Tuesday, November 30, 2010

NCAA FBS Top 25 as of November 27

The 2010 NCAA FBS regular season is almost over, so here is the latest NCAA FBS production Top 25 rank. Notice, even with Boise State's loss this past weekend, they still remain the most productive team in the nation, although the gap between them and the second place team has narrowed considerably.

Rank
School
1
Boise State
2
TCU
3
Oregon
4
Ohio State
5
Alabama
6
Northern Illinois
7
Wisconsin
8
Hawai'i
9
Nevada
10
Stanford
11
Oklahoma State
12
Nebraska
13
UCF
14
Arkansas
15
Iowa
16
Auburn
17
Virginia Tech
18
Oklahoma
19
West Virginia
20
Florida State
21
Air Force
22
Miami (Florida)
23
Arizona
24
Michigan State
25
Missouri

Top 25 for the first week of the BCS (October 16, 2010).
Top 25 for the second week of the BCS (October 23, 2010).
Top 25 for the third week of the BCS (October 30, 2010).
Top 25 for the fourth week of the BCS (November 6, 2010).
Top 25 for the fifth week of the BCS (November 13, 2010).
Top 25 for the sixth week of the BCS (November 20, 2010).

Friday, November 26, 2010

The Iron Bowl - 2010 Edition

In a few minutes, the Iron Bowl between Auburn and Alabama will be under way, and according to the BCS, Auburn is ranked #2 and Alabama is ranked #9. Yet in terms of overall productivity, I have Auburn ranked #12 and Alabama ranked #4. So why does the production ranking differ from the conventional wisdom? Let's take a look at the two teams offense and defense in terms of productivity rank.


Offensive Production Ranks
Auburn ranks as the #7 most productive offense in the nation and Alabama ranks as the #16 most productive offense in the nation. So on the offensive side of the ball, Auburn is the better offense than Alabama. This I do not think is a shock to college football pundints.

Defensive Production Ranks
Alabama is the #2 most productive defense in the nation, just behind Boise State and Auburn is the #55 most productive defense in the nation, just behind Toledo. Thus on the defensive side of the ball, Alabama is a much better team.

Final (Overall) Ranking
So how do you compare the two? What I do is take each teams offensive production and subtract the teams defensive production and calculate the teams overall production. When taking account both the offensive and defensive productivity, Alabama is a more productive team than Auburn. Hence the productivity ranking has Alabama as a higher ranked (i.e. better team) than Auburn. Does that mean that Alabama will win today? No, this is just the statistical analysis of NCAA FBS teams, and these are the results from the statistical analysis.

Thursday, November 25, 2010

Ohio State - Boise State Debate

Recently the presidents of Ohio State and Boise State have made public comments about the worthiness of their respective football teams to play in the NCAA FBS championship game in January. The main issue seems to revolve around each teams strength of schedule. About as recently, I have been writing about the measurement of NCAA FBS strength of schedule and whether strength of schedule is important in regard to a teams winning percent. I conclude that my measure of strength of schedule is not a statistically significant (and therefore) important factor in determining a teams winning percent when taking into account team productivity.

Now we have this debate come up again, so I thought that I would look into the two conferences and the two teams strength of schedule as of November 20, 2010. Let's take them in order.

Strength of Schedule (SOS) of the Big Ten vs. Western Athletic Conference

Now there is a difference between the two conferences, and the Big Ten has an overall stronger conference schedule than the Western Athletic Conference. Here are the two conference teams winning percent and strength of schedule as I measured them using the methodology I outlined in a previous blog.

Team
SOS
Winpct
Illinois
59.455
0.545
Indiana
58.182
0.364
Iowa
57.727
0.636
Michigan
61.909
0.636
Michigan St.
61.545
0.909
Minnesota
53.636
0.182
Northwestern
77.545
0.636
Ohio St.
63.273
0.909
Penn St.
53.000
0.636
Purdue
54.727
0.400
Wisconsin
71.364
0.909
Big Ten Ave.
61.124






Boise St.
73.700
1.000
Fresno St.
70.000
0.667
Hawaii
72.500
0.800
Idaho
61.500
0.400
La.-Lafayette
66.818
0.182
Nevada
84.400
0.900
New Mexico St.
74.455
0.200
San Jose St.
59.200
0.111
Utah St.
66.727
0.364
WAC Average
69.922


Notice that the difference in strength of schedule is higher (easier) for Big Ten teams on average than for Western Athletic Teams. So Gee's comment that teams that play in the Big Ten play more challenging teams on average than teams in the Western Athletic Conference. But just because the Big Ten's strength of schedule is harder than the Western Athletic Conference does not make Ohio State more worthy than Boise State. So let's turn to the two teams themselves.

