In part 1, I introduced how we can reason from graphs, why they're so useful, metrics for analyzing and condensing large information, and more. In part 2, I took a look at the CryptoPunks trading network to introduce a higher level of reasoning of graphs -- random worlds and diffusion models. I then took a little bit of a tangent to discuss how we can use Network and Graph Analysis to look at NBA games. This part will use concepts introduced in that story to further my analysis of Graph Machine Learning. I highly recommend reviewing these previous parts before diving into this one as they set this one up well, and many concepts I won't dive into here are already discussed and shown in each of those.