Explore linear modeling of bivariate data by moving the data points up and down. You'll see how the correlation coefficient r describes the linearity of the data, and how the least-squares regression line fits the data. The point indicating the mean of the x-coordinates ("x-bar") and the mean of the y-coordinates ("y-bar") is shown.
Added is a projection point to facilitate discussion of the predictive potential of the model.

Things to try:
Move the points so the slope of the least-squares line is negative. What happens to the sign on the correlation coefficient?
Adjust the points so they fall in a line. What is the correlation coefficient?
Does the point (x-bar, y-bar) ever come off of the least-squares line?
Explore which points have more influence on the least squares line, the (x-bar,y-bar) point and the predicted value