A new study comparing multiple and single arterial grafting highlights how physician preference can bias observational data.