I found this, I think it gives kinda of an explanation: If the regression of log ((A'/A) -1) on -log B is linear with a slope of -1, then this indicates that the antagonism is competitive and by definition the agonist and antagonist act at the SAME recognition sites. So basically the more you add Y, the less X is bound, which means they have a same structural component (epitope) and must act on the same site? Don't know if this makes sense...

I found this, I think it gives kinda of an explanation: If the regression of log ((A'/A) -1) on -log B is linear with a slope of -1, then this indicates that the antagonism is competitive and by definition the agonist and antagonist act at the

SAME recognition sites. So basically the more you add Y, the less X is bound, which means they have a same structural component (epitope) and must act on the same site? Don't know if this makes sense...http://facpub.stjohns.edu/~yoburnb/pages/dictimages/schild1.html