Sunday, September 12, 2010

An Analysis of AQ OOC Scheduling 1998-2009: More FCS Please!

Editor's Note:  This is the long, detailed analysis of Out of Conference Scheduling for teams from AQ conferences since the inception of the BCS in 1998.  A much shorter summary with the highlights appeared on Playoff PAC here.

Editor's Note Part 2:  The Big Ten 2009 FBS non-AQ & FCS %'s were originally reported incorrectly.  Those numbers have been corrected as have the corresponding averages based which include that year of data.

Editor's Note Part 3:  Some Pac-10 %'s were originally reported incorrectly.  Those numbers have been corrected as have the corresponding averages (8/7/11).

A common college football myth holds that the superiority and excitement of the regular season exists because of the absence of a multiple-team playoff.1  But consider Week #12 from the 2009 season.  Remarkably, only one top-25 team played a game against another top-25 team. Meanwhile, top-ranked Florida played an out-of-conference (OOC) game against a Sun Belt School with 3 wins (FIU) while 2nd-ranked Alabama sweated it out with an FCS team (Chattanooga).  Is this what college football fans expect from the “most exciting regular season in sports”?

Of course, picking one week at random is by no means an appropriate measure of the overall excitement for the regular season.  However, these particularly weak OOC matchups for the top 2 teams made us start thinking about OOC games in general. 

The specific target for the empirical analysis is how OOC scheduling by the teams in the six AQ conferences has changed since the start of the BCS era in 1998.  Additionally, the trends over the last four years were examined as 2006 marked both the start of a standard 12 game regular season in addition to the creation of a 5th BCS game—thus resulting in an increase in the number of at-large BCS slots by 50% from two to four.  Specifically, I wanted to see what has happened to the number of OOC games played versus fellow AQ teams, versus FBS non-AQ teams and versus FCS teams.  After all, when “every week is a playoff,” one crucial loss may keep a team out of the Title Game or out of a BCS Bowl.  In line with our hypothesis, the results of our empirical analysis show a marked shift in OOC scheduling since 1998.  Apparently, to cope with a sport where one loss eliminates the possibility of a title, teams respond by scheduling Chattanooga and other patsies.

I classified every OOC game from 1998-2009 for every team that was in one of the six AQ conferences at any time during that period into one of three categories: AQ, non-AQ FBS, and FCS based on the opponents’ status.  Notre Dame was grouped into the “AQ” category given its separate BCS eligibility rules and its automatic BCS payout regardless of its BCS bowl qualification status.
 
Percentage of OOC games vs. AQ schools
CONF
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
AVG
ACC
64
50
46
47
48
58
44
44
33
46
48
42
47.5
BIG EAST
44
48
31
39
44
40
43
38
44
38
39
40
40.7
BIG 10
38
43
44
45
38
52
45
42
32
30
30
32
39.3
BIG 12
29
34
29
22
35
31
22
19
23
23
31
23
26.8
PAC 10
47
33
39
40
43
45
42
32
42
35
45
50
41.1
SEC
25
22
19
25
35
35
22
28
35
29
31
30
28

The table above shows that four of the six conferences were fairly consistent in the % of games played versus fellow AQ schools while both the ACC and Big 10 showed fairly significant downward trends.  However, the ACC still had the 2nd highest % of games against fellow AQ schools in 2009 and the highest overall average OOC AQ % over the period at 47.5%.  Interestingly the three conferences with the lowest average %’s, the Big 12 (26.8%), SEC (28%), and Big Ten (39.3%), also received the largest number of BCS at-large bids.  Those conferences have received a whopping 21 of the 23 BCS at-large bids awarded to AQ schools (note: Notre Dame is not included in this calculation).  Consequently, the Big 12 and SEC have averaged close to 3/4 of their OOC games against FBS non-AQ schools and FCS schools over the past 12 seasons and in several years reached nearly 80%.

