Watch y=x1 as the input grows: x=10 gives 0.1, x=100 gives 0.01, x=1000 gives 0.001. The outputs shrink toward 0 without ever equalling it, because 1 divided by a positive number is never 0. The line y=0 is a horizontal asymptote. Now move toward x=0 from the right: x=0.1 gives 10 and x=0.01 gives 100. The outputs explode, and the line x=0 is a vertical asymptote.
Vertical asymptotes come from forbidden inputs, where the denominator is 0 and the numerator is not. The graph never touches them, because the function has no value there. Horizontal asymptotes come from the far ends and depend on degrees: equal degrees give the ratio of the leading coefficients, a bigger denominator gives y=0, and a numerator exactly one degree bigger gives a slanted (oblique) asymptote instead.
Two features are easily mistaken for asymptotes. A zero, such as x=2 for x2+1x−2, is a point the graph actually hits. A hole, such as x=1 in x−1x2−1, is one missing point with no explosion around it. And unlike a vertical asymptote, a horizontal one can be crossed in the middle of the graph; it only describes what happens far out.
Think of it like this. An asymptote is like the horizon at the end of a long straight road: you keep getting closer and never arrive. The picture fits vertical asymptotes and far-out behaviour. Unlike the horizon, though, a graph can cross a horizontal asymptote in the middle before settling toward it.
Core idea. An asymptote is a straight line a graph approaches without bound as the input or output grows.