There are several things they could check and factor into the score before banning someone. Client source (if the use the website 100% of the time and suddenly

There are several things they could check and factor into the score before banning someone. Client source (if the use the website 100% of the time and suddenly this was from the Android client for example), IP address (do they tweet from the U.S. exclusively and then suddenly they're tweeting from Moscow?), VPN affiliation (did this tweet originate from a known VPN egress?), and so on. These things _should_ be factored into the "omg ban this account" score IMO, but I have no idea if they are.