Percentage of OOC games vs. FBS non-AQ schools
CONF
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
AVG
ACC
21
39
43
40
45
31
41
47
48
40
23
29
37.3
BIG EAST
44
42
63
58
44
55
43
34
40
48
44
35
45.8
BIG 10
59
51
56
52
55
43
52
52
50
52
50
48 
51.7
BIG 12
63
58
63
59
52
54
67
58
54
60
48
58
57.8
PAC 10
53
64
58
53
45
50
55
48
42
58
48
37
50.9
SEC
67
69
72
64
55
54
64
53
48
52
50
49
58.1

Percentage of OOC games vs. FCS schools
CONF
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
AVG
ACC
14
11
11
13
8
11
15
8
19
15
29
29
15.3
BIG EAST
13
9
6
3
12
5
15
19
15
15
17
25
12.8
BIG 10
3
6
0
3
6
5
3
6
18
18
20
20 
9.0 
BIG 12
8
8
8
19
13
15
11
22
23
17
21
19
15.3
PAC 10
0
3
3
7
13
5
3
19
16
6
6
13
7.8
SEC
8
8
8
11
10
10
14
19
17
19
19
21
13.7

For the non-AQ and FCS games, a clear trend emerged: in recent years, teams have scheduled more and more games against FCS schools.  Games against FCS schools were almost non-existent in the late 90’s and early 2000’s; however, over time, this “patsy scheduling” trend expanded to the currently high levels of about 20% of OOC games among all the conferences.   For some conferences (Big 12, Big 10) these games have come at the expense of AQ games, while others (ACC, Big East, SEC) have traded FBS non-AQ opponents for FCS ones.  Four of the six conferences (ACC, Big East, Big Ten, SEC) saw a record % of OOC games against FCS teams occur in 2009. 

Another area to examine was what has happened in the last four years (2006-2009) compared to the first eight years (1998-2005) of the BCS era, since the NCAA expanded to a standard 12 game regular season at the FBS level in beginning in 2006 (2).  It should be noted that many teams played regular seasons of 12 or more games prior to 2006 due to a variety of factors. First, in 2002 and 2003, most FBS teams played a standard 12 game regular season due to existing NCAA rule 17.9.5.1.  This bylaw allowed for twelve regular season games in any year “in which there are 14 Saturdays from the first permissible playing date through the last playing date in November.”2.   Second, the NCAA previously exempted games played as a preseason “kickoff classic” game from the regular season playing limit restrictions.  Third, road games played against institutions based in Hawaii or Alaska are also granted exceptions to the regular season playing limit according to NCAA rule 17.5.9.2 section (k). 

AVERAGE % of OOC Games Against AQ schools Before/After 12 Gm Regular Season
CONF
1998-2009
1998-2005
2006-09
Difference
ACC
47.5
50.1
42.3
-7.9
BIG EAST
40.7
40.9
40.3
-0.6
BIG 10
39.3
43.4
31.0
-12.4
BIG 12
26.8
27.6
25.0
-2.6
PAC 10
41.1
40.1
43.0
2.9
SEC
28.0
26.4
31.3
4.9
AVERAGE % of OOC Games Against FBS non-AQ schools Before/After 12 Gm Regular Season
CONF
1998-2009
1998-2005
2006-09
Difference
ACC
37.3
38.4
35.0
-3.4
BIG EAST
45.8
47.9
41.8
-6.1
BIG 10
51.7
52.5
50.0
-2.5
BIG 12
57.8
59.3
55.0
-4.3
PAC 10
50.9
53.3
46.3
-7.0
SEC
58.1
62.3
49.8
-12.5
On the AQ side, when comparing the % of OOC games versus AQ opponents from these two categories, the results were mixed.  The Big East was virtually unchanged (-.6%), while the ACC (-9.8%), Big Ten (12.4%), and Big 12 (-2.6%) all saw decreases.  However, the Pac-10 and SEC increased their % of OOC games against AQ schools by 2.9% and 4.9% respectively.  Every single conference saw a drop in the average % of OOC games played against FBS non-AQ schools with particularly large drops from the SEC (-12.5%) and Pac-10 (-7.0%).  But the most telling data was that from the OOC games played against the FCS schools (as seen below) in recent years. 

