Colley Matrix

The Colley Matrix is a computer-generated sports rating system designed by Dr. Wes Colley.

The site is one of more than 40 polls, rankings, and formulas recognized by the NCAA in its list of national champion major selections.[1] In 2018, the Mountain West Conference moved away from using the poll, along with three others, to determine the host site for its conference championship game in football.[2]

Methodology

In his initial paper at Princeton University, Colley states, "The method is based on very simple statistical principles, and uses only Div. I-A wins and losses as input — margin of victory does not matter. The scheme adjusts effectively for strength of schedule, in a way that is free of bias toward conference, tradition, or region."[3] Colley claims that his method is bias-free for estimating the ranking of a team given a particular schedule. The resulting values for each team are identified as a ranking, thus being a realization of Pierre-Simon Laplace’s Rule of Succession.

In a 2018 study released in the journal Science Advances, the ratings system finished third for highest accuracy of ranking synthetic data sets.[4]

National champions

As an NCAA-designated major selector, the NCAA regards the following teams as Colley's national champion selection.[5]

SeasonChampionRecordCoachRef.
2001Miami (FL)12–0Larry Coker[5]
2002Ohio State14–0Jim Tressel[5]
2003LSU13–1Nick Saban[5]
2004USC11–0aPete Carroll[5]
2005Texas13–0Mack Brown[5]
2006Florida13–1Urban Meyer[5]
2007LSU12–2Les Miles[6]
2008Florida13–1Urban Meyer (2)[6]
2009Alabama14–0Nick Saban (2)[6]
2010Auburn14–0Gene Chizik[6]
2011Oklahoma State12–1Mike Gundy[6]
2012Notre Dame12–1Brian Kelly[6]
2013Florida State14–0Jimbo Fisher[6]
2014Ohio State14–1Urban Meyer (3)[7]
2015Alabama14–1Nick Saban (3)[7]
2016Alabama14–1Nick Saban (4)[7]
2017UCF13–0Scott Frost[8][9][10]
2018Clemson15–0Dabo Swinney[11]
2019LSU15–0Ed Orgeron

a Record does not count wins against UCLA, or against Oklahoma in the BCS Championship game on January 4, 2005, as they were vacated by the NCAA.[12]

BCS Rankings

The Colley Matrix was one of the computer rankings used during Bowl Championship Series (BCS) system of determining national championship game participants from 2001-2014. Added in 2001, The Peter Wolfe and Wes Colley/Atlanta Journal-Constitution computer rankings were used in place of the NYT and Dunkel rankings. The change was made because the BCS wanted computer rankings that did not depend heavily on margin of victory.[13]

Criticism

The methodology of the rankings have been questioned by others on the grounds of subjectivity and specifics of the statistical math.[14][15]

Controversies

In the final BCS rankings of 2010, LSU was incorrectly ranked ahead of Boise State, at No. 10 instead of No. 11. The error was a result of Colley failing to input the Appalachian State/Western Illinois FCS playoff game correctly, a mistake that affected an order that helped determine bowl pairings that season.[16]

At the end of the 2012 season, the Colley Matrix ranked Notre Dame as first and Alabama second despite the Crimson Tide defeating the Fighting Irish 42–14 in the 2013 BCS National Championship Game.[17]

References
  1. "National champions: UCF Knights finish season ranked No. 1 in Colley Matrix" Orlando Sentinel. Retrieved 2018-09-01.
  2. "Mountain West announces new host procedure for conference football championship game" San Diego Union-Tribune. Retrieved 2018-09-01.
  3. "Colley’s Bias Free College Football Ranking Method: The Colley Matrix Explained" ColleyRankings.com. Retrieved 2018-09-01.
  4. "A physical model for efficient ranking in networks" Science Advances. Retrieved 2018-09-01.
  5. 2017 NCAA Football Bowl Subdivision Records (PDF). Indianapolis: The National Collegiate Athletic Association. July 2017. pp. 108, 115. Retrieved July 31, 2017.
  6. 2018 NCAA Football Bowl Subdivision Records (PDF). Indianapolis: The National Collegiate Athletic Association. July 2017. pp. 109, 115. Retrieved July 31, 2017. In years where a 'major selector' had a team other than the BCS champion as highest ranked team in its final poll that team is listed below the BCS Champion.
  7. 2018 NCAA Football Bowl Subdivision Records (PDF). Indianapolis: The National Collegiate Athletic Association. July 2017. pp. 109, 115. Retrieved July 31, 2017. In years where a 'major selector' had a team other than the CFP champion as highest ranked team in its final poll that team is listed below the CFP Champion.
  8. "National champions: UCF Knights finish season ranked No. 1 in Colley Matrix". orlandosentinel.com. Orlando Sentinel. January 9, 2018. Retrieved October 12, 2018. The Knights finished the 2017 football season as the No. 1 team in the Colley Bias Free Matrix Rankings, which were released late Tuesday.
  9. Seeley, Andy (January 10, 2018). "Knights Ranked No. 1 - UCF" (Press release). UCF Athletics. Archived from the original on March 22, 2018. Retrieved January 14, 2019.
  10. 2018 NCAA Football Bowl Subdivision Records (PDF). Indianapolis: National Collegiate Athletic Association. August 2018. Retrieved August 31, 2018.
  11. Wesley N. Colley. "Colley's Bias Free College Football Rankings: 2018 rankings, week 17". ColleyRankings.com. Retrieved January 9, 2019.
  12. "University of Southern California Public Interactions Report" (PDF). USA Today. NCAA. Retrieved May 5, 2015.
  13. "BCS formula still subject of debate". ESPN.com. September 16, 2002. Archived from the original on October 18, 2015. Retrieved May 1, 2009.
  14. "The problem with RPI, Elo, and the Colley Matrix" Jelly Juke. Retrieved 2018-09-01.
  15. "Statistics of Colley’s Ranking Methodology" Squared Statistics. Retrieved 2018-09-01.
  16. "The Computer Poll Uprising: Creators of the BCS's Most Controversial Component Look Back" Sports Illustrated. Retrieved 2018-09-01.
  17. "Notre Dame is still No. 1, according to one BCS computer" USA Today. Retrieved 2018-09-01.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.