Another way of analyzing the data might be to see which station pairs have ridership between them above-and-beyond what their total station ridership would sugg

Another way of analyzing the data might be to see which station pairs have ridership between them above-and-beyond what their total station ridership would suggest. For instance, adjacent stations might be expected to have less ridership between them, because in many cases you could walk or bike between them.