Now we need a way to quantify this and we do that much the same way that we
calculate the standard deviation. We take each residual, in other words the
observed minus expected value. We square them to get rid of negatives. That's
the same as finding the area of each of these squares. That essentially we find
the average square. But we correct for the fact that it's a sample, by dividing
by N-2, where N is the number of points. In this case there are six. And then
finally, we take the square root to put it back into terms of distance rather
than squared values. This is called the standard error of the estimate.