For poorly scripted robots, yes. For good scripted bots and humans, no. If you write a ssh attack tool, you can ping the port first I believe. Best thing is to always get your limits up so installing fail2ban is always useful.