We consider approaches for optimally organizing competitive sports leagues in light of competitive and logistical considerations. A common objective is to assign teams to divisions so that intradivisional travel is minimized. We present a bilinear programming formulation based on k-way equipartitioning, and show how this formulation can be extended to account for additional constraints and objectives. We show that our formulation and extensions can be solved directly using modern solvers. We present computational results for all major North American professional sports leagues.