I provide a full characterization of the set of absolutely expedient learning rules introduced in Börgers et al. [2]. The expected change in the expected payoff can be written as a quadratic form on the vector of relative expected payoffs of the strategies. This permits use of standard linear...