Least Squares Regression

Slope determined from one datapoint

By using only one x- and y-value, you could make an estimate of how steep the line should be. In the figure below, a line is drawn from origo to one of the datapoints. The slope b_i of this line can be calculated as:

where y_i and x_i are the coordinates of datapoint i, and b_i is the slope calculated from this datapoint.

plot of dataset

This estimate would of course depend heavily on which datapoint i you choose to use, and thus be a poor estimate of the overall slope although it would be exact for just that point. A better approach could therefore be to do this for all the datapoints, and then compute some sort of mean value of all the b_i's.