In my case they're not scraping a single account, they're scraping half of twitter via my instance, such simple regex options have never worked for me. One of the approaches brought up was to ban anything that wasn't containing a referrer from the site.