Writers of Pro Football Prospectus 2008

Most Recent FO Features

AllenKee13.jpg

» Week 11 Quick Reads

Most of the headlines covering this weekend's Bills-Chargers game focused on Nathan Peterman and his very bad day. Few realized that at the same time, Keenan Allen was having a career day for Los Angeles.

09 Oct 2015

Varsity Numbers Looks Into The Future

by Bill Connelly

Last week, I talked about my rather aggressive phasing out of preseason projection data and some of the outliers that it helped to create in the S&P+ rankings.

I sided with S&P+ on Oklahoma State (overrated by the polls), Northwestern (overrated), Tennessee (underrated), and Alabama (underrated). A week later, I feel pretty good about that decision on OSU and very good about Alabama, but Northwestern and Tennessee both played more to their AP rankings than S&P+ last week.

On the flip side, I sided with the AP poll on Utah (underrated by S&P+), Mississippi State (underrated), Baylor (underrated), West Virginia (overrated), Minnesota (overrated), and Duke (overrated). Duke is making me question myself, but the others played about as I would have guessed last week.

And then, with preseason projections completely phased out, Baylor dropped like a rock again. The Bears are being dinged by some extreme strength of schedule adjustments -- the adjustments themselves aren't intended to be incredibly strong, but the fact that three of their four opponents currently rank 118th (SMU), 125th (Rice), and N/A (Lamar) is pretty damn extreme. And the fact that the Bears have been an awful field-position team because of special teams and turnovers (legitimate concerns, both) against such a bad slate of teams has produced an extreme No. 39 ranking.

The rankings of Baylor and TCU (No. 36) both made me a little bit insecure. I have come to expect some interesting S&P+ outliers early in the season, and phasing out preseason projections so quickly was bound to create more. But ... 39th and 36th!

To either assuage my concerns or simply occupy myself while I wait to see what happens with the Bears and Frogs, I took a look at how this new method of S&P+ ratings had the landscape laid out after five weeks in previous seasons. To look at this, I'll take a gander at the year-end S&P+ top 10 and where each team would have ranked after five weeks.

2014

1. Ohio State (after five weeks: 24th)
2. Alabama (second)
3. Oregon (eighth)
4. Auburn (fourth)
5. Arkansas (10th)
6. Ole Miss (third)
7. Georgia (13th)
8. TCU (38th)
9. UCLA (22nd)
10. Georgia Tech (57th)

These ratings are designed to see into the future at least a little bit, but teams simply change sometimes. Georgia Tech was mediocre in September and awesome in November. Ohio State was good in September and unbelievable in November. Ole Miss was awesome, then collapsed. Et cetera. Regardless, four of the year-end top 10 were outside the top 20 after five weeks. There's still only so much we can learn early.

2013

1. Florida State (fourth)
2. Alabama (seventh)
3. Baylor (first)
4. Stanford (second)
5. Auburn (41st)
6. Missouri (27th)
7. Ohio State (11th)
8. Arizona State (18th)
9. Oklahoma State (16th)
10. South Carolina (12th)

Six of the year-end top 10 in 2013 were within shouting distance of the top, but there were still a couple of late surges to account for: Missouri went from good to awesome, and Auburn went from mediocre to phenomenal. And on the other side, Florida (third after five weeks) went from decent on offense and nearly untouchable on defense to being merely good on defense and miserable on offense. The Gators fell to 33rd.

2012

1. Alabama (first)
2. Texas A&M (15th)
3. Florida (sixth)
4. Oregon (third)
5. Notre Dame (11th)
6. Kansas State (13th)
7. Georgia (10th)
8. Oklahoma (21st)
9. Stanford (12th)
10. South Carolina (29th)

The Year of the SEC (which had the top three teams and five of the top 10) was pretty stable overall. Oklahoma and South Carolina both improved down the stretch, and Texas A&M went from good to holy crap good, while some early hope from Texas (second after five weeks) went down the tubes when a young, banged-up defense briefly collapsed.

