This paper introduces a multi-hypothesis temporal prediction technique that combines two motion vectors (MVs) derived respectively from template and block matching for overlapped block motion compensation (OBMC). It achieves similar prediction performance to bi-prediction while only one MV has to be sent. Based on two signal models, the template MV is shown to approximate the pixel true motion around the template centroid. We then find another MV to best complement the template MV from both deterministic and statistical viewpoints, the latter leading to the search of its optimal sampling location in the motion field. The result is a search criterion with OBMC window functions forming a geometry-like motion partitioning. To compromise between performance and complexity, generalizations to adaptive template design, multi-hypothesis prediction and motion merging are made. Extensive experiments conducted with the HM-3.0 software confirm the effectiveness of the proposed schemes.