Strength of Schedule (SOS): Ohio State vs. Boise State

Listed below are the two teams 2010 schedules and the two teams opponents strength of schedule as of November 20th. Notice that both teams have played very poor teams, such as Boise State playing Wyoming, New Mexico State and San Jose State. Ohio State has played some fairly low productive teams such as Eastern Michigan, Purdue and Indiana. In fairness, Boise State plays their strongest opponent (Nevada - currently ranked #5 in overall productivity) this week and then follows with a very weak Utah State (#99 in overall productivity), while Ohio State plays an above average Michigan (currently ranked #37) in overall productivity.

Team Opponent SOS
Boise St. Virginia Tech 18
Boise St. Wyoming 108
Boise St. Oregon St. 67
Boise St. New Mexico St. 118
Boise St. Toledo 50
Boise St. San Jose St. 116
Boise St. Louisiana Tech 84
Boise St. Hawaii 11
Boise St. Idaho 85
Boise St. Fresno St. 80
Boise St. Nevada DNP
Boise St. Utah St. DNP



Ohio St. Marshall 95
Ohio St. Miami (FL) 17
Ohio St. Ohio 43
Ohio St. Eastern Mich. 114
Ohio St. Illinois 49
Ohio St. Indiana 97
Ohio St. Wisconsin 9
Ohio St. Purdue 100
Ohio St. Minnesota 94
Ohio St. Penn St. 68
Ohio St. Iowa 10
Ohio St. Michigan DNP

Yet all of this does not refute the overall productivity of the two teams. Taking this into account, we see that in terms of team productivity, Boise State is the most productive team in the land (ranked #1 in overall productivity) while Ohio State is ranked #6 in overall productivity. So even though Ohio State has a tougher schedule than Boise State, I have adjusted for conferences in the estimation of team overall productivity and found that Boise State is the more productive team. Additionally, when explaining the factors that drive winning percentage, my measure of strength of schedule is not statistically significant when controlling for points scored and points surrendered; point spread; or total productivity.

Full disclosure: I am an employee of a Big Ten university.

Monday, November 22, 2010

NCAA FBS Top 25 as of November 20

After a great weekend of NCAA FBS football, I have finally calculated the newest ranking as of November 20, 2010. The model still has Boise State ranked #1, but notice that Alabama is moving up the ranking. The data used for this comes from cfbstats.com.

Rank
School

1


Boise State

2


Oregon

3


TCU

4


Alabama

5


Nevada

6


Ohio State

7


Oklahoma State

8


Stanford

9


Wisconsin

10


Iowa

11


Hawai'i

12


Auburn

13


Nebraska

14


Northern Illinois

15


Oklahoma

16


UCF

17


Miami (Florida)

18


Virginia Tech

19


Arkansas

20


West Virginia

21


Air Force

22


Arizona

23


Florida State

24


Florida

25


Michigan State



Top 25 for the first week of the BCS (October 16, 2010).
Top 25 for the second week of the BCS (October 23, 2010).
Top 25 for the third week of the BCS (October 30, 2010).
Top 25 for the fourth week of the BCS (November 6, 2010).
Top 25 for the fifth week of the BCS (November 13, 2010).

Friday, November 19, 2010

Does Strength of Schedule Matter?

The last two blogs have looked at how to measure strength of schedule and the teams actual schedule strength as of November 13th. Now I want to turn my attention to how much effect does strength of schedule actually have on team performance. In order to do this, I will need to analyze all 120 NCAA FBS teams through the weekend of November 13th in order to give (a series) of answers as to how (if at all) does strength of schedule affect NCAA FBS team performance. So, taking the data for all 120 NCAA FBS teams up to and including the weekend of November 13, I have calculated each teams strength of schedule, found their winning percent, points scored, points surrendered and also found their total productivity using the complex invasion sport production model. So, here are the results.

