As this case and others (c.f., GoGaRuCo, which uses the same method) shows, the natural gender split is 50/50. When the submissions pool is 50/50, and they do a blind judging of the submissions, they end up with a 50/50 split in speakers. In other words, they keep demonstrating that there is no essential difference in technology based on gender. When they hide from themselves the gender of the speaker, they end up choosing equally from men and women.
