THE INTRODUCTION

The Phelan Power Principle (PPP) is a computer algorithm used to determine the ranking of each team in the National Collegiate Athletic Association's (NCAA's) Football Bowl Subdivision (FBS).  The formula uses a series of components to calculate each FBS team's relative strength (Team Score).  Teams are ranked based on their Team Score in a lowest-first ascending model. 

THE ANNIVERSARY

This season marks the THIRTIETH anniversary - THREE DECADES - for the PPP!  From its' inception WAY back in August 1996, the mission of the PPP has been to create the most accurate computer ranking system of the major college football programs (back then, known as "Division I-A", but today more familiarly referred to as the "Football Bowl Subdivision (FBS)").  Since then, the PPP has worked diligently in striving to become a nationally-recognized and accepted poll, from its' humble beginnings on a locally-hosted bulletin board (1996 - 2000), to appearing online for a reputable newspaper publication (2001 - 2004), to being self-hosted on its' own domain (2005 - 2018), to being available thanks to the services of Google Sites (2019 - present)!  During this time, college football has seen MAJOR evolution in its' landscape - from the Bowl Alliance, through the Bowl Championship Series, to now already TWO iterations of the current College Football Playoff (CFP) - all while seeing conferences begin, end, realign, and merge, along with the number of teams in the highest-level of NCAA competition continuing to increase.

The PPP has also evolved, with the same purpose in mind: to refine (and hopefully improve!) the method of crowning an FBS National Champion.  As I reflect on the past thirty years, it has been an amazing journey in developing and growing this ranking system, and I am extremely proud of the advancements and success this poll has experienced, and I'm hopeful - GOD-willing - that the NEXT thirty years will be just as successful!  But without the following that has grown over the years, this wouldn't have been possible...

So for everyone who has visited, viewed, commented and/or offered insight, input, recommendations & advice over the years on the PPP, I sincerely offer you a HEARTFELT THANK YOU!!  Your time, assistance, support, and guidance has been IMMENSELY appreciated, and has been instrumental in the PPP's growth and continued existence!  For that I am FOREVER grateful and indebted!

THE HISTORY

The idea for the PPP came during my first semester in graduate school in the Fall of 1996.  As an avid football fan since childhood, I was always interested and perplexed at how major college football would determine its' "national champion" through the use of human polls in order to rank teams, as opposed to the more "traditional" method of a play-off system.  As I began studying the polls, I couldn't help but notice how rankings seemed to change on a whim... whether it be from one team winning convincingly over another, or winning by a margin that so-called "experts" felt was "too close".  It was through this thinking that I learned the meaning and reasoning behind terms such as "running up the score", "ugly win" and even "pretty loss".

At the time the Internet was just beginning to come to the forefront in terms of "on-demand information", so previously-existing computer polls did not garner the attention they do today.  Watching the polls made me begin to wonder:  what kind of results would a computer generate in comparison to humans, when only raw data was entered... leaving out the biasedness of pre-season perceptions of how humans felt a given team would perform that season?  Thus was born the PPP ranking system.  Looking back now, 1996 turned out to be a perfect year to begin the ranking system, as that year would mark the first year of overtime in college football, thus eliminating the extra step of dealing with ties in game outcomes.  With the guarantee of a winner and loser in each game, the rankings would be a more accurate representation of which team is truly "better" than another. 

THE POLL

The PPP rankings are conducted weekly during the FBS regular season, along with after the first-round CFP playoff, bowl & CFP quarterfinal (when all are completed), and CFP semi-final games.  In addition to these rankings, a Preseason Poll two weeks prior to the start of the season and a Final Poll after the conclusion of all games for the season are also conducted.  The Preseason Poll is designed to act only as a baseline starting point for each team.  The poll does not take into account issues relating to recruiting, transfers/departures, injuries, coaching changes, etc.; it is strictly based on the final record results from the previous year, while applying the current year's scheduled opponents and their corresponding final record results, in order to create an initial starting point for each team.  Once this has been completed, all teams are dis-joined from each other, and each team's corresponding record is reset to 0-0.  Once the season begins, teams will be joined as they begin playing each other.  As the season progresses, each team's Team Score will become not only a reflection of that team's strength individually, but also of that team's strength in relation to the other FBS teams in the poll.

THE FORMULA

The formula uses an "Increasing Strength" model in its' weekly calculations.  Specifically, some formula components are removed or restricted early in the season, when teams have not played many opponents and are not joined to many other teams in the system; as the season progresses, teams will become joined to other teams and their corresponding opponents.  At different intervals during the season, the effect of some formula components will lessen as other components increase, in order to provide a normalized Team Score based on the relative competition for all teams at that point of the season.

