Understanding what prevents or enables the evolution of resistance to disease is critical for both understanding the natural world and also managing many modern societal problems, from agriculture to medicine. In previous work, we had shown, for what we believe to be the first time in an animal, that a cost of resistance (the price paid to be protected against a pathogen) meets critical assumptions in evolutionary theory. Here, we examined that same cost of resistance, which plays a part in preventing hosts from evolutionary ‘defeating’ their natural enemies, and showed that it doesn’t obey other common assumptions in evolutionary modelling. Using a laboratory model (grain / pantry moths) we showed that evolution along a resistance – development-time trade off, where the cost of resisting a virus is taking a longer time to grow, doesn’t behave symmetrically. Moths which were bred to take longer to grow were not more resistant to a virus than moths which were bred to grow faster. This is in stark contrasts to numerous studies showing that moths which are bred to be more resistant to a disease always take longer to grow. We frame our findings in the context of genetic underpinnings of complex pathways which lead to resistance, and conceptual understanding of competitive and non-competitive phenotype space (individuals cannot be best at everything, but they can be worst at everything).