In propositional domains using a separate test set via random sampling or cross validation is generally considered to be an unbiased estimator of true error. In multirelational domains previous work has already noted that linkage of objects may cause these procedures to be biased and has proposed corrected sampling procedures. However, as we show in this paper, the existing procedures only address one particular case of bias introduced by linkage. In this paper we therefore introduce generalized subgraph sampling, a sampling procedure based on bin packing, which ensures that test sets are properly chosen to match the probability of reencountering previously seen objects and which includes previous approaches as a special case. Experiments with data from the Internet Movie Database illustrate the performance of our algorithm.