A likely accident scenario is a drunk driver ramming another vehicle in the side. The drunk would be protected by the crumple zone of his car while his victim would be more likely to be injured.
Most crashes don’t happen quite that way, though. It isn’t a likely event in any case; merely, it’s a possible event.
One thing you have to think about in a collision that nature is the rotational effects. Car A hits car B in the driver’s door area. Well, now the driver’s side of B is moving towards the passenger side of A. So, you get a secondary collision of driver B with car A, while driver A is still distance from the area of impact. Double collisions of this nature are particularly deadly because you have the penetration into the driver’s compartment. After that, of course, you get the rotation, which puts the driver’s body in an odd position anyway. So, now you have those rotational forces and translational forces and the driver (if conscious) trying to equalize them. Then, bam, the second collision, which by this point is probably at least the third collision driver B has had with the side of his own car.
But it’s not all that common for the collision to happen exactly bumper to driver’s side door. They happen, of course, but quite rarely.
I have no statistics to cite, but when reading news articles about crashes due to drunk driving it often sounds as if the crash was head to head, or something close to it. The drunk driver “crosses the center line” and smashes into the other vehicle.
But about two years ago, in my town, there was one of those rare cases of bumper to diver’s side crashes. The victim was thrown from his car and killed, while the drunk suffered only minor injuries. What made it worse was that the victim probably would have survived (said the highway patrol) if he’d been wearing a seatbelt. I knew him slightly, the man was a fine guitarist, excellent musician.