Despite well-established no-go theorems on a perfect linear optical Bell state analyzer, we find a numerical trend that appears to approach a near-perfect measurement if we incorporate eight or more un-entangled ancilla photons into our device. Following this trend, we begin a promising inductive approach to building an ideal optical Bell measurement device. In the process, we determine that any Bell state analyzer that (even occasionally) bunches all photons into only two of the output modes cannot perform an ideal measurement and we find a set of conditions on our linear optical circuit that prevent this outcome.