I decided to release the code that I am using for my scatter plots. Mostly because I am not using it often enough to keep it for myself, so I hope that it is useful for somebody else. Use it at your own risk. Feel free to remove my name or change anything. And so on (we all know these proclamations...).
In my opinion, the output is pretty cool, especially if you combine it with Inkscape or Illustrator (which works well if you save the plots as pdf)
As an example, take the plot I have here: http://sportstribution.blogspot.de/2014/03/three-quick-points-on-kyle-korver-and.html
You just have to use the same axes settings and different filter or data points and you get two plots that you can easily merge (and color differently).
Especially for those of you that have the data readily at hand (I'm looking at you, Nylon guys!) it could be worth to give it a try.
This code was my first try in learning Python and it is therefore far from perfect. It is only 400 lines and not so many packages, so it might also be interesting for those of you that are as well interested in learning Python.
The code should work with Python2.7 and the right packages. If someone knows how to update this, I am glad to listen.
Oh, I almost forgot to add the github: https://github.com/SportsTribution/SportsTribution-Scatter.git