Let's examine the following rational function: $f(x) = \frac{3x^3+2}{x^2-x-7}$.
Considering that the degree of the polynomial in the numerator is 1 greater than that of the denominator, it can be assumed that the function possesses no horizontal asymptote, but possess a slant, or oblique, asymptote.
As a result of long division, the slant asymptote appears to be $y = 3x + 3$. However, on a graph, the function $y = 3x + 3$ intersects with the original function, $f(x) = \frac{3x^3+2}{x^2-x-7}$, at the coordinate $(-.9583333..., .125)$.
Why is this the case? What condition prevents $y = 3x + 3$ from being a true asymptote of the function $f(x) = \frac{3x^3+2}{x^2-x-7}$ if long division produces a result declaring otherwise?

Notice that intersection of the function with the asymptote does not prevent it from being an asymptote. Broadly speaking, "asymptote" for $x$ to infinity, for example, means that the farest we go with $x$, the closest the function goes to the line, but it may intersect it an infinite number of times.
For example, consider $\frac{x^2+\sin(x)}{x}$ and look what happens.