YES(?,O(n^1))
* Step 1: PolyRank WORST_CASE(?,O(n^1))
    + Considered Problem:
        Rules:
          0. start(A,B,C) -> eval(A,B,C)          True                 (1,1)
          1. eval(A,B,C)  -> eval(C,-1 + B,1 + A) [100 >= A && B >= C] (?,1)
        Signature:
          {(eval,3);(start,3)}
        Flow Graph:
          [0->{1},1->{1}]
        
    + Applied Processor:
        PolyRank {useFarkas = True, withSizebounds = [], shape = Linear}
    + Details:
        We apply a polynomial interpretation of shape linear:
           p(eval) = 101 + -1*x1 + x2 + -1*x3
          p(start) = 101 + -1*x1 + x2 + -1*x3
        
        Following rules are strictly oriented:
        [100 >= A && B >= C] ==>                      
                 eval(A,B,C)   = 101 + -1*A + B + -1*C
                               > 99 + -1*A + B + -1*C 
                               = eval(C,-1 + B,1 + A) 
        
        
        Following rules are weakly oriented:
                  True ==>                      
          start(A,B,C)   = 101 + -1*A + B + -1*C
                        >= 101 + -1*A + B + -1*C
                         = eval(A,B,C)          
        
        
* Step 2: KnowledgePropagation WORST_CASE(?,O(n^1))
    + Considered Problem:
        Rules:
          0. start(A,B,C) -> eval(A,B,C)          True                 (1,1)              
          1. eval(A,B,C)  -> eval(C,-1 + B,1 + A) [100 >= A && B >= C] (101 + A + B + C,1)
        Signature:
          {(eval,3);(start,3)}
        Flow Graph:
          [0->{1},1->{1}]
        
    + Applied Processor:
        KnowledgePropagation
    + Details:
        The problem is already solved.

YES(?,O(n^1))