Compu-Picks 2012 Analysis: Week 13_1

Mr Pac Ten
Posted Nov 26, 2012


Compu-Picks 2012 Analysis: rating the top and bottom teams in college football after week thirteen

As of the end of Saturday's games, these are the Compu-Picks top 35 and bottom 30 (plus a couple extras). Remember that this is a predictive model, designed to pick games and show how good a team actually is. Its results can be very different from what you'll see elsewhere. The workings of the model are confidential (it is, after all, designed to make winning picks), but I'm happy to answer questions about the models' results.

Rank BCS Rank Team League Score Schedule Rank * Result Rank
1 2 Alabama SEC 0.96 10 1
2 5 Oregon Pac-12 0.92 6 2
3 4 Florida SEC 0.77 3 12
4 9 Texas A&M SEC 0.76 4 10
5 6 Kansas State Big 12 0.69 17 7
6 1 Notre Dame Indep 0.69 32 5
7 3 Georgia SEC 0.63 36 4
8 10 South Carolina SEC 0.62 22 15
9 8 Stanford Pac-12 0.62 13 20
10 11 Oklahoma Big 12 0.61 8 22
11 15 Oregon State Pac-12 0.60 2 34
12 7 Louisiana State SEC 0.60 11 21
13 13 Florida State ACC 0.50 41 6
14 23 Oklahoma State Big 12 0.49 7 33
15 14 Clemson ACC 0.46 43 11
16 18 Texas Big 12 0.46 27 29
17 NR Southern California Pac-12 0.44 25 37
18 4* Ohio State Big Ten 0.39 57 13
19 NR Arizona State Pac-12 0.36 31 41
20 24 Utah State WAC 0.35 75 8
21 16 UCLA Pac-12 0.35 37 30
22 NR Fresno State Mountain West 0.34 67 14
23 NR Texas Christian Big 12 0.32 24 53
24 12 Nebraska Big Ten 0.32 40 28
25 NR Brigham Young Indep 0.31 49 25
26 25 San Jose State WAC 0.31 72 16
27 NR Mississippi State SEC 0.31 29 47
28 NR Arizona Pac-12 0.30 5 61
29 19 Michigan Big Ten 0.30 39 36
30 NR Wisconsin Big Ten 0.30 48 31
31 NR Texas Tech Big 12 0.30 19 55
32 NR Mississippi SEC 0.29 15 59
33 NR Baylor Big 12 0.28 23 56
34 NR Vanderbilt SEC 0.28 38 45
35 20 Boise State Mountain West 0.27 95 9
36 22 Northwestern Big Ten 0.24 53 35
43 21 Northern Illinois MAC 0.19 124 3
48 17 Kent MAC 0.14 113 17
95 . UTSA WAC -0.31 110 85
96 . Florida Atlantic Sun Belt -0.32 60 102
97 . Temple Big East -0.34 73 98
98 . Houston C-USA -0.34 117 77
99 . Marshall C-USA -0.34 109 79
100 . Florida International Sun Belt -0.36 100 89
101 . Colorado State Mountain West -0.38 90 94
102 . Boston College ACC -0.39 52 111
103 . New Mexico Mountain West -0.40 98 92
104 . Central Michigan MAC -0.41 120 80
105 . Memphis C-USA -0.44 118 86
106 . Western Michigan MAC -0.44 122 82
107 . Texas State WAC -0.44 83 99
108 . Buffalo MAC -0.45 92 100
109 . Texas-El Paso C-USA -0.46 96 101
110 . Nevada-Las Vegas Mountain West -0.49 91 105
111 . Army Indep -0.50 101 103
112 . Colorado Pac-12 -0.51 34 124
113 . Miami (Ohio) MAC -0.53 105 106
114 . Illinois Big Ten -0.53 55 119
115 . Alabama-Birmingham C-USA -0.58 108 108
116 . Eastern Michigan MAC -0.60 103 113
117 . South Alabama Sun Belt -0.60 97 112
118 . Tulane C-USA -0.64 106 114
119 . Hawaii Mountain West -0.65 79 120
120 . Idaho WAC -0.69 82 121
121 . Southern Mississippi C-USA -0.70 107 118
122 . Akron MAC -0.71 115 117
123 . Umass MAC -0.80 121 122
124 . New Mexico State WAC -0.88 114 123

This week's extra table will be an update of the league ratings posted a few weeks ago. I'm also posting a more in-depth look at league ratings, since while the table below does show a substantial amount of detail, there's still a lot more to the story than it can show in that amount of space.

I'm going to try and cycle through various potentially interesting outputs through the year; let me know if there's something in particular that you might like to see in future versions of this article.

