Threads, Loops, and Upsets

 In GridMap, when two teams have played each other, they are conneced by what GridMap calls a "thread". A thread is shown on GridMap as an arrow pointing from the winning team to the losing team. Eventually, teams may become connected in a series of continuous threads. For example when team A defeats team B, and team B defeats team C, team A and C are now linked on the continuing thread from A, through B, to C. If team A later defeats C, no new thread connecting A directly to C is needed, since A is already connected to C through B. If however, as occasionally happens, team C defeats team A, this creates what GridMap calls a "loop": A > B > C > A. Any positioning (or ranking) of these teams (other than tied), will result perforce in a lower positioned/ranked team having defeated a higher positioned/ranked team. The process of determining where to "cut" a loop, and therefore which team to place at the top of the resultant thread, is what GridMap calls "resolving" the loop. Once a loop has been resolved, the game between the team at the bottom of the thread who defeated the team at the top of the thread is what GridMap calls an "upset", often referred to in other systems as a ranking violaton. The fewer the number of upsets, or ranking violations in a system, the more game results the system corretly reflects, and the more accurate the system can be considered to be. The 3 team loop above is a simplistic example that results in just one upset no matter where the loop is cut, or resolved. But many loops, due to the broad interconnectivity of college football schedules, are much more complicated, often involving ten or more teams. GridMap has developed a set of principles and logical rules that, even with these more complicated loops, has successfully minimized upsets, and resulted in one of the two most accurate sytems in all of college football.