First, I ran a linear regression looking at how strength of schedule impacts winning percentage only. The regression estimated is: winning percent = f(strength of schedule). The result from this statistical estimation is that estimated coefficient for strength of schedule is positive and statistically significant, with a t-statistic greater than two in absolute value, and an adjusted r-squared equal to 0.05. Thus only looking at strength of schedule I find that it does impact winning percentage and that teams that have easier schedules have higher winning percentages. The problem is that the variation in strength of schedule "explains" very little of the variation in NCAA FBS team's winning percent - only about 5%. So, let's see if we can do better.

The second linear regression looks at how strength of schedule impacts winning percentage along with the amount of points the teams scores and the amount of points the team surrenders. (I also ran the regression on point spread - points for minus points against). Now strength of schedule variable does not work out well. The results from this regression are that strength of schedule is statistically insignificant (or that statistically strength of schedule has zero impact on winning percentage) when also taking into account the number of points scored and points surrendered. The regression does rather well on the whole, with an adjusted r-squared = 0.84, and the estimated coefficient on points scored is positive and statistically significant at the 99% level of confidence, and the estimated coefficient on points surrendered is negative and statistically significant at the 99% level of confidence.

The same goes for a regression on strength of schedule and the point spread (i.e. points scored minus points surrendered). Strength of schedule is statistically insignificant and point spread is positive and statistically significant at the 99% confidence level.

Given this result, I conclude that strength of schedule does not matter when looking at teams winning percent. It did not matter in the NCAA FBS paper I wrote up about team production, and still does not. So all this talk about one teams strength of schedule better than another teams strength of schedule is rather a waste of time.

Thursday, November 18, 2010

NCAA FBS Strength of Schedule as of November 13

In the previous post, I talked about how I measure NCAA FBS strength of schedule. Now I want to use this strength of schedule measure and examine the teams in the top 25 of the BCS rankings after the weekend of November 13th. Here they are in the table below.

Rank
Team
BCS Average
SOS
1
Oregon
0.9753
78.60
2
Auburn
0.9687
55.09
3
TCU
0.8966
71.09
4
Boise State
0.8634
77.56
5
LSU
0.8243
57.20
6
Stanford
0.7553
72.20
7
Wisconsin
0.7258
73.60
8
Nebraska
0.7203
80.20
9
Ohio State
0.6674
69.30
10
Oklahoma State
0.6601
63.60
11
Alabama
0.6151
64.00
12
Michigan State
0.6066
57.80
13
Arkansas
0.5133
61.20
14
Oklahoma
0.4728
61.80
15
Missouri
0.4563
60.90
16
Virginia Tech
0.3676
71.20
17
South Carolina
0.3244
46.50
18
Nevada
0.3016
79.78
19
Texas A&M
0.2788
56.90
20
Iowa
0.219
61.60
21
Mississippi State
0.177
48.50
22
Arizona
0.1413
68.00
23
Utah
0.1109
76.70
24
Miami (FL)
0.0885
54.40
25
Florida State
0.0415
57.70

As you will notice, Oregon's strength of schedule is larger (easier) than both Boise State's and TCU's. Yet, I have not heard anyone question the legitimacy of Oregon playing in the NCAA championship game. I wonder why?

Wednesday, November 17, 2010

Measuring Strength of Schedule

Over the last few weeks, I have heard a lot of discussion about Boise State and TCU and why they should not be given the same consideration for the national championship game as teams from automatic qualifying conferences. One of the arguments is couched around the perception that Boise State and TCU play easier schedules than other top ranked teams. This got me to thinking about how to measure strength of schedule. There are existing measures of SOS in both NCAA FBS football and NCAA division I basketball.

