Showing posts with label NCAA. Show all posts
Showing posts with label NCAA. Show all posts

Thursday, February 5, 2009

How can Soccer Improve College Football? Part 1

Everybody is talking about how we can fix the mess in collage football known as the BCS, including President Obama.  Generally speaking, unless your favorite team is in the BCS title game you dislike the system.

The Europeans have an interesting system in place for the Champions League tournament (as do other European "football" championships and cups).  The Champions League takes the best soccer club teams from each country's league and determines the champion of Europe.


Here is how it works:

76 teams from the 52 European soccer associations participate.  The better quality leagues (as determined by a statistical system) send more teams.  So the tournament is made up of the following:

Leagues 1-3 send their top four teams (12 teams)
Leagues 4-6 send their top three teams (9 teams)
Leagues 7-15 send their top two teams (18 teams)
Leagues 16-52 send their top one team (37 teams)

The tournament follows an interesting format and begins with three qualifying rounds.

Qualifying Round 1: The teams ranked 49-76 (lowest ranked) are randomly paired off into 14 match-ups.  Each pair play two games against one another - one game at each team's home venue.  The combined score, aggregate, determines the winner.  If a tie exists after the 2nd game, tie-breakers are used to determine the winner.  The 14 winners move on to the next round.

Qualifying Round 2: The teams ranked 35-48 and the 14 round 1 winners are randomly paired off into a new group of matches.  The winners are determined just as they are in round 1.  The 14 winners move on.

Qualifying Round 3: The teams ranked 17-34 (18 teams) are mixed with the 14 teams that qualified past round 2.  This creates 16 games and 16 teams to move on to the Group Stage.  The winners are again determined by an aggregate score.

Group Stage: This stage is made up of the 16 round 3 winners and the top 16 teams in the tournament (teams 1-16).  The teams are split into eight groups of four teams.  These teams play "round robbing" and the top two teams from each group advance to the first knockout round.  Each team in a group plays everyone else in their group twice.

The rest of the tournament follows a typical structure.  The teams are paired off in a bracket and then move to the quarter-finals, semi-finals, and finals.  All rounds are aggregate, except the finals.


So, a long story short, all teams get a shot.  This is true regardless of which league they are from or how "good" that league is relative to the other leagues.  The catch is that the lesser leagues have a longer road to travel and do not send as many teams.  If we use this model for NCAA Division 1 football (bowl championship division), I think we have something interesting.  The powerhouse conferences (SEC and BIG 12) would send more teams followed by the BIG 10 and PAC 10.  The mid-majors would send a couple followed by the smaller conference such as the Sun Belt Conference.  And, if a school is "too good" for a conference, we can award them the Notre Dame trophy of excellence and give them an NBC television contract.

See my next post for a "prototype" of this concept applied to college football.

How can Soccer Improve College Football? Part 2

Now that I've described the European Champions League format (note, the aggregate scoring was not used for the below example, only one game is needed for football), it is time to apply that concept to college football.  The first step is to rank the conferences to determine how many teams from each conference will qualify for the championship tournament.  Here is a quick ranking that I put together based on 2008 for the conferences and the number of teams from each:

Tier 1
1) SEC (4 teams)
2) Big 12 (4 teams)

Tier 2
3) Pac 10 (3 teams)
4) Big East (3 teams)

Tier 3
5) Big 10 (2 teams)
6) WAC (2 teams)
7) ACC (2 teams)

Tier 4
8) Conf USA (1 team)
9) Mountain West (1 team)
10) Mid-American (1 team)
11) Sun Belt (1 team)

This gives us 24 teams in the tournament.  Note that the tiers can be adjusted each season based on conference performance.  Just as with the Champions League in Europe, the better teams will enter in later rounds.  Here is a proposed breakdown for college football (based loosely on 2008 performance):

The first set of teams to enter:
24) Tier 4 Champion (Sun Belt - Troy)
23) Tier 4 Champion (Mid Am. - Ball State)
22) Tier 4 Champion  (Mnt. West - Utah)
21) Tier 4 Champion (Conf USA - East Carolina)
20) Tier 1 3rd Runner Up (Big 12 - Missouri)
19) Tier 1 3rd Runner Up (SEC - Ole Miss)
18) Tier 2 2nd Runner Up (Big East - West Virginia)
17) Tier 2 2nd Runner Up (Pac 10 - Oregon)

The teams to enter Round 2:
16) Tier 3 Runner Up (ACC - Boston College)
15) Tier 3 Runner Up (Big 10 - Ohio State)
14) Tier 3 Runner Up (WAC - Fresno State)
13) Tier 1 2nd Runner Up (SEC - Georgia)
12) Tier 1 2nd Runner Up (Big 12 - Texas Tech)
11) Tier 2 Runner Up (Big East - Pitt)
10) Tier 2 Runner up (Pac 10 - Oregon State)
9)  Tier 3 Champion (lowest ranked of the three tier 3 champs) (WAC - Boise State)

Round 3 new entries:
8) Tier 3 Champion (ACC - Virginia Tech)
7) Tier 3 Champion (Big 10 - Penn State)
6) Tier 1 Runner Up (SEC - Alabama)
5) Tier 1 Runner Up (Big 12 - Texas)
4) Tier 2 Champion (Big East - Cincinnati)
3) Tier 2 Champion (Pac 10 - USC)

Quarterfinal entries:
2) Tier 1 Champion (Big 12 - Oklahoma)
1) Tier 1 Champion (SEC - Florida)

Now we have the teams and when they enter.  My next step was to simulate the tournament.  Using a random number generator I randomly paired the teams for each round up and through the quarterfinals.  Then I used a crude algorithm to determine the winner.  This algorithm used a weight system to give higher ranked teams an advantage since they would be a better team.  Weights were equal to teams that entered in the same phase.  So, for example, Texas and Cincy would not have a weight but Texas/Florida would be weighted in favor of Florida.  If Florida were to play Utah, a more severe weight would be installed since Utah entered in the first round.

Now on to the tourney!

Round 1
Missouri beats Ball State
Utah beats Oregon
Troy beats West Virginia
Ole Miss beats East Carolina

The four winners (MU, Utah, Troy, Ole Miss) move to Round 2 with the Round 2 entrants.  Again, all teams were randomly paired for round 2.

Round 2
Missouri beats Oregon State
Texas Tech beats Georgia
Boston College beats Troy
Ohio State beats Pitt
Utah beats Boise State
Old Miss beats Fresno State

Same thing for Round 3
Ohio State upsets Cincy
Penn State over Alabama
Boston College upsets Texas
Utah beats Missouri
Virginia Tech beats USC

Quarterfinals (the final two teams enter, last round to use a random draw)
Ohio State beats Boston College
Florida beats Utah
Oklahoma beats Virginia Tech
Penn St beats Texas Tech

Semi Finals
Ohio St beats Florida
Penn St beats Oklahoma

Finals
Penn St wins over Ohio State.

So there you go.  Obviously this would need some tweaking.  Overall rankings might be needed and seeding each round might make for more fair match-ups.  The Utah/Missouri matchup shows this as two early entry teams play in a later round.  On the other hand, the random draw creates a fair bit of excitement and adds an element of surprise.  

You can also argue how the teams are initially entered and the feasibility of this working with the current season structure.  Teams would have to cut out some none conference games.  Also, it would be idea if the conferences all had the same number of teams and played the same number of games and determined their champions in a similar manner.

Leave comments and let me know how you'd tweak it up!