[Edi]

I know. There's a discussion somewhere that discusses the relative merits ofa couple of equations, but I'm not going to try to comb through the forums to try and find it. As I recall, there was a fairly elegant solution to the problem by simply changing the equation a tiny little bit that would account for the discount in calculating dispel difficulty. As such, it could be fixed with relatively little effort as long as it's reported and if we can find that old discussion to save Johan some time in looking at and analyzing the code for why it behaves that way.