The NCAA division I basketball is calculated as: 2/3 opponents' winning percentage and 1/3 opponents' opponents' winning percentage. The Sagarin ratings (linked above for NCAA FBS football) are non-intutive, and frankly I am not sure how they are calculated even after reading the notation at the top, and the NCAA basketball strength of schedule measurement seems arbitrary or contrived.

Since I am unaware of a measure of strength of schedule that I feel is simple I thought that I would come up with my own measure of strength of schedule based on the actual rankings of each FBS school for each team. Thus the most productive team will have a rank equal to 1 and the least productive team will have a rank equal to 120, since there are 120 FBS schools. Here it is. Strength of schedule is the average of each team's opponents production rank. Teams with a high strength of schedule number have a weaker strength of schedule, and teams with a low strength of schedule number have a stronger strength of schedule. That seems simple.

Now, there are some issues with calculating strength of schedule.
  • The first is what to do with FBS schools that played non-FBS schools. What I decided was since there are 120 FBS schools this year, then for each non-FBS school a team plays was to measure all non-FBS schools as having a Strength of Schedule equal to 121. This is not perfect since a non-FBS football team may be better than the worst FBS school, which has a rank of 120.
  • Second, highly ranked teams (i.e. whoever is ranked #1) cannot play themselves, so they may potentially have a lower Strength of Schedule rating. Likewise, low ranked teams may potentially have a higher Strength of Schedule rating, since they cannot play themselves.
  • Third, I am only calculating the Strength of Schedule for the games that have been played, not for games that are currently on the schedule that have not been played. This will change the final strength of schedule measurement.
  • Fourth, the difference in rank from the best to second best team is only 1, but that may not really be the difference in their production, and hence my measure of strength of schedule is not totally measuring the difference in each team's opponents perfectly. I could also base strength of schedule on each team's final production numbers, but have not as of now.

With those four issues presented (and possibly others that I have not worked on) let's use an example to see how the strength of schedule is calculated for each NCAA FBS team.

Thus, for the week ending November 13, my model estimates that Boise State is the #1 ranked NCAA FBS team in the nation. So using Boise State's 2010 schedule here is how I calculate each NCAA FBS schools Strength of Schedule.

Date
Opponent
Rank
9/6/2010
Virginia Tech
16
9/18/2010
Wyoming
115
9/25/2010
Oregon St.
72
10/2/2010
New Mexico St.
117
10/9/2010
Toledo
64
10/16/2010
San Jose St.
112
10/26/2010
Louisiana Tech
91
11/6/2010
Hawaii
18
11/12/2010
Idaho
93
11/19/2010
Fresno St.

11/26/2010
Nevada

12/4/2010
Utah St.


Since Idaho was the last team that Boise State played Boise State's Strength of Schedule is measured as the average of the teams rank in terms of overall production, which for Boise State equals 77.556.

Looking at the number we know that a team that plays a purely average schedule will have a strength of schedule rank equal to 59.5, so overall Boise State has a weaker schedule than an average schedule. Guess what? Of the schools in my top 25 for as of November 13, Boise State does not have the easiest strength of schedule, and the team that has a easier strength of schedule than Boise State is not TCU. Check back tomorrow to find out who!

Tuesday, November 16, 2010

NCAA FBS Complex Invasion Sport Model - An Overview

Since the simple model of NCAA FBS production is rather limited, as I explained earlier, I wanted to create a model of NCAA Football Bowl Subdivision (FBS) productivity that goes beyond the number of points scored by the team and the number of points scored against the team; or as some in the media seem to focus on the point spread between points scored and points surrendered.

In order to do this, I am analyizing both the team's offensive production (in a later blog) and the team's defensive production (in a later blog). But for now, I want to give an overview of how I am setting up the complex invasion sport NCAA football production model, and give credit to those who have thought (and wrote) about this idea before me.

