This preprint documents the conceptual and methodological framework of a comparative protocol for analyzing human translation and algorithmic output in translator training. The proposed approach focuses on the comparison of event configurations in discourse across languages and translation modalities. The protocol, illustrated through Russian–Italian translation examples, is intended as a transferable methodological tool for studying structural tendencies in both human translation and algorithmic output.