When a system includes too many possible blockings, selecting the minimal restrictive nonblocking solution as supervisor sometimes constitutes a conservative solution. Then relaxing the nonblocking condition becomes an inevitable fact to improve the overall performance. On the other hand, selecting the complete satisfying solution as supervisor may cause serious system failures due to some of these possible blockings. Then a supervisor linking these two solutions is needed. Therefore in this paper the balance between blocking and success is investigated employing an optimization approach. Firstly, we introduce a new performance measure which depends on numeric values obtained from strings that correspond to blocking and success. The proposed formulation captures the fundamental trade-off motivated by the classical optimization approach. Besides a new algorithm that explores the best result according to this performance measure is introduced.