Tennessee and Close Game Variance Ratio

One thing that has always fascinated me is the propensity for some college football teams to have wild swings in wins from season to season. Phil Steele started this metric, and I gave it a name and update all of you. Now, before I go much farther into this, let me link you to 2007's CGVRs and also link you to an explanation of CGVRs in case this is all new to you.

Even more interesting to me than CGVRs is trying to figure out why teams that fall in the parameters don't produce results that follow the theory. Every year you have a handful of teams who won more than their fair share of close games, only to have a better season the next time around. Tennessee has been one of those teams. Here are the W/L records, CGVRs, and following W/L records for every Volunteer season under The Papa.

Year	Record	CGVR	Expectation	Next Season	Correct?
1992	9-3	E		10-2	N/A
1993	10-2	-1		8-4	N/A
1994	8-4	-2		11-1
1995	11-1	+4		10-2
1996	10-2	E		11-2	N/A
1997	11-2	+4		13-0
1998	13-0	+5		9-3
1999	9-3	-1		8-4	N/A
2000	8-4	+2		11-2
2001	11-2	+3		8-5
2002	8-5	+1		10-3	N/A
2003	10-3	+2		10-3
2004	10-3	+5		5-6
2005	5-6	E		9-4	N/A
2006	9-4	+2		10-4

What does this tell us? Not a lot more than we already assume through CGVRs. There have been 9 years where a Fulmer-led team had a CGVR that fell within the parameters. In 6 of them, or 66.7%, resulted "correctly." What I find a tad more interesting is that in 8 of those 9 seasons where UT had a CGVR in the parameters, 8 of them we had positive CGVRs. That tells me that The Papa wins more close games than he loses (which I really like). In 5 of those 8, or 62.5%, our team followed the theory and had the same or worse record the following season.

The exciting prospect of this (sort of) is that Fulmer is better than the rest of the country, on average, at not following a "lucky" season with a "lesser" season. The national average for positive CGVRs is 77% of them either have the same or worse record the next year, whereas Fulmer's teams have the 62.5% ratio.

What does all of this mean for 2008? Well, if you click the link at the very top for the 2007 CGVRs, you'll see that Tennessee's 2007 squad racked up 3 net close wins, which is considered "lucky." Nationally over the last 6 years, 73% of teams with a +3 CGVR have finished with the same or worse record the following year; that's not good news for the prospects of our 2008 season. However, because of the comparison above, I'm not panicking, as The Papa has shown he bucks the trend, at least a little bit.

In the end, it boils down to this: the +3 CGVR is a sign that Tennessee won't improve on the 10-4 record of 2007.

APP Breakdown: The SEC

I have finished the SEC analysis for Adjusted Performance Percentage. Keep in mind that a score of 100 means 100%, or that team performed on par with what their opponents allowed. Anything higher means they excelled, but anything lower means they underperformed. We'll start with Total Offense Adjusted Performance Percentage:

1. LSU 116.1
2. Kentucky 113.9
3. Arkansas 111.1
4. Florida 104.0
5. Alabama 103.7
6. Vanderbilt 102.9
7. South Carolina 100.7
8. Georgia 97.9
9. Tennessee 97.5
10. Auburn 91.0
11. Mississippi State 83.8
12. Mississippi 76.1

I think the two biggest surprises we see here are that Alabama was 5th in the conference on offense, while Tennessee was a dismal 9th! A lot of fans point to David Cutcliffe's return as the reason Tennessee went from 5-6 to 9-3, but during SEC play, Tennessee didn't play as well as we all thought.

Best Rush Offense (Best ROAPP):

Not surprising - Arkansas not only racked up a lot of yards on the ground, but scored a 164.9 for their ROAPP. Wow! Was it due to Darren McFadden and Felix Jones, or the 3 monster offensive linemen who just graduated? We'll find out in 6 months.

Worst Rush Offense (Worst ROAPP):

Again, no surprise here. Tennessee's terrible run blocking and inconsistent tailback play created a 64.5 ROAPP, meaning Tennessee performed 35.5% below what their opponents were allowing. Awful.

Best Pass Offense (Best POAPP):

Did you see this coming? Surely you had to, as they passed for 2,178 yards in the SEC, but a mere 178 more yards than LSU. What am I getting at? Kentucky scored a 133.7 in POAPP, while LSU and Tennessee scored 118.2 and 118.3, respectively. That's a 15.5% difference! Kentucky's pass offense wasn't just better than everybody else in the SEC, it was miles ahead! Oh, and they return every important skill position player on offense. Go ahead, say they won't make a bowl again.

