Having a polished public implementation can lead to a massive increase in the number of citations a paper recieves, if it is really a useful system. Some of my

Having a polished public implementation can lead to a massive increase in the number of citations a paper recieves, if it is really a useful system. Some of my papers I think would have received far fewer citations if I had not released the code. Of course, if it is a really niche area with only a handful of researchers, this may not be true.