Let y1, y2, . . ., yn be an optimal sequence of 0/1 values for x1, x2, . . ., xn, respectively.
If y1 = 0, then y2, y3, . . ., yn must constitute an optimal sequence for KNAP(2,n,m). If it does not, then y1, y2, . . ., yn is not an optimal sequence for KNAP(l,n,m).
(The two solutions would be the same so if one is not optimal the other was not to begin with)
If y1 = 1, then y2, y3, . . ., yn must constitute an optimal sequence for KNAP(2,n,m - w1). If it isn't, then there is another 0/1 sequence z2, z3, . . ., zn such that
Again, the principal of optimality applies (the subsolutions KNAP(2,n,m) and KNAP(2,n,m - w1) were solutions to the subproblems).
(If this new solution is better we still have found a subproblem choice that is optimal given our first choice...we continue)
Hence, the sequence y1,z2, z3, . . ., zn is a sequence for the problem with greater value.