This doesn’t imply that evidential and non-evidential success are opposed in general; just that whatever shape memespace has, it will have a convex hull that can be drawn across this border.
Is there any mathematical reason why the hull can’t be concave?
The only applicable mathematical meaning for the word “hull” that I know of here refers to the convex hull. You can draw a variety of non-convex figures that enclose a set of points of course, but I’ve not heard those referred to as “hulls”.
I’m not trying to be a smartass about the word hull; I’m just curious to know if there is a good mathematical reason why the shape of the boundary you mention in the post would necessarily be convex.
I’m sorry, I’m not following you. The hull is convex by definition, no matter where the points are.
Thinking about it though, the appropriate figure to consider isn’t the convex hull, but the set of points which are not dominated by any other points. That can produce a concave figure, but it’s still true to say that when you switch between them, you have to lose on one axis to gain on another, again by definition.
Is there any mathematical reason why the hull can’t be concave?
The only applicable mathematical meaning for the word “hull” that I know of here refers to the convex hull. You can draw a variety of non-convex figures that enclose a set of points of course, but I’ve not heard those referred to as “hulls”.
I’m not trying to be a smartass about the word hull; I’m just curious to know if there is a good mathematical reason why the shape of the boundary you mention in the post would necessarily be convex.
I’m sorry, I’m not following you. The hull is convex by definition, no matter where the points are.
Thinking about it though, the appropriate figure to consider isn’t the convex hull, but the set of points which are not dominated by any other points. That can produce a concave figure, but it’s still true to say that when you switch between them, you have to lose on one axis to gain on another, again by definition.
That answers my question.