Would the publication of anonymous referee reports and editorial decision letters of published papers benefit the scientific debate? Results from a trial seem to suggest this.