Worst Pass Offense (Worst POAPP):

The arrival and play of Yaw Yaw Yaw Brent Schaeffaw (Brent Schaeffer for those who don't know Orgeron-ese) was nothing short of a nightmare. Let me put it this way, Tennessee fans: the Rebels' pass offense was WORSE than Tennessee rush offense, if you can believe it. A wretched 63.5POAPP is more than 12% worse than the Razorbacks' score. Ouch.

Next, I did the same analysis for the defenses in the SEC. There are a few changes when you compare it yards allowed. Here is Total Defense Adjusted Performance Percentage:

1. LSU 121.3
2. Florida 112.1
3. Georgia 109.6
4. Alabama 106.1
5. Auburn 104.7
6. Arkansas 103.0
7. Mississippi State 102.7
8. Tennessee 102.4
9. South Carolina 98.9
10. Mississippi 94.9
11. Vanderbilt 90.0
12. Kentucky 72.9

What stands out? Alabama at #4 is the biggest surprise. Tennessee at #8 is actually better than I expected. Not in terms of SEC rank, but in terms ofTDAPP score. They actually performed 2.4% better on defense than what their opponents' offenses were used to. The 4 teams to be in the bottom of bothTOAPP and TDAPP are Ole Miss, Mississippi State, South Carolina, and Tennessee.

Best Rush Defense (Best RDAPP):

The Gators had 4 or 5 NFL draft picks on the defensive line in 2006, with 3 future picks at linebacker behind them. Only four teams had better than a 100 score inRDAPP: Florida, LSU, Mississippi State, and Alabama. Of those 4, the first 3 had amazing RDAPP scores - yes, Miss State was amazing in rush defense (134.9).

Worst Rush Defense (Worst RDAPP):

Kentucky couldn't stop anyone who wanted to rush the ball last season. Their RDAPP was 72.1, which means that they not only didn't stuff the run well, but they also played teams with very good running offenses. They'll need to get tougher in the trenches to have a chance in the SEC East.

Best Pass Defense (Best PDAPP):

The Tigers had studs all across the line last season, the year AFTER they lost 3 starters to the NFL draft. Their passing opponents were all over the board in terms ofPOAPP, so it just goes to show that the Tigers played extremely well when teams had to pass.

Worst Pass Defense (Worst PDAPP):

Deja vu? Kentucky's pass defense was a smidgen better than their rush defense, but still horrid. A 73.4 PDAPP is 15% worse than their closest competition for absolutely blowing coverage. Their entire defense was a joke, and without any improvement, they'll finish 6-6 next season. If you gave Kentucky a new defense, say just Ole Miss (94.9TDAPP), Kentucky would have fought Florida for the SEC East Division Title. I'm not joking.

Lastly, I wanted to see who the best overall team in the SEC was, statistically. I added each team's TOAPP with their TDAPP to create Complete Adjusted Performance Percentage:

1. LSU 118.7
2. Florida 108.1
3. Arkansas 107.1
4. Alabama 104.9
5. Georgia 103.8
6. Tennessee 100.0
7. South Carolina 99.8
8. Auburn 97.9
9. Vanderbilt 96.5
10. Kentucky 93.4
11. Mississippi State 93.3
12. Mississippi 85.0

Again, how do we use this for preseason predictions? Look at Alabama - 4th best team in the SEC, but they went 6-6. Expect a big jump in wins, especially when you consider their -2 CGVR. Couple that with new head coach Nick Saban and a boatload of returning talent and you've got a potential surprise SEC West Champion. Georgia is another team I expect to win more games next year. They lose a back-up running back to the draft and a few, key defensive players, but the core of both sides of the ball will be back and stacked. These statistics imply that Kentucky should take a step back next season, but they lose little to the draft, so I'm a little confused on their progress right now.

In the end, these stats don't win games, but they can absolutely give us ideas into who we can expect to show up in 2007. As for Tennessee, I expect the same old, same old.

Interesting Stats

Head over to Matt Melton's blog, Statistically Speaking, and check out his discussion on his SDPI statistic, specifically the SEC Rewind. Matt's found a pretty nice tool in terms of performance evaluation, as well as a tool that you can use in your repertoire for preseason predictive measures.

Matt has also helped me on my project, Adjusted Performance Percentage, or APP, and it's many branches. You'll be able to find things like ROAPP, or Rush Offense Adjusted Performance Percentage. Anyway, I have the SEC done, so expect that pretty soon.

