also sleep better or attend more classes. A scatter plot shows that two variables move together, not why.

Look at four things before you add any trendline:
Google Sheets
Excel
Desmos
y_1 ~ m x_1 + b on a new line.TI-83 / TI-84 graphing calculator
Python
import numpy as np
import matplotlib.pyplot as plt
x = np.array([1, 2, 2.5, 3, 4, 4.5, 5, 6, 6.5, 7, 8, 9])
y = np.array([52, 58, 55, 63, 66, 71, 68, 75, 79, 77, 85, 88])
m, b = np.polyfit(x, y, 1)
r = np.corrcoef(x, y)[0, 1]
plt.scatter(x, y)
plt.plot(x, m * x + b)
plt.title(f"y = {m:.2f}x + {b:.2f}, R² = {r**2:.2f}")
plt.xlabel("Hours studied")
plt.ylabel("Exam score")
plt.savefig("scatter.png", dpi=200)
This gives the same result as the chart above: y = 4.55x + 47.56, R² = 0.97.
Statistics classes often ask you to show the working. Here it is for the 12 students above, using the least-squares formulas that Sheets, Excel and Desmos use.
| Step | Formula | Result |
|---|---|---|
| Mean of x | x̄ = Σx ÷ n = 58.5 ÷ 12 | 4.875 |
| Mean of y | ȳ = Σy ÷ n = 837 ÷ 12 | 69.75 |
| Sum of cross-products | Sxy = Σ(x − x̄)(y − ȳ) | 312.125 |
| Sum of squares of x | Sxx = Σ(x − x̄)² | 68.5625 |
| Sum of squares of y | Syy = Σ(y − ȳ)² | 1,466.25 |
| Slope | m = Sxy ÷ Sxx | 4.5524 |
| Intercept | b = ȳ − m·x̄ | 47.557 |
| Correlation | r = Sxy ÷ √(Sxx · Syy) | 0.9844 |
So the line is y = 4.55x + 47.56, and R² = 0.9844² ≈ 0.97.
Check the residuals. A residual is the actual y minus the predicted y. The largest one here belongs to the student who studied 2.5 hours and scored 55: the line predicts 58.9, so the residual is −3.9. No residual is far from the rest, so there is no outlier pulling the line.
| Mistake | Fix |
|---|---|
| Using a line chart for two numeric variables | Use a scatter chart, so x values are placed by value |
| Swapping x and y | Put the explanatory variable on x, the response on y |
| Forcing a straight line through a curve | Check the form first; try a polynomial or log trendline |
| Reporting a trend from a handful of points | Say how many points you have; collect more if you can |
| Extending the line far beyond the data | Show the fit only across the data range, or mark the extension |
| Thousands of overlapping dots | Use smaller, semi-transparent dots, or a heat map |
| Claiming one variable causes the other | Say "is associated with" unless you ran an experiment |
| Alt text that only says "scatter plot" | Describe the relationship, such as "scores rise about 4.6 points per study hour" |
Google Sheets and Excel are the best free scatter plot makers for most people, because they also add a line of best fit with its equation and R². Desmos is best for math classes, RapidTables for a quick plot without an account, and Datawrapper for scatter plots with trend lines on websites.
Put x values in the first column and y values in the second, insert a scatter chart, then add a linear trendline. In Google Sheets it is under Customize > Series > Trendline; in Excel it is under Chart Elements > Trendline. Turn on the equation and R² to show the fit.
R² (the coefficient of determination) shows how much of the variation in y the trendline accounts for, from 0 to 1. An R² of 0.97 means the line accounts for about 97% of the variation in that data set. It does not prove that x causes y.
r, the correlation coefficient, runs from −1 to 1 and shows the direction and strength of a straight-line relationship. For a simple linear fit, R² is r squared, so it is always between 0 and 1 and has no sign.
A scatter plot places every point by two numeric values to show a relationship. A line graph spaces x values evenly and connects points to show change over time or another ordered sequence.
Yes. Color the dots by group to add a category, or use a bubble chart, where dot size shows a third number. Google Sheets and Excel both make bubble charts.
Calculate the means of x and y, then the slope m = Σ(x − x̄)(y − ȳ) ÷ Σ(x − x̄)² and the intercept b = ȳ − m·x̄. The worked table above shows each step for 12 data points.
Yes. Scatter plot, scatter graph, scatter chart, scatter diagram and XY graph all mean the same chart.
A scatter plot is usually one step in an analysis: the data, the chart, the fit and your conclusion belong together. In AFFiNE you can keep the source table, the exported chart and your notes in one doc, or lay out several plots side by side on the AFFiNE whiteboard.
Update note (November 1, 2026): This article was first published in June 2025 and rewritten in November 2026. We replaced a long, generic overview with a dated comparison of six scatter plot makers checked on each vendor's site, an original worked example with regression results we calculated (y = 4.55x + 47.56, R² = 0.97), a guide to reading scatter plots, steps for adding a line of best fit in Google Sheets, Excel, Desmos, a TI-84 and Python, a by-hand calculation with residuals, graph-paper steps, and new FAQs on r and R². We replaced the shared cover and unlabeled images with original graphics. AI tools assisted with drafting, and the content was reviewed and edited by the AFFiNE team.