The electromagnetic total power radiated from the plasma in a tokamak is often estimated by a weighted summation of some bolometer channels. Although this method may be reliable for limiter plasmas, theoretical considerations and a comparison with total radiated powers determined from tomographic reconstructions show that for diverted plasmas with much radiation from the divertor region such estimates can be unreliable, unless the used lines of sight are parallel. The sparseness of the sampling, however, has little effect on the total radiated power calculation if the beam widths are similar to the distance between lines of sight.