I would not use **shade**.  Shade has no conceptual mapping to a score and would be more difficult to distinguish from one shade to the next -- I'm assuming you're using a shade of the same color (e.g., green) vs. different colors.  Different colors would still have no visual mapping to "score".

Here is an example:

![enter image description here][1]

(note: I've removed some specifics from the image)

This sounds like what you mean by "size of the circle".  This chart gives me an idea of "how many to-do items" and a "pain score" on the axis.  The size of the circle reflects how popular the category is.

I can quickly gather that Orange is very "painful", has a lot of "to-do items" associated with it, and by the size I can tell that this area is pretty popular.  Alternatively I can see that Blue is not as "painful" as "Red", but is more popular (I can explore to find out why)!


  [1]: https://i.sstatic.net/yLi8V.png