Before making conclusions, this graph needs to be controlled for at least one outside factor: the overall geographic distribution of companies. If one were to

Before making conclusions, this graph needs to be controlled for at least one outside factor: the overall geographic distribution of companies. If one were to take the complete bipartite graph of [All Startups, All Fortune 500s] and plot the distances between connected vertices, I would not be surprised to find a distribution graph similar to the one shown above. Most companies are in population centers, and most population centers are on the coasts. I have no reason to doubt that the subset of that graph that contains acquisition relationships is, in fact, more heavily skewed towards proximity. But, that is currently a conjecture, and the data does not support the conclusion (although it does not disprove it, which is something).