The next visualization, figure 6.7, is an even more striking example of the kind of discovery only quantification can generate. Amid the flurry of references in every episode, I did not expect to see such a clear, positive trend line from season 2 to season 6, demonstrating a steady increase in the number of references used. Seasons 1 and 7 are outliers; perhaps season 1 relied more heavily on references as the writers were figuring out the show's tone and style, looking to make quick jokes or associations for purposes of characterization, while season 7 did not rely on references as much because the writers were focused on the characters and bringing the narrative to a satisfying conclusion. Regardless, the trajectory for the bulk of the series is clear, with a 50 percent increase over the course of seasons 2–6. The potential for synergy and speculation in this security grew over time as the complexity of its intertextuality expanded.

XREF: Connects to general observations about how series evolve their referential density and in-jokes as they develop an established audience.

Andrew deWaard, Derivative Media, loc. 376