2011

1. LSU (sixth)
2. Alabama (first)
3. Oklahoma State (ninth)
4. Boise State (fourth)
5. Oklahoma (second)
6. Oregon (13th)
7. Michigan (seventh)
8. Texas A&M (third)
9. Stanford (eighth)
10. Wisconsin (10th)

Wow, this qualifies as about the most stable year ever. Nine of the top 10 after five weeks were in the top 10 after the season. See? S&P+ can sometimes actually see the future as programmed!

(The odd team here, of course, is Texas A&M, which was so good at blowing winnable games that it put 2015 Tennessee to shame. The Aggies had already lost by a combined five points to OSU and Arkansas teams that would go on to finish a combined 23-3, and they would blow three more easy leads and finish 7-6 before changing coaches and quarterbacks and storming to 11 wins the next year.)

OK, one more.

2010

1. Boise State (ninth)
2. Auburn (19th)
3. Ohio State (second)
4. Alabama (first)
5. Oklahoma (15th)
6. Stanford (10th)
7. Notre Dame (32nd)
8. South Carolina (16th)
9. Arkansas (30th)
10. TCU (eighth)

This was an odd year. With Cam Newton behind center, eventual national champion Auburn was the quintessential "you become your record" team. The Tigers won close game after close game early on, against teams both good and iffy, then surged down the stretch. Meanwhile, both Notre Dame and Arkansas developed into something far more than what they showed in September. And at the same time, teams that looked surprisingly good early on (No. 3 Arizona, No. 4 Missouri, No. 5 Iowa, No. 6 Miami) regressed to different degrees. These four teams all finished between 14th and 24th, but none could keep up their early level.

So what have we learned? That your five-week rating is more likely to be telling than not, but that teams do still change. If Baylor's strange rating is simply a product of fluky field-position issues, and the Bears are destined to play like a top-10 team, then the odds are good that they'll eventually grade out as such. Or they finish 8-4 this year and I shout, "THE NUMBERS KNEW IT ALL ALONG." Either way, I'll pretend the numbers knew what they were talking about.

Honing in

S&P+ and every other system of ratings in the world, gets a read on different teams with different speeds. Sometimes, you are what you are right from the start; other times, you're a big jumbled mess of random results early on. I thought it would be interesting to take a look at which teams S&P+ has gotten a bead so far, and which ones have left the system clueless.

