 ZPP

$\mathbf{ZPP}=\mathbf{RP}\cap \mathbf{co}\text{-}\mathbf{RP}$. Equivalently, it consists of problems having an always-correct randomized algorithm with polynomial expected running time.