In the landscape, if there is to be any prospect of scientific prediction, it is crucial that there be states which are distinguished in some way. The obvious candidates are states which exhibit symmetries. Here we focus on states which exhibit discrete symmetries. Such states are rare, but one can speculate that they are cosmological attractors. We investigate the problem in model landscapes and cosmologies which capture some of the features of candidate flux landscapes. In non-supersymmetric theories we find no evidence that such states might be cosmologically favored. In supersymmetric theories, simple arguments suggest that states which exhibit R symmetries might be. Our considerations lead us to raise questions about some popular models of eternal inflation.