FanPost

QB Efficiency and Success - Analysis and Projection

About a month ago, Dr. Saturday published an article discussing QB Rating and team success. The article, 'The Big Picture: Start paying more attention to pass efficiency, the most useful statistic in football.' argues that passer efficiency is the 'single best reflection of success anywhere in your average box score'... better than total offense, total defense, turnover margin, etc. While the article doesn't necessarily argue cause and effect (not all top teams lead the nation in passer efficiency - see LSU), the evidence is convincing that if you have an efficient QB, your team will succeed.

Top 8 teams in passing efficiency 2011

  1. Auburn (Cam Newton). 14-0
  2. Boise State (Kellen Moore). 12-1
  3. Stanford (Andrew Luck). 12-1
  4. Wisconsin (Scott Tolzien). 11-2
  5. Alabama (Greg McElroy). 10-3
  6. TCU (Andy Dalton). 13-0
  7. Arkansas (Ryan Mallett). 10-3
  8. Ohio State (Terrelle Pryor). 12-1

QB Rating = (8.4 x yds + 330 x TD + 100 x completions - 200 x interceptions)/Attempts

Before we further examine the premise of Dr. Saturday's article, I'd like to introduce Tomahawk Nation to a stat used by SB Nation's California site (californiagoldenblogs.com). They use a stat called Ultimate Adjusted Yards Per Attempt (uAYA) that is designed to account for both QB passes and runs. I personally like uAYA better as it accounts for mobile QBs and for heavy QB-running teams like Nevada or Nebraska, and it also gives a number that is more meaningful (yards per play vs. random calculated number).

uAYA = (Pass Yds - Sk Yds + Rush Yds + 20 * Pass TD + 18 * Rush TD - 45 * INT - 25 * Fum) / (Pass Att + Sacks + Rush Att)

The reasoning behind the specific numbers for passing touchdowns, rushing touchdowns, interceptions, and fumbles in the uAYA equation can be found through the links at the californiagoldenblogs.com article. In summary, passing touchdowns are worth 20 yards, rushing TDs worth 18, interceptions incur a 45 yard penalty, and fumbles incur a 50 yard penalty (which is halved base on a 50% fumble recovery chance).

Top 10 QBs uAYA 2010

  1. Kellen Moore (Boise St) 10.5995 (12-1)
  2. Andrew Luck (Stanford) 9.6604 (12-1)
  3. Cam Newton (Auburn) 9.1397 (14-0)
  4. Andy Dalton (TCU) 9.1294 (13-0)
  5. Colin Kaepernick (Nevada) 9.7368 (13-1)
  6. Ryan Mallett (Arkansas) 8.7187 (10-3)
  7. Chandler Harnish (NIU) 8.5944 (11-3)
  8. Greg McIlroy (Alabama) 8.4933 (10-3)
  9. Ryan Lindley (SD St) 8.475 (9-4)
  10. Scott Tolzien (Wisconsin) 8.375 (11-2)

Dr. Saturday's article really got my mind going about FSU and EJ Manuel. If QB efficiency was that closely related to success, how does FSU look? Why weren’t we on this list in 2010? Why were these teams on the list in 2010? How will FSU and opponents look in 2011? Will EJ be able to jump up the efficiency ranks?

Before we dig in to 2010, I'd like to note that this is not an attempt to use Passer Efficiency to determine who the ‘best offense’ was.  For that, it is better to use advanced metrics that account for strength of schedule, 'garbage time', etc.  However, Dr. Saturday was not trying to list the top offenses, but rather examine the correlation between winning and passer efficiency and that is what will be discussed here.

Also, this is not an attempt to determine who the best QB is, as there is obviously more that goes into passing efficiency than just QB skill.  There are certainly a variety of factors that help QB efficiency: great defense; talented, experienced receivers; a healthy, veteran offensive line.  Some teams get it with great quarterback play (Stanford, Auburn); some get it with horrible strength of schedule (Boise, TCU); some by a great running game (Wisconsin, Alabama); or any combination of these things.

With those qualifiers out of the way, it seemed pretty apparent that uAYA and QB rating are pretty closely related to overall team success, but I wanted to see how closely they are related to offensive strength, so I compared it with more complex statistics. (overall F/+, offensive F/+). 

I also wanted to look at other factors that could affect QB efficiency (Offensive Strength of Schedule, Defensive F/+) to see if teams were only efficient because they played an easy set of defenses, or if they were able to be efficient because they could rely on their defense and not be pressed into trying to score quickly.