Team Proj. Wins Diff (Proj. and
actual wins)
Avg. Diff
(Proj. and actual
scoring margin)
Against
the
spread
S&P+ Rk Preseason
Rk
Alabama 4.13 -0.13 -2.3 40% 1 2
Oklahoma 2.88 1.12 6.6 25% 2 11
Michigan 3.77 0.23 5.9 80% 3 30
Clemson 3.18 0.82 2.3 100% 4 17
Florida 3.06 1.94 12.0 60% 5 37
USC 3.13 -0.13 11.9 50% 6 12
West Virginia 3.03 -0.03 6.8 33% 7 39
Florida State 3.59 0.41 -4.2 50% 8 20
LSU 3.07 0.93 -3.3 75% 9 9
Stanford 3.64 0.36 4.1 75% 10 18
Duke 3.44 0.56 7.7 50% 11 55
Notre Dame 3.29 0.71 6.0 80% 12 16
Louisville 3.02 -1.02 1.0 75% 13 34
Ole Miss 3.91 0.09 3.4 50% 14 3
Toledo 2.17 1.83 12.2 25% 15 75
Wisconsin 3.92 -0.92 -2.0 60% 16 28
Arkansas 3.41 -1.41 -6.5 60% 17 8
Boston College 3.17 -0.17 8.1 33% 18 52
Utah 2.25 1.75 17.2 25% 19 41
Penn State 4.14 -0.14 -10.0 75% 20 31
Navy 2.87 1.13 13.1 67% 21 60
Texas A&M 3.51 1.49 7.4 60% 22 21
Iowa 3.13 1.87 8.7 75% 23 59
Minnesota 3.00 0.00 -9.8 50% 24 45
Ohio State 4.56 0.44 -10.8 40% 25 1
Team Proj. Wins Diff (Proj. and
actual wins)
Avg. Diff
(Proj. and actual
scoring margin)
Against
the
spread
S&P+ Rk Preseason
Rk
Georgia 3.89 0.11 -1.6 25% 26 5
Texas Tech 2.47 0.53 8.3 25% 27 51
Western Kentucky 3.53 0.47 5.6 80% 28 61
Northwestern 2.74 2.26 15.0 50% 29 62
Tennessee 3.42 -1.42 0.8 50% 30 14
Illinois 3.22 0.78 5.2 25% 31 57
Louisiana Tech 3.33 -0.33 5.6 50% 32 54
BYU 2.18 0.82 2.8 60% 33 49
Michigan State 4.08 0.92 -7.6 80% 34 10
Georgia Tech 3.70 -1.70 3.7 50% 35 19
TCU 4.20 0.80 4.4 33% 36 15
UCLA 4.12 -0.12 -5.1 80% 37 6
Temple 2.53 1.47 7.8 25% 38 58
Baylor 3.58 0.42 9.5 33% 39 13
Air Force 1.91 0.09 12.6 67% 40 81
Nebraska 3.23 -1.23 -1.5 60% 41 32
Tulsa 1.83 0.17 2.1 50% 42 108
Vanderbilt 2.05 -0.05 6.4 50% 43 77
Miami-FL 3.11 -0.11 1.0 0% 44 26
Pittsburgh 2.64 0.36 -2.0 100% 45 35
Washington 2.23 -0.23 9.6 0% 46 64
Kansas State 3.14 -0.14 -0.1 67% 47 38
California 3.38 1.62 10.5 50% 48 50
Syracuse 2.25 0.75 10.6 67% 49 79
Mississippi State 3.11 -0.11 2.9 100% 50 22
Team Proj. Wins Diff (Proj. and
actual wins)
Avg. Diff
(Proj. and actual
scoring margin)
Against
the
spread
S&P+ Rk Preseason
Rk
Arizona State 2.99 0.01 -8.1 50% 51 23
NC State 3.91 0.09 10.8 50% 52 44
Iowa State 2.34 -0.34 1.0 67% 53 86
Oklahoma State 3.57 1.43 5.4 100% 54 47
Indiana 2.38 1.62 5.2 50% 55 74
Oregon 3.96 -0.96 -14.3 50% 56 4
Boise State 3.97 0.03 11.1 25% 57 25
Appalachian State 2.49 0.51 13.8 33% 58 97
Houston 1.91 2.09 23.8 0% 59 88
Middle Tennessee 2.41 -0.41 19.2 25% 60 84
Bowling Green 2.09 0.91 4.3 20% 61 99
Ohio 2.41 1.59 10.6 0% 62 92
Massachusetts 1.12 -0.12 -0.8 50% 63 117
Memphis 3.43 1.57 12.4 75% 64 68
Oregon State 1.75 0.25 0.3 67% 65 80
East Carolina 2.41 0.59 2.4 75% 66 78
Northern Illinois 2.68 -0.68 2.1 50% 67 82
North Carolina 3.23 0.77 9.1 33% 68 40
Auburn 3.20 -0.20 -10.6 75% 69 7
Virginia Tech 3.39 -1.39 -4.5 50% 70 27
Central Michigan 1.89 0.11 5.8 50% 71 91
Cincinnati 2.90 0.10 2.9 50% 72 48
South Carolina 3.19 -1.19 -12.8 40% 73 33
Texas 2.47 -1.47 -13.5 60% 74 29
Team Proj. Wins Diff (Proj. and
