Estimate how many binary low-density parity-check codes have self-orthogonal duals. [Note that we don't expect a huge number, since almost all low-density parity-check codes are ‘good', but a lowdensity parity-check code that contains its dual must be ‘bad']