First, Bill Gerrard in 2007 wrote a paper in the International Journal of Sport Finance, where he coined the phrase "complex invasion sports" and wrote about how to set up modeling a team sport production function. Basically Gerrard tells us that the actions on the field can be attributed to the team's offense and the team's defense. From the offensive perspective the team actions either increase or decrease the offenses scoring opportunities and from the defensive perspective the teams actions allow the team's opponent to either increase or decrease their scoring opportunities. Obviously, not all scoring opportunities result in a score, so the production model needs to incorporate the rate (or efficiency) at which the scoring opportunities are converted into scores on the offensive side of the ball and the rate at which the team's opponents scoring opportunities are converted into scores on the defensive side of the ball. The offensive conversion rate relates to the teams scoring and the defensive conversion rate (i.e. the team's opponent) relates to the opponents scoring, or the team's points surrendered, which result in the final game outcome. It is with points scored and points surrendered where the simple team production model starts. Thus the simple team production model misses a lot of the actions that take place on the field and impact the final game outcome.

Second, from the book The Wages of Wins (by Berri, Schmidt and myself), we present a production model of the National Football League (NFL) that revolves around four areas in terms of the offensive and defensive productivity. Dave Berri provides a much more rigorous presentation of the NFL model in a book chapter he wrote in 2007, and it is the model from both sources that I draw heavily on in the creation of the NCAA FBS production function. Given that I am standing on the shoulders of giants, let me go through the four general factors that I model NCAA FBS productivity.

NCAA FBS production is determined by:
1. the ability of the offense to acquire the football,
2. move the football down the field,
3. maintain possession of the football, and
4. the efficiency at which teams score points.

Offenses that are relatively more productive in achieving these four factors than other NCAA FBS team offenses are deemed to be more productive and thus will be ranked higher than offenses that are not as productive. Defense is modeled in the same way, but better defenses are ones that in essence do the opposite of the offense.

Specifically, defense is also modeled by:
1. the ability of the defense to keep the football away from their opponent,
2. reducing their opponent from moving the ball toward a scoring attempt,
3. reducing their opponent from maintaining possession of the football, and
4. reducing their opponents efficiency of scoring points.

So that's an overview of the NCAA FBS Complex Invasion Sport Model. What I still need to determine are:
1. the actual on-field variables for each of the four production factors,
2. statistically determine if the variables are significant with respect to the offense scoring or the defense keeping their opponent from scoring,
3. the marginal impact (or weight) each variable has on offense/defense scoring,
4. the impact that different conferences have on offense/defense scoring, and
5. anything else I have forgotten while typing this up.

UPDATE:  The model is published here:  An NCAA Football Bowl Subdivision Production Function.

Monday, November 15, 2010

NCAA FBS Top25 as of November 13

Well given TCU's poor relative performance last week against San Diego State University (a school the model predicts is a very efficient team), and the same for Oregon, both have moved a little out of contention from the overall top spot in the NCAA FBS top rank. Now Boise State is ranked number 1, and given some of the discussion in the media, there seems to be some consensus that this is how others feel (atleast between Boise State and TCU). Here is the new Top 25 based on my model. The exact top 25 ranking is below. The data for this analysis came from cfbstats.com.

Rank
School
1
Boise State
2
Oregon
3
TCU
4
Ohio State
5
Alabama
6
Oklahoma State
7
Nevada
8
Auburn
9
Wisconsin
10
Nebraska
11
Miami (Florida)
12
Iowa
13
Oklahoma
14
Arkansas
15
Stanford
16
Virginia Tech
17
Northern Illinois
18
Hawai'i
19
Arizona
20
UCF
21
Air Force
22
West Virginia
23
Florida State
24
Michigan State
25
Utah

Top 25 for the first week of the BCS (October 16, 2010).
Top 25 for the second week of the BCS (October 23, 2010).
Top 25 for the third week of the BCS (October 30, 2010).
Top 25 for the fourth week of the BCS (November 6, 2010).

Monday, November 8, 2010

NCAA Top 25 as of November 6

TCU now is the most productive team in the nation after their outstanding performance this weekend on both sides of the ball. Boise State has jumped over Oregon from last week as well in the production model of ranking NCAA football teams. The exact top 25 ranking is below. The data for this analysis came from cfbstats.com.