actual wins)
Avg. Diff
(Proj. and actual
scoring margin)
Against
the
spread
S&P+ Rk Preseason
Rk
Purdue 1.93 -0.93 1.6 75% 75 67
Wake Forest 2.41 -0.41 5.5 25% 76 90
Marshall 3.99 0.01 -4.1 38% 77 43
Florida Atlantic 1.91 -0.91 -7.4 50% 78 94
Akron 1.82 0.18 7.9 25% 79 112
South Florida 1.43 -0.43 10.4 100% 80 93
Ball State 2.37 -0.37 -4.4 50% 81 104
Virginia 1.66 -0.66 -12.6 100% 82 42
Maryland 2.49 -0.49 -7.0 75% 83 53
UNLV 1.42 0.58 17.4 100% 84 120
Arizona 3.07 -0.07 -1.8 67% 85 36
UTSA 1.28 -0.28 0.1 80% 86 127
Georgia Southern 3.35 0.65 3.0 50% 87 72
Utah State 1.90 0.10 -1.2 33% 88 65
Kentucky 2.55 1.45 1.7 50% 89 46
Arkansas State 2.43 -0.43 -1.7 25% 90 71
San Diego State 2.79 -0.79 -3.6 75% 91 76
Army 1.72 -0.72 10.0 33% 92 126
New Mexico State 1.42 -1.42 -8.0 50% 93 124
Eastern Michigan 1.17 -0.17 8.8 60% 94 128
Colorado State 3.50 -1.50 -3.9 38% 95 73
Washington State 2.32 -0.32 -1.8 33% 96 56
San Jose State 2.30 -0.30 3.2 75% 97 103
Texas State 1.95 -0.95 -12.6 0% 98 110
Connecticut 1.73 0.27 5.2 25% 99 115
Western Michigan 1.59 -0.59 -0.7 33% 100 63
Team Proj. Wins Diff (Proj. and
actual wins)
Avg. Diff
(Proj. and actual
scoring margin)
Against
the
spread
S&P+ Rk Preseason
Rk
Missouri 3.67 0.33 -4.6 0% 101 24
Kent State 2.49 -0.49 -6.2 63% 102 95
Troy 1.52 -0.52 1.2 33% 103 122
Colorado 3.02 -0.02 3.3 38% 104 69
Buffalo 1.74 0.26 15.1 25% 105 123
Old Dominion 2.42 -0.42 -14.6 0% 106 105
Southern Miss 2.36 0.64 13.3 50% 107 107
UL-Lafayette 1.96 -0.96 -5.8 67% 108 85
Hawaii 1.44 0.56 -1.9 50% 109 121
Rutgers 2.26 -0.26 5.6 50% 110 70
Tulane 1.87 0.13 -7.0 100% 111 89
Nevada 2.28 -0.28 -3.5 75% 112 83
Florida International 1.88 0.12 4.2 75% 113 106
South Alabama 1.40 1.60 -1.3 75% 114 109
New Mexico 2.76 0.24 6.9 75% 115 98
UL-Monroe 1.32 -0.32 2.2 33% 117 101
Miami-OH 1.19 -0.19 -2.9 100% 118 113
SMU 1.84 -0.84 -2.9 100% 119 119
Fresno State 1.61 -0.61 -7.8 50% 120 96
Idaho 1.39 -0.39 -6.3 50% 121 116
Georgia State 1.53 -0.53 -0.9 100% 122 111
Charlotte 1.95 0.05 -9.4 33% 123 118
Wyoming 2.04 -2.04 -8.3 75% 124 114
Kansas 1.06 -1.06 -5.2 50% 126 102
Rice 1.63 0.37 0.9 25% 127 87
Central Florida 2.31 -2.31 -10.2 67% 128 66
UTEP 1.71 0.29 -9.5 50% 130 100
North Texas 0.87 -0.87 -12.2 75% 131 125

Preseason projections account for some of the larger differences -- teams like Duke, Toledo, Utah, Navy, and Iowa have achieved far greater things than the preseason numbers suggested, and S&P+ has been playing catch-up. For other teams -- LSU, Clemson, Stanford, Notre Dame, Louisville, Pitt, Oklahoma State, etc. -- it's been pretty dead-on from the start.

Posted by: Bill Connelly on 09 Oct 2015

2 comments, Last at 12 Oct 2015, 12:37pm by Tomlin_Is_Infallible

Comments

1
by gofastjoey :: Sat, 10/10/2015 - 9:13pm

well after the first half it looks like you are correct about TCU.

2
by Tomlin_Is_Infallible :: Mon, 10/12/2015 - 12:37pm

and after the 2nd half?

--------------------------------------
The standard is the standard!