Use a straight edge to approximate the y-intercept. It turns out to be (0,0). Next, choose a point at the other end of the graph. One such point might be (9.5, 19.5).
Thus, the slope of this regression line is m = rise / run = 19.5 / 9.5 = 2.05.
I would say that the best fit based upon these calculations would be
y = 2x + 0.
Choice B is not exactly the same: y = 2x + 1.
However, Choice B is closest to the calculated y = 2x + 0.