The benefits of "bet-hedging" strategies have been assumed to be the main cause of phenotypic diversity in biological populations. However, in their recent work, Healey et al (2016) provide experimental support for negative frequency-dependent selection (NFDS) as an alternative driving force of diversity. NFDS favors rare phenotypes over common ones, resulting in an evolutionarily stable mixture of phenotypes that is not necessarily optimal for population growth.