APP, ROAPP, POAPP, etc are easy to calculate. You find the average yards allowed rushing (if you decide you want to know the ROAPP) by the 8 teams YOUR team faced in the SEC (adjust it for conferences who play more or less than 8 games). Then, take your school's rushing offense and divide it by the average allowed by the 8 opponents.

For Tennessee, as an example, their 8 SEC opponents allowed an average of 987 yards rushing in the SEC, while Tennessee rushed for 645 yards. 645/987 = 65.3%, or the worst ROAPP in the SEC.

After finishing the SEC portion of the project, I found that these ratings differ very little than if you just ranked every SEC team by the number of yards gained or allowed. It DOES change, however, and in the end, I will be adding the Total Offense APP (TOAPP) and Total Defense APP (TDAPP) to create Complete APP (CAPP). As a preview, Tennessee rates lower than Alabama in CAPP for 2006.

Until next time, here is Phillip Fulmer's Greatest Locker Room Moments Volume 1:

Narrowing the Field

I was doing some quick research on the National Champions of this decade, and I noticed a little bit of a trend for these teams. Obviously, seven national champions isn't the best of sample sizes, but in the current landscape of college football, it's enough for me to mention.

Here is what I found: 6 of the last 7 BCS Champions returned 6 defensive starters and six offensive starters, with one of the six being the quarterback. Ohio State was the lone outlier.

So, who are the the teams who meet this criteria as of spring practice (If your team is listed, and I tabulated incorrectly, let me know)?

Ole Miss
South Carolina
Kentucky
Vanderbilt
Purdue
Northwestern
Indiana
Texas A&M
Oklahoma State
Colorado
Kansas
Miami
Virginia Tech
North Carolina
Florida State
Boston College
Southern Cal
California
Arizona State
Oregon
UCLA
West Virginia
South Florida
Connecticut
Cincinnati

That's a lot of teams! However, to narrow these teams down to a much more defined list, I found that every single BCS Champion of this decade had 7 or more wins the previous year. From Oklahoma to Florida, they all had at least 7 wins. So, the teams like Indiana and Connecticut will be removed from the list, leaving the pool of teams that will win the National Title as follows:

All logos used from http://www.gnu-bee.com/football/index.shtml; sorry South Florida and Cincinnati - I know you win 7+ games and fit the criteria, but this is much prettier ... and do you REALLY think you have a chance?

2006 CGVRs

Sorry this took so long to get to you loyal readers (reader?). I'm not proficient enough in HTML to make a big table for this, but maybe in the future, it will look better. Anyway, here are the qualifying CGVRs for 2006.

+4 or more:
Wake Forest +5 (11-3)
Maryland +5 (9-4)
Auburn +5 (11-2)
Florida +5 (13-1)
Oregon State +4 (10-4)
Rice +4 (7-6)
Boise State +4 (13-0)
Iowa State +4 (4-8)
Arkansas State +4 (6-6)

+3:
Kansas State +3 (7-6)
Texas Tech +3 (8-5)
Wisconsin +3 (12-1)
Notre Dame +3 (10-3)
Ohio +3 (9-5)
Troy +3 (8-5)

+2:
Boston College +2 (10-3)
Texas A&M +2 (9-4)
LSU +2 (11-2)
Tennessee +2 (9-4)
Georgia +2 (9-4)
Kentucky +2 (8-5)
Ohio State +2 (12-1)
Purdue +2 (8-6)
Rutgers +2 (11-2)
Cincinnati +2 (8-5)
Southern Miss +2 (9-5)
Idaho +2 (4-8)
Louisiana-Lafayette +2 (6-6)

-4 or more:
North Carolina State -5 (3-9)
Memphis -5 (2-10)
Eastern Michigan -5 (1-11)
Florida International -5 (0-12)
Louisiana-Monroe -5 (4-8)
New Mexico State -4 (4-8)
Colorado -4 (2-10)

-3:
Duke -3 (0-12)
Missouri -3 (8-5)
Oklahoma State -3 (7-6)
Vanderbilt -3 (4-8)
Michigan State -3 (4-8)
Illinois -3 (2-10)
Air Force -3 (4-8)
Alabama-Birmingham -3 (3-9)
Miami (OH) -3 (2-10)

-2:
Clemson -2 (8-5)
Florida State -2 (7-6)
Alabama -2 (6-7)
Mississippi State -2 (3-9)
Iowa -2 (6-7)
Minnesota -2 (6-7)
San Diego State -2 (3-9)
UNLV -2 (2-10)
Houston -2 (10-4)
Nevada -2 (8-5)
North Texas -2 (2-10)

Expect the teams listed with positive CGVRs to be unable to improve on their 2006 records, while teams listed with negative CGVRs to be able to improve on theirs.

Close Game Variance Ratio

One of the resources I use is something I call the "Close Game Variance Ratio" or CGVR. It's an idea I picked up from Phil Steele's preseason magazines, and so far, it's been a pretty strong predictive tool.

The way to find a team's CGVR is to take the difference of games won by 8 points or less from the games lost by 8 points or less. For instance, Tennessee in 2006 beat Air Force by 1, Alabama by 3, South Carolina by 7, and Kentucky by 5, giving Tennessee 4 close wins. UT lost to Florida by 1 and LSU by 4, giving UT 2 close losses. Two from four is +2. So, Tennessee's CGVR is +2.

The idea behind the theory is that teams with higher CGVRs were extremely lucky, or, to sound a little more serious, had positive variance. I believe variance tends to balance itself out, usually from year to year, but sometimes longer. Conversely, teams with lower, negative CGVRs will usually get no worse in record. A 4-7 team with a -3 CGVR will probably improve their record the next season, or at the very least, stay at the same level.

The evidence? I still consider my research to be done over a small sample size, but it's looking pretty good so far. I calculated every team's CGVR for every season from 2001-2005, and so far, the outcomes are pretty telling. First, we'll look at teams with high CGVRs:

+4 and up - Has occurred 34 times, with 91% of the teams having the same or worse record the next season
+3 - Has occurred 24 times, with 75% of the teams having the same or worse record
+2 - Has occurred 49 times, with 71% of the teams having the same or worse record

Solid evidence. Now, for negative CGVRs:

-4 and worse - Has occurred 22 times, with 86% of the teams having the same or better record
-3 - Has occurred 37 times, with 68% of the teams having the same or better record
-2 - Has occurred 50 times, with 71% of the teams having the same or better record

As far as I can tell, this theory can continually be a solid resource in preseason predictions. Obviously, preseason predictions are about as common as the sniffles, but it's still fun to be right. If you used this resource before the 2006 season, you could have predicted Arkansas's improvement (-4 in 2005), Alabama's and Penn State's record drop (both +2 in 2005), Oregon's troubles (+3), Arizona's 3 win jump (-4), and other teams' fortunes.

Later this week, I will outline which teams from 2007 fall into the important CGVR ranges.

Players of the Last Week You Haven't Heard of

We've all heard about Troy Smith, Mario Manningham, Erik Ainge, Garrett Wolfe, etc etc etc. Each week, I think I'd like to point out some very productive players who won't garner facetime on ESPN, FSN, houserockbuilt, and others. :)

• Bryan Cupito - Minnesota QB: Against Penn State, Cupito completed 25 of 36 passes for 347 yards and 2 TDs against an athletic PSU defense. With the Gophers unable to run the ball consistently, Cupito took the offense on his shoulders and carried the team to a very close loss. And Gopher fans, please quit sobbing about the bad call in OT ... you were lucky to get that far.
• La-Rod Stephens-Howling - Pittsburgh RB: Jesus H. Christ Almighty! I rejoice for your performance, but damn your name all to Hell. If I ever mention you again, La-Rod, you will be known as LaRSH ... yes, that's easier. LaRSH's previous season high in 2006 was 66 yards against the famous Citadel defense. Against the Orange of Syracuse, he busted out for 221 yards on 27 carries. One was 70 yards, but even 151 on 26 is incredible. Good job last week LaRSH.
• P.J. Hill - Wisconsin RB: Another 200+ performance from an unknown stud RB. Hill had actually shown a lot of skills earlier this season, rushing for over 100 in 4 of 5 games. But against Northwestern, he did whatever he wanted. 35 carries, 249 yards, and a touch. He's a big boy, too, but he's short. 5'11" and 242 yards. During his long run, he was caught from behind (because he's big and slow for a RB), but the defender looked like a lamb hitting the hood of a car.
  
• Earl Bennett - Vanderbilt WR: One of my favorite unknown players in the SEC. Bennett had 876 yards receiving last season ... as a freshman(!) ... in the SEC (!!!). Against Ole Miss this past weekend, he had 10 catches for 179 yards. ONE HUNDRED AND SEVENTY NINE YARDS! Plus, he told this guy to talk to the hand ...

This will become a weekly update and will probably become more streamlined. So, for the 0 of you who frequent this blog, stick with it.