A multiploid genotype comprises a number of chromosomes, and a ‘dominance’ mechanism underlying its interpretation. The simplest dominance mechanism uses a ‘mask’ chromosome, genes in which determine which gene from which chromosome is expressed at each locus. Previous work using contrived test problems showed that a multiploid was sometimes better than a normal genetic algorithm (GA), and sometimes not. Multiploidy seemed particularly helpful in cases where a normal GA would be likely to irretrievably lose important genetic material. Here we continute this investigation in the context of more realistic problems: the multiple knapsack problem, and the set covering problem. There are many complex effects, but experiments tend overall to reinforce the above suggestion.