You're confusing two different types of angular momentum. To see the difference, consider a classical system of planets orbiting around a sun. Orbital angular momentum is associated with the motions of the planets going around the sun in their orbital paths. Spin angular momentum is associated with the rotation of a planet around its own axis. Two planets can follow the same orbital path, in the same direction, at different locations on the path of course, but spin around their axes in opposite directions. The fact that they don't collide with each other doesn't have anything to do with those spins being in opposite directions, right?
In a quantum-mechanical system like electrons "orbiting" around an atomic nucleus, an electron doesn't follow a nice classical-type orbit. Instead, we have an orbital wave function which gives us a probability distribution for where the electron might turn out to be if we were to measure its position somehow. But we can still associate an "orbital" angular momentum with this orbital wave function.
Similarly, the electron is not literally a tiny ball that spins around its own axis, but we can still associate a "spin" angular momentum that has a fixed, unvarying magnitude, but can be oriented in different directions. In the same atom, we can have two electrons with the same orbital wave function and orbital angular momentum, but different orientations for the "spin" angular momentum.
(Aha, I see DrC snuck in while I was making a sandwich.)