We consider a game theoretic model to study allocation of credit to authors; in particular, we find that alphabetical ordering can lead to higher research quality, while ordering by contribution results in a denser collaboration network and more publications.