Each week teams are given the same base score to begin with.  Then the current week's results are factored in and then linked to all previous results for the season to-date.  Filters exist to prevent a team from making extraordinarily large gains or losses based on the results of a single game.  What this means is that a team's Score and corresponding Ranking are designed to be a measure of the "entire body of work to-date" for a team's season, as opposed to the "hear and now awe" that sometimes exists on the human side, when for example one team convincingly overwhelms what is considered by humans to be a good/great opponent, and in-turn makes a significant climb in the human polls.  I liken the formula to another popular American sport:  auto-racing (NASCAR in-particular), where a driver with one win and ten other Top Five finishes may rank higher than a driver with five wins and no other Top Five finishes.

THE BCS COMPUTERS COMPARISON (1998-2013)

From 1998-2013, the Bowl Championship Series (BCS) had been the mainstay in determining the major college football "national champion".  As the BCS grew and adapted, computer rankings were added to and removed from its' ranking system, each unique in their algorithm for calculating rankings for the Division 1-A/FBS teams.  As a litmus test of sorts, the PPP was compared with each of the BCS computers for each week the BCS rankings were calculated during a season, as a way of measuring whether the calculated results in the PPP rankings correlated closely with those of the expert formulas used for the BCS; in each comparison, the PPP rankings were highlighted when the results were either lower or higher than those of the approved BCS computer models.  After sixteen seasons of comparisons, the PPP demonstrated consistently that not only were its' overall average rankings approximate to those of its' BCS counterparts, but in multiple cases the PPP excelled in more accurately determining the Division 1-A/FBS "national champion" prior to the start of the bowl season, as evidenced by its' top-ranked team in the final regular-season polls during this time frame.

THE COLLEGE FOOTBALL PLAYOFF (2014-2023)

Beginning in the 2014 season, FBS college football fans were given what they have so greatly wanted and longed for:  a "play-off" at the NCAA's highest-level of competition! Gone were weekly computer and human polls in determining the top two teams to compete for the "national championship" (although they both still "vote" for a champion at season's end) - replaced instead by a panel of experts tasked with determining the top four teams at the end of the regular-season to compete in a two-round single-elimination bracket in an attempt to determine a "true" FBS National Champion! This change has provided a unique opportunity for the PPP:  a chance to itself grow and further develop its' own unique brand and identity. To that end, the PPP has welcomed this new era of college football by beginning to include an entire new slate of components - in addition to those that have been actively in-use since its' inception - to provide team evaluations at a more granular level than has ever previously been considered and implemented!

THE EXPANDED COLLEGE FOOTBALL PLAYOFF (2024- )

Beginning with the 2024 season, the College Football Playoff (CFP) has been expanded from four to twelve teams! With this new change, the playoffs have expanded from two to four rounds, with the four highest-ranked (by the same panel of experts enacted in 2014) teams earning first-round byes, while the teams ranked fifth through twelfth compete with the higher seed hosting the lower (i.e. 12 @ 5, 11 @ 6, 10 @ 7, and 9 @ 8) at their home stadium. The Quarterfinals will occur during bowl season, the Semifinals when the previous championship game would occur, and finally the Championship game just over a week later.  As with the initial playoff iteration, this change has provided a chance to review how the PPP has fared during the CFP era; specifically, in its' ability to match the four teams in the CFP rankings. In reviewing the final CFP Rankings for the original four-team model, the PPP managed to have the same four teams in 33/40 (82.5%) top four ranking spots; of those 33, 17 were exact ranking matches - highlighted by perfectly matching all four teams for the 2016 and 2022 seasons. Not surprisingly, the worst season for matching was 2020, where the PPP only matched one of the four CFP teams (although that one was a perfect match to the CFP's spot!).

THE RESULTS

Despite being one of the "younger" computer models for calculating rankings for the NCAA FBS, the PPP has proven to be an accurate and successful college football ranking system.  The results - while unique and independent of any other human or computer input - have consistently fared favorably when compared to other well-respected computer models that have been in existence for much longer.  As such, the PPP will continue to build its' own unique legacy with each new season, with the goal being that in the future it too will be mentioned in the same breath as those other respected models. Moreover, it is my hope that as the major college football landscape continues to grow and adapt to future environments, so too will the Phelan Power Principle!

All Site Content © 1996-2025 Adaronius Enterprises.  All Rights Reserved.