AVERAGE % of OOC Games Against FCS schools Before/After 12 Gm Regular Season
CONF
1998-2009
1998-2005
2006-09
Difference
ACC
15.3
11.4
23.0
11.6
BIG EAST
12.8
10.3
18.0
7.8
BIG 10
9.0 
4.0
19.0
15.0
BIG 12
15.3
13.0
20.0
7.0
PAC 10
7.8
6.6
10.3
3.7
SEC
13.7
11.0
19.0
8.0

From 2006-2009, every conference except the PAC-10 has averaged nearly 1 out of 5 OOC games against FCS opponents.  Every conference increased the average % of games compared to the first eight years of the BCS era with the Big Ten (+15%) and ACC (+11.6%) leading the way.  The Big Ten example is particularly telling because from 1998-2005, they averaged playing an FCS team in 1 out of every 25 OCC games (lowest average % among all the conferences) while from 2006-2009 this average was 1 out of every 5.3 games.  This increase in scheduling FCS teams is also a byproduct of two additional factors.  First, the NCAA rule 18.7.2.2.1 which allows one win against FCS schools (which meet certain scholarship criteria) to count towards the 6 victories needed for bowl eligibility every year.2  Second, the increased competitiveness of FBS non-AQ schools against AQ schools during both regular season and post-season games.  Therefore, I frankly don’t blame the AQ teams.  Under the current format where teams repeatedly hear that “every week is a playoff”, what incentive is there for a team to increase its chance of losing (and thus being “eliminated”) by scheduling harder opponents (AQ and FBS non-AQ schools).  Especially when you consider that many FCS teams will gladly travel to play a road game without any requirement of a future home game, common among AQ school agreements, due to the large amounts of cash they receive to take a beating (Note: don’t tell this to Boise State who has had trouble scheduling AQ teams for this type of agreement in 2011).

As mentioned above, in 2002 and 2003 and 2006-2009 almost all teams played at least 12 games in the regular season so it might be more relevant to compare what happened to OOC scheduling in those years versus 1998-2001 and 2004-2005 (when teams were permitted to schedule a maximum of 11 regular season games not including Hawaii and pre-season kickoff exceptions).  The averages for each of the three groups, as seen in the tables below, were much smaller than those found by comparing the first eight years of the BCS to the past four.

AVERAGE % of OOC Games vs. AQ schools in 11 Game vs. 12 Game Regular Seasons
CONF
11 GM RS
12 GM RS
Difference
ACC
49.2
45.8
-3.3
BIG EAST
40.5
40.8
0.3
BIG 10
42.8
35.7
-7.2
BIG 12
25.8
27.7
1.8
PAC 10
38.8
43.3
4.5
SEC
23.5
32.5
9.0
AVERAGE % of OOC Games vs. FBS non-AQ schools in 11 Game vs. 12 Game Regular Seasons
CONF
11 GM RS
12 GM RS
Difference
ACC
38.5
36.0
-2.5
BIG EAST
47.3
44.3
-3.0
BIG 10
53.7
49.7
-4.0
BIG 12
61.3
54.3
-7.0
PAC 10
55.2
46.7
-8.5
SEC
64.8
51.3
-13.5
AVERAGE % of OOC Games vs. FCS schools in 11 Game vs. 12 Game Regular Seasons
CONF
11 GM RS
12 GM RS
Difference
ACC
12.0
18.5
6.5
BIG EAST
10.8
14.8
4.0
BIG 10
3.5
14.5
11.0
BIG 12
12.7
18.0
5.3
PAC 10
5.8
9.8
4.0
SEC
11.3
16.0
4.7

The averages for the % of OOC games against AQ schools trended the same direction as the pre/post-2006 averages for every conference except the Big 12 and Big East.  The Big 12 and Big East both went from decreasing average %’s to increasing average %’s while the Big Ten only had a decrease of 7.2% of AQ OOC games compared to its 12.4% decrease when comparing pre-2006 versus post-2006.  Additionally the SEC showed an increase in the % of OOC games played versus AQ schools of 9%.  Similarly every conference showed a decrease in the % of OOC games played versus FBS non-AQ schools, but 4 of the 6 six conferences decreased by a larger amount than the pre-2006 and post-2006 averages.