Rank
School
1
TCU
2
Boise State
3
Oregon
4
Ohio State
5
Nevada
6
Iowa
7
Alabama
8
Auburn
9
Oklahoma State
10
Wisconsin
11
Utah
12
Nebraska
13
Miami (Florida)
14
UCF
15
Stanford
16
Arizona
17
Hawai'i
18
Florida
19
Virginia Tech
20
Oklahoma
21
Northern Illinois
22
Florida State
23
San Diego State
24
Michigan State
25
Arkansas

Top 25 for the first week of the BCS (October 16, 2010).
Top 25 for the second week of the BCS (October 23, 2010).
Top 25 for the third week of the BCS (October 30, 2010).

Friday, November 5, 2010

Explaining the College Football Model - Simple Model

OK, I have posted these top 25 ranks for NCAA FBS teams, and the question is: how do I determine who should be in the top 25? Or better yet, we have the AP top 25, USA Today top 25, Harris voting poll and the BCS rankings, so why do we need another NCAA football top 25? Fair enough.

The top 25 that I am posting is not my personal preferences or who I think should be in the top 25, but is rather the results of a statistical analysis of NCAA football offense and defense. So to start this off, I first will look at the broad picture and start with a simple model of NCAA football.

Specifically, the simple model of NCAA football is a production function. By production function what I mean is that it relates output - which for the simple model is the individual teams winning percent - to inputs - which for the simple model are the individual teams points scored and points surrendered. This should be rather familiar, since the definition of a win is having greater points scored than surrendered.

With this definition in mind, I then statistically test this simple model of NCAA football production function using each NCAA FBS team's winning percentage, points scored and points surrendered for the 2008 and 2009 NCAA FBS seasons. Since during the 2008 and 2009 season there are 120 NCAA FBS schools, this results in 240 rows of data, 120 for the 2008 season and 120 for the 2009 season. After "running a linear regression" on this data, where the dependent variable is winning percentage and the independent variables are points scored and points surrendered, I highlight the following three results of the linear regression.

1. The coefficient (or weight) on points scored = 0.001 and is statistically significant above the 99% level of confidence.
2. The coefficient (or weight) on points surrendered = -0.001 and is statistically significant above the 99% confidence level.
3. The R-square and adjusted r-square (for those interested) is 0.83, which can be interpreted as the variation in points scored and points surrendered "explain" 83% of the variation in the teams winning percent over those two years.

Given that the model is rather simplistic, why would I be interested in this type of analysis? There are two fundamental reasons. The first is to determine if there is a statistical difference in terms of winning percentage between offense and defense. As we can see from the estimated coefficients, they are equal in terms of determining winning percentage, which makes the analysis easier under the complex model. The second is to show that just using point spread (the difference between points scored and points surrendered) is inferior to the complex model.

While the model does rather well at explaining why teams win, the big problem with this simple model is that it does not allow me to investigate what actually happens on the field to how productive the team actually is. In other words, the model just says that if a teams scores 3 additional points then that will result on average a 0.003 increase in winning percentage or if the team allows their opponent to score 3 more points then that will result on average a 0.003 decrease in winning percentage.

What is missing is what happens if the team throws an interception, or allows a sack or recovers a fumble? How do on field actions impact the production/efficiency of the team's offense or defense?

Periodically, over the rest of this month, I hope to address the "complex invasion sport production function" model - i.e. the model that allows me to rank NCAA FBS teams each week.

Thursday, November 4, 2010

Measuring Conference Strength - Overall

So, given the conference offense and defense strengths, I have ranked the conferences (as of October 30, 2010) by taking the average conference offensive productivity and subtracted the average conference defense productivity and come up with the average conference overall productivity. Since the percentage differences do not make sense to add together, I have just ranked them from highest to lowest.

The biggest surprise (to me) is that the Big East is actually better overall than the SEC, Pac 10 or Big 12. This is mainly due to the Big East's lead in defense relative to all the other conferences. I suspect that this will decline as the season progresses, but it may not.



Overall
1
Big 10
2
Independent
3
Big East
4
SEC
5
Pac 10
6
Big 12
7
ACC
8
WAC
9
Mountain West
10
CUSA
11
Sun Belt
12
Mid American

Measuring Conference Strength - Defense

In the previous post, I calculated conference strength for the offenses of the 11 NCAA FBS conferences and the three independent schools. Now I want to turn my attention to the defense. In order to do so, I will have to calculate the defensive productivity of each conference, and it is done similarly as for the offensive side of the ball, with one big exception. On the defensive side of the ball, the lower the number the better the defense. In other words the more defenses stop their opponents from scoring (and on field actions that lead to their opponent scoring) the better the overall defense. Thus better defenses have smaller numbers overall.

