Shadow mapping and Closest Percentage Filtering

Hi. This may be a silly question, so at the risk of making an idiot of myself, I’ll ask away.

I’m currently reading about shadow maps. Now shadow maps have aliasing problems and thus produce jagged shadow edges. I can see this happening when you have shadow maps of a fairly low resolution or when you only use a small portion of the shadow map to shadow an entire image. Thus a lot of pixels from the image plane will map to the same pixel in the shadow map. This generates the jagged edges.
What I don’t understand is that a lot of papers state that Reeves Percentage Closer Filtering can help this problem. I understand how the filtering is done and that it should be untertaken whenever a pixel from the image plane covers several pixels in the shadow map, but when will this actually happen? If the shadow map is low res, then how can a tranformed pixel cover several pixels in the map?
Have I gotten the concept completely wrong? Is aliasing and the problem with the low res maps the same at all? Would anyone care to explain this in detail please?

Thanks in advance Anders

The percentage closer filtering can work just like linear interpolation, except what you interpolate is the post-test color value for each of the four depth samples, rather than the depth samples themselves. This will make edges of the shadow blurry, which may reduce jaggies.