I think there's a bug in the down filtering in scrapr-data.R? Your blog post has 3rd down Romo to Bryant but that's a 2nd down play in the data?