First, this paper establishes an effective partition relationship in the finite integer set and an effective splitting relationship in the coalition set, and devises an EOCS (effective optimal coalition structure) algorithm, which only evaluates bipartite effective splittings of coalition to find the optimal value from bottom to top, so it decreases the number of bipartite splitting. Secondly, the correctness of the EOCS algorithm is proved based on the Kleene closure function. Moreover, this paper proves that the EOCS lower bound is Ω(2.818ⁿ) by the integration limit theorem, and discovers that the EOCS upper bound is O(2.983ⁿ) by the time serial analysis technique. Finally, this paper compares the EOCS algorithm with other algorithms to point out that the EOCS algorithm can find optimal coalition structure in O(2.983ⁿ) time whether the coalition values meet which probability distributions or not. The DP (dynamic programming) algorithm and the IDP (improved dynamic programming) algorithm proposed by Rothkopf and Rahwan can find an optimal solution in O(3ⁿ). The EOCS algorithm’s design, correctness proof, and time complexity analysis are all improvements of Rothkopf and Rahwan’s related work.