On Quantifying the Topsy-Turvyness of a College
Football Season: Introduction

Autar Kaw
December 21, 2007
Introduction to TT factor                                                              Page   1   2   3   4  
    
Throughout the 2007 college football season, media and fans alike cannot recall a
more topsy-turvy season.  

In just week#1, No. 5 Michigan lost to unranked Division II Appalachian State and
was booted out of the AP Top 25.  In week#3, No. 11 UCLA not only lost to
unranked Utah but also by a wide 44-6 margin.  UCLA was also dropped from the
AP25 rankings.  

And then came the curse of being ranked #2.  University of South Florida, where I
work, was ranked #2 at the end of Week 7 and lost to unranked Rutgers 30-27 in
Week#8.  This was the same week when four of the top ten AP poll teams lost
their games.

Although everyone can see the anecdotal evidence of a topsy-turvy season, is it
possible that the media and fans over exaggerated the topsy-turvyness.  Were
there other seasons that were more topsy-turvy than 2007 and we did not
notice?

To answer this question scientifically, Professor Autar Kaw of the University of
South Florida has developed a simple algorithm to quantify the topsy-turvyness
from week to week and that of the season.  

The quantity is called the topsy-turvy factor (TT factor).  The TT factor is calculated
for each week (called Week TT factor); then the Week TT factors are used to
calculate the Season TT factor.

Using the AP football rankings from the last six football seasons, Professor
Kaw confirms quantitatively that 2007 has been a topsy-turvy season, but its
topsy-turvyness is NOT too different from that of the 2003 season.

Next: Comparisons of TT factors for 2002 thru 2007 Seasons