There are a number of ways, all under goodness of fit (which I forgot some decades back).

The simplest way (if the numbers are positive and you have a good regression line) is to calculate the y value on the line for the x value of the point.

The vertical distance between the line is y(line) - y(point).

That is one way and that is not the only way.

The simplest way (if the numbers are positive and you have a good regression line) is to calculate the y value on the line for the x value of the point.

The vertical distance between the line is y(line) - y(point).

That is one way and that is not the only way.