We calculated our rankings by measuring how tied attendance figures were to winning percentage since 1991, and we discounted expansion teams that have come along since. Through the use of multiple regression analysis, we determined how quickly fans supported the team when they started winning and how quickly they dissipated once performance slumped. The faster that fans boosted attendance and the more hastily they abandoned poor performance indicated fans who were less loyal.Maybe I'm just too taken with the popular stereotypes of certain fan bases, but I will say I was surprised with many of the rankings on both lists. So surprised, in fact, that I can't help but wonder if there's some monster flaw in the way they calculated it all.
We also controlled for new stadium construction and the boost it gives franchises, in order to avoid confusing the novelty of a new ball field for loyalty. Ticket prices were also controlled. If a team had to reduce prices during a bad season in order to maintain attendance, that's not loyalty. But if a team jacked up ticket prices and brought in even more fans, it's clear that the club has broad support.
I'm too dumb to figure that out -- I suspect that some intangible stuff like ownership changes and anomalous, pre-1991 hangovers affected it a bit -- but the statistically-oriented among you can find the full methodology here.