Linear extrapolation

Linear extrapolation is a method to predict future data points given known data points, by assigning a line that lies closest to all the known data points on average.

The linearly extrapolated data yields good predictions if either the underlying data points follow a linear relation or if the predicted data is not too far beyond the known data. In general more than two known data points are required to perform any meaningful linear extrapolation.

Further reading: https://en.wikipedia.org/wiki/Linear_regression