Hello All. I am trying to prove that this constrained maximization problem is correct for this warfare model, and I cannot for the life of me solve it. The fucking authors say the solution can be easily found. See pic related. The Bi's are just constants. If you use lagrange multiplier you get every vector component of the gradient to be
x_i/[(x_i)^2-(B_1)^2)]^(1/2) = lambda.

How the fuck do I solve this problem? Been working on it for hours. Fuck this author for saying it is easy.