EFI Documentation              
                     
    8/27/2006 16:48                
                     
                     
    The EFI rankings will return to the simplified sequential version this year.      
                     
    The ratings are mathematically very simple. I won't bore you with the details other than to say  
    the EFI index is essentially a "power rating" system which is heavily driven by scoring margins.  
                     
    The system is primarily designed to predict the outcomes of future games and will    
    generally be more volatile than many other ranking systems which are more mathematically  
    sophisticated and which are more "retrodictive" (as opposed to "predictive") in nature.  
                     
    Forecasts of future games are very simply calculated by taking the direct difference    
    between the EFI ratings for any two given opponents.        
                     
    The entire system is contained and executed within a series of EXCEL spreadsheets.  
                     
    **************************************************************************************************************************
                     
    I would like to acknowledge Peter Wolfe for use of his historical game scores, Ken Massey for graciously
    including my rankings in his College Football Ranking Comparison listings, and David Wilson  
    for his very thorough and informative American College Football WWW Virtual Library website.  
                     
    **************************************************************************************************************************
                     
    Who am I and why should I have any credibility at this football ranking stuff? I have an enginneering degree
    (Tennessee Tech, 1972), with a major in engineering science and minors in mathematics, physics, and 
    engineering mechanics. I worked for 31 years in the telecommunications industry, most of which was
    spent in demand analysis (forecasting demand for telecommunications products and services using
    quantitative techniques such as time series analysis and multiple regression, etc.). My whole academic
    and professional life has pretty much been involved with numbers. I have followed college football
    very closely since my pre-teen years and spend much time keeping up with the game using  
    football magazines/annuals, television, radio, and especially internet sources. I developed my first rating
    system when I was in high school in 1966 (using pencil and paper only), and have been trying different
    ratings methods for many years. I love the numbers aspect of college football, but am also a  
    devoted fan of the game. My favorite teams are Tennessee Tech (Div IAA, Ohio Valley Conference,
    my alma mater), the Tennessee Volunteers (you can't be from Tennessee and not follow the Vols), and - 
    believe it or not - the Duke Blue Devils (I've lived in Durham NC for more than 20 years and, despite
    the dismal track record of Blue Devil football over the last quarter century or so, I've gradually become
    a fan of the home town team since it gives me the opportunity to get away from the numbers and actually 
    watch games in the stadium).