People have frequently noted some connection between counterfactuals and causation. While it seems backwards to suggest that causation should be defined in terms of counterfactuals, it is reasonable to connect the two concepts, and explains why some counterfactuals are more reasonable than others, as we noted in the last post.
For example, “If I dropped this cup, it would fall to the floor,” is more reasonable than “If I dropped this cup, it would fly up to the moon,” because we are considering the operation of causes like gravity which could cause falling to the floor, and which could not cause (merely by dropping) an object to fly to the moon. In particular, since causes eliminate alternatives, they give us a reason to say “this would have happened rather than that.”
Nonetheless, we cannot get any sort of absolute determination out of this. One would attempt to get a determinate outcome by specifying the counterfactual as clearly as possible: “If I dropped this cup, and everything else was the same.” The “nearest possible world” idea is trying to get at this. However, this is not in fact completely determinate because “everything else” can’t be entirely the same, and what else needs to change is not determinate. In order to drop the cup, there would need to be a course of events that led up to the dropping, and there are many different courses that could have done that. The same thing will happen if you to specify exactly what led to the dropping of the cup; there will need to be something that led to your specification. Thus, at the very least, you will not typically be able to get absolute determination in this way.
Naturally, there is nothing to prevent us from coming up particular examples where we can get complete determination by using something which is always true anyway, or by using logical implication from the counterfactual, e.g. “If I dropped this cup, 2 and 2 would still be 4,” or “If I dropped the cup, it would have been dropped.” But these are not typical cases.