Friday, December 27, 2013

FBS 2013: The Drummond Rankings

Anyone familiar with college football knows there has never yet been an actual national championship worthy of the name.  Up to now, three general systems have been used, all with serious flaws.  Originally, football teams competed in their regions and there was no real consensus about who was best at the end of the year.  The bowls set up alignments by conference and the press polls named their favorite as champion, but this was always unofficial, since in many years the acknowledged contenders never played each other. At every other level, from pee wee football through the NFL, a playoff determines the champion, but for reasons of greed and arrogance the NCAA refused to set up a valid for Division I-A, later called the FBS, teams.   Instead, a corrupt system called the Bowl Championship Series (BCS) was imposed, which was no ‘series’ at all, since two hand-picked teams got to play for the championship according to a subjective system which seemed to favor large state schools, and every other team was left out, regardless of how well they played.  The new ‘playoff’ proposed to begin next year is simply that same corruption extended to a four-team clique, as evidenced by the total lack of published criteria about how a team qualifies for the playoff – a committee with no published qualification criteria will announce the four teams on whatever basis they liked, including sheer favoritism.  The new system is no better than the old, and deserves no support from the fans, whose demand for a legitimate playoff system has been long, clear, and yet ignored by the NCAA.

As I wrote before, every other level of football, from the beginners to the professional leagues, uses a playoff system and has clear metrics on just how a team qualifies.  Ultimately, every valid system is simple in its construction, and two elements decide the teams – record, and opponents’ record (or Strength of Schedule).  I have no official standing with the NCAA, but back in 1980 I sent DeLoss Dodds (who was the NCAA’s point of contact for the playoff debate back then) a simple proposal to use the existing system to have thirty-two teams playoff to decide the championship, using sixteen bowls.  Dodds sent back a form letter which basically ignored the idea, which has been the NCAA’s consistent and contemptible treatment of their responsibility for decades.  I recognize now that there are reasons to limit a playoff (mostly due to the costs and difficulty for fans to attend five weeks of games away from their hometown.  At the least, however, an eight-team system can be established, perhaps sixteen if the first round follows the FCS model and allows the higher-ranked teams to play at home early on.  The key point is finding an equitable way to rank the teams for qualification and standing.  And I believe the system I created in 1980 is still the most reasonable.

For each of the FBS teams, I use a point system as follows:  award ten points for each win, and take away five points for each loss, regardless of the opponent.  Then add additional points according to wins by teams the FBS team has defeated; for each team the FBS school defeats, award an additional point for each win that team has over an FBS team.  Better record is tie-breaker for teams with same point totals. This system is easy to work, awards points for wins over good teams, and does not reward teams for either easy schedules or losing to teams with good records.

Accordingly, my system applies to the 2013 season and creates the following rankings (with BCS rankings in parentheses):

  1. Florida State 13-0, 190 pts  (#1 BCS)
  2. Auburn 12-1, 183 pts   (#2 BCS)
  3. Ohio State 12-1, 176 pts  (#7 BCS)
  4. Stanford 11-2, 169 pts  (#5 BCS)
  5. Michigan State 12-1, 163 pts  (#4 BCS)
  6. Baylor 11-1, 159 pts  (#6 BCS)
  7. Northern Illinois 12-1, 153 pts  (#23 BCS)
  8. UCF  11-1, 153 pts  (#15 BCS)
  9. Missouri 11-2, 152 pts  (#8 BCS)
  10. Alabama 11-1, 151 pts  (#3 BCS)
  11. Fresno State 11-1, 149 pts  (#20 BCS)
  12. South Carolina 10-2, 143 pts  (#9 BCS)
  13. Louisville 11-1, 142 pts  (#18 BCS)
  14. Oklahoma State 10-2, 141 pts (#13 BCS)
  15. Arizona State 10-3, 141 pts (#14 BCS)
  16. Oklahoma 10-2, 139 pts  (#11 BCS)
  17. Oregon 10-2, 128 pts  (#10 BCS)
  18. Clemson 10-2, 125 pts  (#12 BCS)
  19. Duke 10-3, 125 pts  (#24 BCS)
  20. Rice 10-3, 124 pts  (BCS unranked)
  21. Bowling Green 10-3, 121 pts  (BCS unranked)
  22. UCLA 9-3, 114 pts  (#17 BCS)
  23. Ball State  10-2, 113 pts  (BCS unranked)
  24. USC 9-4, 113 pts  (#25 BCS)
  25. LSU 9-3, 111 pts  (#16 BCS)
  26. Wisconsin 9-3, 108 pts  (BCS unranked)
  27. Miami (Fl) 9-3, 106 pts  (BCS unranked)
  28. Notre Dame 8-4, 106 pts  (BCS unranked)
  29. BYU 8-4, 104 pts  (BCS unranked)
  30. East Carolina 9-3, 101 pts  (BCS unranked)
  31. Virginia Tech 8-4, 100 pts  (BCS unranked)
  32. Georgia 8-4, 99 pts (#22 BCS)  
   35Texas A&M 8-4, 96 pts (#21 BCS)

If all we want is to pit #1 vs #2, my system is in agreement with the BCS and the two human polls, and in fact also with the six computer ranking systems the BCS uses for its ranks.  But if we extend to a simple four-team playoff, things change.  BCS darling Alabama gets 3rd place in their scoring, but based on actual wins and how their opponents did they only rate 10th place.   My third and fourth teams, Ohio State and Stanford, only get seventh and fifth place in the BCS, in large part because they won early and lost late, while teams which lost early and won late got extra support in the BCS system.  I rolled the numbers all the way thorough #32 to show how the BCS starts to fall apart as we get further along; UCF and Northern Illinois did a better job than the BCS granted them, and there is an increasing disparity between the polls and objective results as we look at teams in the two and three-loss range. 

So how would my system plug in a playoff?  As a compromise to the logistics and the networks, an eight-team playoff is both functionally feasible and worth calling a playoff.  I’d take the, what, thirty-nine existing bowls and re-assign teams according to a simple three-level system of designation:

The top eight teams are in the playoffs.  They play at the top four locations, and each week the top surviving teams stay where they are and host the others.  While some fans won’t be able to afford making up to two additional trips to see their school, the rising interest in a bonafide NCAA FBS playoff will ensure seats get filled; 

The next teams are assigned according to conference and bowl alignment agreements;

The remaining bowls are filled according to bowl preference and team acceptance. 

That would produce the following playoff line-up if we applied it to 2013’s season:

Rose Bowl:  [1] Florida State vs. [8] UCF
Sugar Bowl: [2] Auburn vs. [7] Northern Illinois
Orange Bowl:  [3] Ohio State vs. [6] Baylor
Fiesta Bowl:  [4] Stanford vs. [5] Michigan State

Let’s say FSU beats UCF, Auburn beats N Illinois, Baylor beats OSU, and Michigan St beats Stanford.  That gives the following second-round playoff match-ups:

Rose Bowl:  [1] Florida State vs. [5] Michigan State
Sugar Bowl:  [2] Auburn vs. [6] Baylor

Now let’s say the two games both end up in upsets; Michigan State upsets Florida State and Baylor beats Auburn.  The same two bowls would be used, and the higher-seeded team stays put.  That would give us:

Rose Bowl:  [5] Michigan State vs. [6] Baylor

It’s workable, it’s based on common-sense standards for ranks, and it’s not your canned-product football we’ve choked on for decades.