 

uAYA

QB Rating

Off. F/+

Off. SOS

Defensive F/+

Kellen Moore

10.5995

182.6

9

111

1

Andrew Luck

9.6604

170.2

2

71

15

Cam Newton

9.1397

182

1

3

31

Andy Dalton

9.1294

166.5

15

98

4

Colin Kaepernik

8.7368

150.5

8

91

68

Ryan Mallett

8.7187

163.7

5

4

28

Chandler Harnish

8.5944

157.8

29

116

64

Greg McElroy

8.4933

169

4

15

11

Ryan Lindley

8.475

149.4

30

74

63

Scott Tolzien

8.375

165.9

6

50

36

Ricky Stanzi

8.3053

157.6

34

58

20

Brandon Weeden

8.1515

154.1

11

48

29

Tyrod Taylor

8.128

154.8

10

18

23

Darron Thomas

8.0441

151

16

73

13

Denard Robinson

8.0238

149.6

3

14

115

Tyrelle Pryor

7.9541

157.9

17

46

3

Aaron Murray

7.9441

154.5

27

31

44

Taylor Martinez

7.5335

138.8

50

42

5

Mike Hartline

7.3593

146.4

23

30

78

What can we gather from this data?

  • QB efficiency corresponds well to offensive strength.  Out of the top nineteen QB's in uAYA, thirteen led their offense to a top a 20 F/+ finish
  • Facing tough defenses didn't seem to hurt as 5 of the top 19 teams faced a top 20 set of defenses.
  • Bad defenses didn't exclude high efficiency numbers (Michigan, Northwestern).  But this shows what FSU fans know all to well, that an efficient offense alone will not win you a lot of games.

Next, I tried to look for what the top teams had that made it possible to be so efficient in 2010?

 

 

uAYA

Returning Off. Starters

Returning OL starts

Return Rec. Yards

Kellen Moore

10.5995

11

94

91%

Andrew Luck

9.6604

8

66

88%

Cam Newton

9.1397

7

111

89%

Andy Dalton

9.1294

9

67

88%

Colin Kaepernik

8.7368

9

46

86%

Ryan Mallett

8.7187

9

77

83%

Chandler Harnish

8.5944

8

52

73%

Greg McElroy

8.4933

8

46

78%

Ryan Lindley

8.475

9

67

76%

Scott Tolzien

8.375

10

107

73%

Ricky Stanzi

8.3053

6

36

68%

Brandon Weeden

8.1515

4

12

62%

Tyrod Taylor

8.128

8

54

95%

Darron Thomas

8.0441

10

85

66%

Denard Robinson

8.0238

7

82

72%

Tyrelle Pryor

7.9541

9

93

77%

Aaron Murray

7.9441

10

155

89%

Taylor Martinez

7.5335

8

75

64%

Mike Hartline

7.3593

5

31

73% 

 

  • I honestly thought returning OL starts would play a bigger role in QB efficiency than is seen here.  Returning OL starts had a wide range.  Average OL returning starts was 74.
  • I was surprised about OL starts.  I was shocked when I looked at the returning receiving yards as it seems to be the most important determinant in QB efficiency.  The list is very top heavy for returning receiving yards, and at first glance seems linear from top to bottom.  The average returning receiving yards for the top 20 teams was 76%. 
  • All but three teams returned 7 or more offensive starters.  

After looking at the nation's top teams of 2010, I looked specifically at FSU and opponents.

 

                                                                   Conference Play Only

 

 

Atlantic

uAYA

 

Coastal

uAYA

FSU (6-2)

6.85

 

VT (8-0)

8.46

Maryland (5-3)

6.39

 

UNC (4-4)

6.42

NCSt (5-3)

5.99

 

Miami (5-3)

5.68

BC (4-4)

3.64

 

GT (4-4)

5.13

Clemson (4-4)

3.4

 

Virginia (1-7)

4.75

Wake (1-7)

3.39

 

Duke (1-7)

4.32

 QB Efficiency again corresponds well to success.  For the complete conference season average, only UNC "underperformed" compared to their uAYA rank, but I think TJ Yates probably had more variability in uAYA than any other QB... great performances against FSU and Virginia (11.178 and 13.88 respectively) and horrible performances against Miami and VT (1.35 and 0.829 respectively)

 

                                                         Single game QB Efficiency

 

 

FSU uAYA

Opponent uAYA

FSU QB Rating

Opponent QB Rating

Oklahoma

3.91

9.74

58.9

187.8

BYU

6.97

0.273

141.98

77.03

Wake

7.88

2

137.87

93.46

Virginia

5.359

4.489

119.82

112.14

Miami

7.32

4.345

146.06

76.38

BC

1.795

2.344

109.29

70.75

NCSt

5.718

6.133

131.30

122.33

UNC

7.489

11.178

164.93

202.22

Clemson

6.35

3.229

127.7

100.17

Maryland

4.343

3.176

121.46

104.21

UF

9.056

2.146

185.26

91.73

VT

6.6676

9.359

150

178.54

USC

4.969

2.905

142.4

98.27


In single game numbers, both uAYA and the traditional QB rating were very well correlated with success, with each measure going 12 of 13 for predicting the winner of the game.