On the FCS side, the results were nearly uniform.  All six conferences played a greater % of their games against FCS opponents in seasons where they played more regular season games, while every conference except the Big Ten (+.1%) saw the difference between the average %’s decreased compared to differences between the pre-2006 and post-2006 numbers.  Thus when comparing apples to apples or years with 12 game regular seasons to those with 11 game regular seasons, the numbers are not quite as bleak but still show an overall trend towards scheduling easier opponents.  Finally, I should mention that I am by no means denigrating the effort and ability FCS schools put up against the big boys.  In recent years, these teams have put up heroic efforts to both nearly defeat AQ Schools (Iowa-Northern Iowa 2009) and grab stunning upsets over AQ schools with five such upsets having already taken place this season.  However, the fact of the matter is that a large majority of FBS versus FCS games are blowouts that usually result in nothing more than a tune-up for the first-string players of the FBS team while at the same time the largest amount of playing time many second and third-string players will see.

Before I conclude, there are a number of things which I should mention regarding the data.  First, the ACC and Big East each experienced membership changes during the past 10 years and these changes should be considered when looking at the %’s for each category.  For example, Miami and Florida State are rivals who have played every year from 1998-2009. Before Miami joined the ACC, this game would have counted as an OOC game for both schools; however, starting in 2004 when Miami & Virginia Tech joined the ACC from the Big East, this game no longer qualified as such since it is now a conference game.  The same is true for the annual Virginia Tech-Virginia rivalry game.  Another thing to consider is ongoing, usually in-state, rivalries which exist between AQ schools.  For example, Georgia & Georgia Tech play every year as do Iowa and Iowa State.  Therefore, these schools generally start each year with at least one OOC game on their schedule against an AQ team and as a result, the numbers mentioned above should be not be viewed in a vacuum without some context.  When considering these rivalry games, the % of OOC games versus AQ opponents for the SEC looks even worse as it usually has at least 3 of these matchups (Georgia-Georgia Tech, UF-FSU, South Carolina-Clemson) every year.  Also, several schools changed status between FBS AQ, FBS non-AQ, and FCS classification.  Again this makes the SEC look even worse with Louisville and Kentucky as an example. They have played every year since 1998 with Louisville counting a non-AQ school for the first seven meetings and as an AQ school for the last five.

Although I’ve examined a large amount of data and broken the numbers down a variety of ways, they all lead to some very simple conclusions regarding how FBS teams from the six AQ conferences construct the three to five OOC games they play each year.  First, since the BCS went into effect in 1998, AQ teams, as a whole, have decreased their % of OOC games played against fellow AQ schools.  Second, teams have played an increasing % of OOC games against FCS schools with three of the six conferences setting records in 2009 at 21%, 25% and 29%.  Third, from 2006-2009, after the NCAA switched to a 12 game regular season and the BCS added a 5th BCS bowl game, teams from all conferences increased the average % of OOC games played versus FCS schools, some by as much as 15%.   Fourth, the Big Ten, Big 12, and SEC have earned 91% of the BCS at-large berths awarded to AQ schools and not coincidentally, have the lowest average % of games versus fellow AQ schools over all three of the time periods I examined.  Last and most importantly, when given the option or mandate of playing an additional regular season game, some conferences increased the average % of OOC games versus fellow AQ schools while some decreased their %, but ALL conferences increased the average % of games played against FCS schools.  Therefore I can’t help but conclude that the current BCS system encourages teams to schedule weaker opponents and reduces the potential for exciting regular season matchups between top quality teams.  Furthermore, an 8 or 12-team playoff would still require near-perfect records for admission while at the same time encouraging AQ schools to schedule tougher OOC games for several reasons.  First, they would help prepare teams for the high level of competition they would face in a postseason playoff.  More importantly they eliminate the level of perfection required for national championship consideration under the current system which has resulted in weaker OOC schedules.   

1 Bill Hancock, BCS Executive Director, recently said of the FBS regular season, “..we have the best, most compelling regular season of any sport.” http://detnews.com/article/20100731/SPORTS0203/7310341/BCS-chief--Forget-about-playoff-system

2 2010-2011 NCAA Division I Manual - http://www.ncaapublications.com/productdownloads/D111.pdf

No comments:

Post a Comment