Football Final Four Participants Showing Regional Computer Points Seeding

Year	Division	Champion	Regional Seed	Runner Up	Regional Seed	Semi-Finalist	Regional Seed	Semi-Finalist	Regional Seed
2000	1	Upper Arlington	2	Solon	1	Colerain	2	Massillon Perry	1
	2	Olmstead Falls	2	Piqua	2	Marysville	1	Buchtel	6
	3	Canton CC	1	Van Wert	8	Portsmouth	1	Copley	2
	4	Youngstown Ursuline	2	Coldwater	4	Germantown Valley View	2	Newark Licking Valley	1
	5	Amanda-Clearcreek	3	Bedford St. Peter Chanel	1	Reading	4	Liberty Center	1
	6	Maria Stein Marion Local	1	Mogadore	1	Delphos St. John's	1	Newark Catholic	6
2001	1	Cleveland St. Ignatius	7	Cincinnati St. Xavier	1	Massillon Washington	1	Dublin Coffman	7
	2	Toledo St. Francis DeSales	2	Columbus Watterson	5	Green	3	Vandalia Butler	4
	3	Mentor Lake Catholic	5	Columbus St. Francis DeSales	8	Bellevue	6	Kettering Alter	3
	4	Kenton	1	Newark Licking Valley	2	Akron St. Vincent St. Mary	7	Coldwater	4
	5	Bedford St. Peter Chanel	1	Marion Pleasant	1	Woodsfield Monroe Central	2	Sidney Lehman	3
	6	Maria Stein Marion Local	1	Mogadore	1	Shadyside	4	Columbus Grove	5
2002	1	Cincinnati Elder	1	Warren G. Harding	2	Findlay	2	Massillon Washington	8
	2	Dayton Cham. Jul	2	Macedonia Nordonia	4	Toledo St. Francis	1	Olmstead Falls	5
	3	Columbus Watterson	4	Cleveland Bendectine	1	Germantown Valley View	2	Akron Hoban	2
	4	Kenton	1	Portsmouth West	7	Kettering Alter	1	Youngstown Ursuline	8
	5	Marion Pleasant	1	Smithville	1	Amanda-Clearcreek	2	Delphos St. John's	6
	6	Mogadore	1	Dola Hadin Northern	2	Columbus Grove	1	Newark Catholic	4
2003	1	Cincinnati Elder	1	Lakewood St. Edward	3	Toledo St. Johns Jesuit	7	Dublin Scioto	3
	2	Avon Lake	2	Trenton Edgewood	1	Chardon	3	Columbus Brookhaven	1
	3	Cleveland Benedictine	1	Kenton	5	Steubenville	1	Newark Licking Valley	4
	4	Versailles	2	Cleveland Villa Angela-St. Joseph	6	Bellaire	4	Delta	4
	5	Gahanna Columbus Academy	4	Amanda-Clearcreek	1	Warren JFK	2	Delphos St. John's	3
	6	Columbus Grove	1	Maria Stein Marion Local	8	Monroeville	4	Newark Catholic	1
2004	1	Cincinnati Colerain	3	Canton McKinley	5	Worthington Kilbourne	4	Cleveland Glenville	5
	2	Columbus Brookhaven	2	Avon Lake	1	Dayton Carroll	2	Tallmadge	1
	3	Cleveland Benedictine	4	St. Marys Memorial	5	Columbus Watterson	3	Canal Fulton Northwest	3
	4	Youngstown Cardinal Mooney	3	Versailles	7	Coldwater	2	Martins Ferry	2
	5	St Henry	8	Amanda-Clearcreek	1	Hamler Patrick Henry	1	Smithville	5
	6	Dola Hardin Northern	2	Norwalk St. Paul	2	Danville	8	Hicksville	3
2005	1	Cincinnati St. Xavier	1	Massillon Washington	3	Hilliard Davidson	1	Lakewood St. Edward	1
	2	Toledo Central Catholic	2	Canfield	2	Tallmadge	1	Dayton Carroll	6
	3	Steubenville	1	Columbus St. Francis DeSales	3	Akron Buchtel	3	Clyde	1
	4	Coldwater	2	Youngstown Cardinal Mooney	1	Bellaire	1	Monroe Lemon-Monroe	2
	5	Hamler Patrick Henery	1	Cincinnati Hills Christian Academy	1	North Lima South Range	1	Centerburg	6
	6	Delphos St. John's	5	Steubenville Catholic Central	3	Newark Catholic	5	Bascom Hopewell-Loudon	1

A quick discussion of tournament theory is of interest. Most believe that an ideal tournament crowns the "best" team in any league (as opposed to a team that gets on a roll at the right time.) In order to do this, two issues must be balanced: 1) tournament participant selection and 2) tournament design.

One can imagine a tournament that consists of one team. This is actually just the old polls which crowned a champion at the end of the season. This "tournament" puts an infinite premium on regular season performance. This structure is unacceptable since it is often impossible to identify the "best" team without post-season play, especially if there is minimal head-to-head play among the top teams.

Alternatively, a playoff in which every team in the league participates without need for qualification leaves the regular season much less important. A tournament without need for qualification leaves the possibility that a team that is not necessarily the "best" will get on a winning streak at the end of the season and be crowned champion.

Design of the tournament includes the number of games to be played and the seeding hierarchy.

In a tournament with more games (made necessary by increased pool of participating teams), it becomes more likely that the "best" team will have a bad game and lose. While this structure doesn't necessarily favor the very worst of teams, it make it much less likely that the "best" team will win.

The idea of seeding a tournament rewards those teams that have performed well during the regular season. There are various types of seeding, each bestowing upon stronger teams different amounts of favoritism. A tournament with strong seeding makes it much more difficult for a team to win if they did not do well in the regular season. A tournament with weak seeding gives less preference to stronger teams and may make it less likely that the "best" team will be champions.

A very interesting study of playoff structure in regards to I-A college football is:
Annis, David H. and Wu, Samuel S. (2006) "A Comparison of Potential Playoff Systems for NCAA I-AA Football," The American Statistican, 151-157.

The combination of computer point ratings and the structure of the playoff system itself seems to be working quite well. The fact that superior playoff performance has been exhibited by the highest seeded teams validates the use of the Harbin computer rating system. The ever-decreasing success for lower seeded teams indicates that a similar trend would continue if more teams were allowed in.

Allowing more teams in would simply lessen the likelihood of finding the truly "best" team in each division.

	Champions	Runners-up	Semi-Finalists	Total
1 seeds	15 (42%)	14 (39%)	22 (31%)	51 (35%)
2 seeds	11 (31%)	6 (17%)	13 (18%)	30 (21%)
3 seeds	3 (8%)	4 (11%)	10 (14%)	17 (12%)
4 seeds	3 (8%)	2 (6%)	10 (14%)	15 (10%)
5 seeds	2 (6%)	4 (11%)	5 (7%)	11 (8%)
6 seeds	0	1 (3%)	6 (8%)	7 (5%)
7 seeds	1 (3%)	2 (6%)	3 (4%)	6 (4%)
8 seeds	1 (3%)	3 (8%)	3 (4%)	7 (5%)