 

                                                                              Complete season numbers 2010

 

 

uAYA

Passer rating

Returning Offensive Starters

Returning OL Starts

Returning Receiving Yards

Offensive SOS

Tyrod Taylor

8.128

154.8

8

54

95 %

18

Landry Jones

7.2275

146.3

9

42

81 %

19

EJ Manuel

7.0821

153.3

8

146

45 %

2

TJ Yates

6.7625

145.5

8

57

60 %

25

Danny O'Brien

6.7609

134.5

7

38

89 %

11

Russell Wilson

6.2522

127.5

7

30

78 %

9

Stephen Morris

5.9944

125

6

47

76 %

6

Christian Ponder

5.8546

135.7

8

146

45 %

2

Josh Nesbitt

5.7232

105.4

7

37

32 %

29

Mark Verica

5.484

122.8

7

56

56 %

44

Sean Renfree

5.1864

120.7

8

74

84 %

37

Kyle Parker

5.1186

117.2

6

88

34 %

5

Jacory Harris

4.8863

116.6

6

47

76 %

6

John Brantley

4.4094

116.4

6

87

25 %

23

Tanner Price

4.1804

106.8

6

45

91 %

17

Chase Rettig

3.9394

105.5

8

108

22 %

32

Dave Shinskie

3.9167

106.7

8

108

22 %

32

Tajh Boyd

3.547

107.7

6

88

34 %

5

While this table might not seem as useful as those above because there aren't any evident trends, this is expected because there is a much larger difference in talent and quality of teams than in the charts above. 

But if we think of some of the 'surprises' of 2010, I think we can again draw some conclusions.

Surprise #1 - UF's offense falls apart - Again the returning receiving yards stand out.  While Meyer and company certainly didn't help matters with the QB shuffling, they were fighting an uphill battle after losing Tebow and the large majority of receiving yards.

Surprise #2 - Danny O'Brien played extremely well for a freshman QB.  Not only did he have an experienced receiving corp with 89% of receiving yards returning, he had very good talent at the position with Torrey Smith.

Surprise #3 - Christian Ponder was the heavy preseason favorite for ACC Offensive POTY, and many FSU fans, myself included, geared up for what we thought would be a great year and potentially a run at the Heisman trophy for CP7.  Ponder played well, but he was fighting a losing battle, playing injured with only 45% of receiving yards returning. He fell well short of both Heisman and ACC Ofenisve POTY, but the talent was still there, as Coach Fisher tried to tell the media all year long.

 

Hopefully, by now you're convinced that QB efficiency is a fairly good predictor of success.  Hopefully you also see that there seems to be fairly strong indicators of what will make a QB efficient in any given year:
 
1. Overall talent is obviously necessary - no amount of returning receiving yards, returning OL starts, etc. is going to allow a Duke or Wake QB to lead the ACC in efficiency
 
2. QB talent is also necessary - Even surrounded by talent and experience, a poor QB is going to be a poor QB (see Jacory Harris)

 3. QB talent is not sufficient - Two very talented QBs 'struggled' in 2010 relative to their talent.  Ponder (2010 uAYA = 5.855; 2009 uAYA ~ 7.2) was 5th in the ACC for QB efficiency behind Taylor, Yates, O'Brien, and Wilson, but is clearly the best QB of the lot.  Blaine Gabbert (2010 uAYA = 5.8737; 2009 uAYA = 7.16) was 9th in efficiency in the Big 12 in 2010.  Gabbert, the #10 overall NFL draft pick only had 22% returning receiving yards in 2010 and his efficiency dropped accordingly.

4. Returning Receiving yards seems to have the largest effect on QB efficiency.  It would make sense to me that Offensive Line quality would probably play a larger part in QB efficiency than seen here, and I think it is likely that Returning OL Starts isn't the best predictor of O-Line quality.  Until a better predictor of O-Line quality is found, returning receiving yards will be more important.

 

Now it's time for the projection.  Looking to 2011, what do FSU and opponents return?

