Residuals and the pmcc
Looking at residuals and lines of best-fit
Try moving the points to see the effect upon the various statistics. You can either look at any line or lock to the "centre of gravity" of the points. How can you make r as small as possible?