Given that better defenses have smaller numbers overall, worse defenses have higher numbers, so a conference with a defense of 1.500 is 50% as productive as the most productive defense.

So with that in mind, here are the conference strengths as of October 30, 2010.

Conference Defense
Big East
1.000
Independent 1.204
Big 10
1.215
SEC
1.275
ACC
1.279
Big 12
1.343
Pac 10
1.382
WAC
1.463
Mountain West 1.483
Sun Belt
1.494
CUSA
1.589
Mid American 1.591

Measuring Conference Strength - Offense

One of the big debates among the NCAA FBS football landscape is over how to adjust for different conferences. In other words, teams like Boise State or TCU are discounted due to the perception that they play in a conference that is weaker than teams that automatically qualify for a BCS bowl. So the first question is, is that true? And the second question is if it is true, how much of a discount should be placed on the conference?

Now the NCAA FBS production model includes conference fixed effects - which is a fancy way of saying the I control for teams in the same conference and thus control for teams in different conferences when calculating offensive and defensive productivity. But even since that happens, not all conferences are equal, right. Right!

Let me first explain how I calculated conference strength for the offense. After I calculated the offensive total productivity (model will be blogged about in the coming weeks), I added the total productivity of each school in each conference, which gives me total conference offensive production. Given that not every conference has the same number of teams, I then divided the total conference offensive production by the number of schools to get the average conference offensive production.

Once that is done, I took the conference that had the highest average conference offensive production and divided each conference by the largest average conference offensive production to get a number that would be between zero and one. On the offensive side, the higher the average conference number the more productive is that conference. Thus if a conference has an average conference offensive production equal to 0.90, then I conclude that conference is 90% as productive as the best conference. The conference with the highest average conference offensive production will always be 1.00 or 100%.

Calculation Summary (for Offense):
1. Add all the offensive production number for each team in a given conference which I call Total Conference Offensive Production.
2. Divide Total Conference Offensive Production by the number of teams in their conference, which I call Average Conference Offensive Production.
3. Rank Average Conference Offensive Production from largest to smallest.
4. Divide each conference by the largest Average Conference Offensive Production, which is the Percentage Conference Offensive Production.

So here are the conference strengths for the offense as of October 30, 2010.

Conference
Percentage
Big 10
1.000
Pac 10
0.981
Independent
0.976
Big 12
0.941
SEC
0.933
WAC
0.875
ACC
0.813
CUSA
0.804
Big East
0.753
Mountain West
0.752
Mid American
0.664
Sun Belt
0.629

Wednesday, November 3, 2010

Welcome!

Over the last few months I have been working on an NCAA Football Bowl Subdivision (FBS) model of offensive, defensive and total production in order to determine a ranking for FBS schools that relates on field performance to how FBS schools are ranked. I have written about this on my other blog, but I thought that I would create a new blog devoted just to NCAA college football analysis. Thanks to cfbstats.com for posting the NCAA FBS data.

So to get us started, I will link to the other posts from my Hawkonomics blog and give the top 25 as of last week. Later this week, I will be posting about measuring conference strength, and then giving the back story of how I use statistics to rank the teams below.

Top 25 for the first week of the BCS (October 16, 2010).
Top 25 for the second week of the BCS (October 23, 2010).

Here is my ranking (re-posted) through October 30, 2010.

Rank
School
1
Oregon
2
TCU
3
Boise State
4
Ohio State
5
Utah
6
Alabama
7
Iowa
8
Arizona
9
Hawai'i
10
Nebraska
11
Wisconsin
12
Auburn
13
Nevada
14
Oklahoma
15
Stanford
16
UCF
17
Oklahoma State
18
Virginia Tech
19
Miami (Florida)
20
West Virginia
21
Florida State
22
Northern Illinois
23
Michigan State
24
Baylor
25
Missouri