 

Returning Offensive Starters

Returning Receiving Yards

Returning

QB Rating

Returning

QB uAYA

FSU

8

100%

153.3

7.0821

VT

6

91%

n/a

n/a

Duke

7

85%

120.7

5.1864

Clemson

8

82%

107.7

3.547

Wake

8

78%

106.8

4.1804

Boston College

8

78%

106.7

3.939

Virginia

8

63%

n/a

n/a

Miami

6

58%

116.6

4.8863

NC St

8

44%

n/a

n/a

Maryland

7

36%

134.5

6.7609

UF

7

81%

116.4

4.4094

Oklahoma

8

74%

146.3

7.2275

 

This might be an overly simplistic approach at a projection for 2011, but it looks as though FSU will run away from the rest of the ACC in QB efficiency in 2011.  VT could challenge for the top spot if Logan Thomas plays well, and I fully expect to meet them in the ACC Championship Game again.  But the Atlantic Division should be a one-horse race, as EJ Manuel and FSU have a chance to dominate the Atlantic Division in QB Efficiency: 

1) FSU is the leader at overall talent, QB talent, returning offensive starters, and returning receiving yards. 

2) Three of the five other teams in our division will be in the first year with a new offensive coordinator (Clemson, Maryland, and BC). 

3) Maryland and NCSt both return < 45% of receiving yards from 2010.

My prediction is that EJ will average 8+ for uAYA.  I would be very surprised if Miami, BC, Wake, Virginia, Duke, NC State or Maryland finished with a uAYA > 6.  Clemson and UF are harder to judge.  Both teams have talent and both return > 80% of their receiving yards.  Will they be able to pick up enough of the offense in the first year to be efficient?  I doubt it, but they could surprise me.

 

Final Thoughts:

Many FSU fans criticized Ponder heavily for the decreased production in 2010.  Meanwhile, Coach Fisher continuously told the media Ponder was as good as ever, but our 'knucklehead' receivers weren't where they needed to be, literally and figuratively.  I think most college football fans, myself included, truly don't understand how details really affect offensive performance.  We're stuck just seeing the results, like when it looks like Ponder left the pocket too early, but in reality he had to because his inexperienced receivers ran right into coverage.  Or when Ponder double pumps and looks hesitant to throw, but actually he is setting up to throw before the receiver makes his break, and the inexperienced receiver breaks 2-3 steps early forcing Ponder to readjust and wind up again.  Those examples are pure speculation and might not be the issue at all, but there is no doubt that inexperienced receivers can really slow down an offense.

How many FSU fans will jump to compare EJ's 2011 numbers to Ponder's 2010, saying EJ should've been playing all along?  I'm not saying EJ isn't as good as Ponder now, or that he won't be better than Ponder at some point.  I just hope fans recognize that EJ will be throwing to a completely different, more veteran receiving corp than Ponder did, even though it will consist of the exact same players.

Vikings fans might have a legitimate complaint for taking Ponder at 12th overall.  However, I believe that if Ponder had the opportunity that Dalton, Mallett, or Luck had (> 80% returning receiving yards), his production would've been significantly better, and he might not have made it past Tennessee or Jacksonville at 8th or 10th overall.  FYI for any Vikings fans reading this: FSU graduated two senior receivers in 2009, including our leading receiver.  To further add to the losses, our most talented underclassmen receiver, Jarmon Fortson (6'3" 230), was dismissed from the team in the summer for a failed drug test among other issues.  This left FSU with two slot-type receivers, Bert Reed (5'10" 165) and Taiwan Easterling (5'11" 190) as our only two returning receivers with any significant amount of receptions.  Outside of Reed and Easterling, our receivers had a combined 2 catches.

 Many FSU fans are looking to 2012 for the year when FSU will make a run for the National Championship.  I think that 2011 also presents a unique opportunity for FSU.  Five of the best six teams we face will be installing new offenses (UF, Clemson, Miami, Maryland, BC).  Our defense will continue to get better through 2012, but I suggest that the gap between FSU's defense and our opponents' offenses could be larger in 2011 than 2012, when very talented teams like UF and Clemson will be in the second year with their new system.  This obviously doesn't account for schedule, since we don't play OU in 2012 and play both Clemson and UF at home in 2012.  But, I could honestly see the neutral-site point spread being better in 2011 than 2012. 

Afterthought: Many fans were worried after EJ's performance in the spring game.  Let's not forget that while playing behind a patch-work O-Line in a game where he couldn't run, he was also only throwing to 22% returning receiving yards on the gold team (Bert Reed).  Things will look very different when he has Rodney, Willie, Bert, Taiwan, plus all the young guys to work with.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Tomahawk Nation

You must be a member of Tomahawk Nation to participate.

We have our own Community Guidelines at Tomahawk Nation. You should read them.

Join Tomahawk Nation

You must be a member of Tomahawk Nation to participate.

We have our own Community Guidelines at Tomahawk Nation. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker