Scatterplots and Informal Lines of Best Fit

A scatterplot shows pairs of values as dots. If those dots roughly follow a straight path, you can draw one line through the middle of the cloud and use it to predict values you don't have data for. This works because the line captures the general trend — the steady relationship between x and y — even though no single real data point has to sit exactly on it. Real data is messy; the line smooths that mess into one usable rule.

To place the line by eye, aim for balance: about the same number of points above the line as below it, with the gaps roughly similar in size on both sides. You're not trying to touch every point — you're trying to split the scatter evenly. Once the line is placed, you can treat two points on it like any two points on a graph: find the slope between them, then use that slope to step forward or backward to the x-value you care about.

This connects directly to the algebra you already know about lines: slope, point-slope form, and plugging in x to solve for y. The new twist is the warning that comes with it — a best-fit line only describes the range of x-values where you actually have data. Push it far outside that range and you're extrapolating, assuming the pattern keeps behaving the same way with no evidence that it does. Good predictions stay close to the data; shaky ones wander far past it.