League Rating OOC Schedule Rating Home/Away/Neutral Splits OOC vs Top 5 OOC vs 6-15 OOC vs 16-35 OOC vs 36-62 OOC vs 63-89 OOC vs 90-109 OOC vs 110-119 OOC vs Bottom 5
SEC 0.37 -0.08 28 / 10 / 3 0 - 0 2 - 1 2 - 1 8 - 5 8 - 1 6 - 0 3 - 0 4 - 0
Big 12 0.34 -0.10 13 / 8 / 0 0 - 0 0 - 1 1 - 1 3 - 1 5 - 1 8 - 0 0 - 0 0 - 0
Pac-12 0.25 0.11 15 / 11 / 1 0 - 0 1 - 3 6 - 4 4 - 1 2 - 1 1 - 1 3 - 0 0 - 0
Indep 0.10 -0.06 21 / 18 / 0 0 - 0 2 - 1 3 - 3 4 - 6 6 - 2 6 - 2 1 - 1 2 - 0
Big Ten 0.06 -0.13 27 / 12 / 1 0 - 1 0 - 4 3 - 2 5 - 2 3 - 3 7 - 2 5 - 0 3 - 0
Big East -0.01 -0.13 13 / 17 / 0 0 - 0 0 - 2 0 - 1 2 - 5 5 - 4 4 - 1 4 - 0 2 - 0
ACC -0.02 0.05 19 / 14 / 2 0 - 2 0 - 6 0 - 3 1 - 8 7 - 1 3 - 0 2 - 1 1 - 0
Sun Belt -0.09 0.05 8 / 22 / 0 0 - 5 0 - 3 0 - 4 1 - 3 2 - 3 3 - 1 3 - 0 2 - 0
WAC -0.15 -0.09 10 / 16 / 0 0 - 1 0 - 2 1 - 4 2 - 1 2 - 3 4 - 3 3 - 0 0 - 0
Mountain West -0.18 -0.04 12 / 19 / 1 0 - 1 0 - 0 1 - 10 0 - 4 1 - 5 2 - 0 4 - 1 3 - 0
MAC -0.26 -0.07 12 / 25 / 0 0 - 1 0 - 2 0 - 3 3 - 6 5 - 6 2 - 2 4 - 1 2 - 0
C-USA -0.33 0.06 21 / 21 / 0 0 - 1 0 - 3 1 - 11 0 - 12 1 - 7 3 - 1 0 - 0 2 - 0
League Rating Bowl Record OOC vs AQ's OOC vs Non-AQ's OOC vs Big 12 / SEC OOC vs Pac-12 OOC vs ACC / Big East / Big Ten OOC vs Sun Belt / MWC / WAC OOC vs CUSA / MAC
SEC 0.37 0 - 0 8 - 6 25 - 2 0 - 1 2 - 0 6 - 5 13 - 2 12 - 0
Big 12 0.34 0 - 0 5 - 2 12 - 2 1 - 0 0 - 1 4 - 0 7 - 0 5 - 2
Pac-12 0.25 0 - 0 6 - 5 11 - 5 1 - 2 0 - 0 5 - 1 6 - 4 3 - 0
Indep 0.10 0 - 0 14 - 6 10 - 9 1 - 0 3 - 2 10 - 4 8 - 5 2 - 4
Big Ten 0.06 0 - 0 6 - 9 20 - 5 1 - 2 1 - 3 4 - 1 5 - 1 14 - 3
Big East -0.01 0 - 0 7 - 9 10 - 4 3 - 0 0 - 1 4 - 7 2 - 0 6 - 4
ACC -0.02 0 - 0 6 - 17 8 - 4 1 - 8 0 - 1 5 - 5 3 - 2 4 - 0
Sun Belt -0.09 0 - 0 3 - 16 8 - 3 2 - 12 0 - 1 1 - 3 0 - 0 7 - 2
WAC -0.15 0 - 0 3 - 6 9 - 8 0 - 3 1 - 1 2 - 2 5 - 2 2 - 3
Mountain West -0.18 0 - 0 3 - 11 8 - 10 0 - 3 3 - 4 0 - 4 4 - 5 2 - 2
MAC -0.26 0 - 0 8 - 17 8 - 4 1 - 4 0 - 1 7 - 12 3 - 2 1 - 1
C-USA -0.33 0 - 0 1 - 23 6 - 12 1 - 12 0 - 2 0 - 9 5 - 10 1 - 1

Some thoughts on the list:

1) Please note that AA games are NOT counted for these ratings. This includes the schedule rankings. At some point later this year, I will post an adjusted schedule list that does account for the AA games, but they are not ready at this time. Please keep this in mind when looking at the schedule rankings, since a "true" schedule ranking would note these games.

2) One consistent theme that pops up when I've done these analyses the past few years is that Compu-Picks gives a lot more weight to schedule strength and dominance than does the BCS, and a lot less weight to simple W/L record and head to head. The same thing is true this time around.

3) As usual, I'm only posting the Compu-Picks ratings for the top and bottom teams (top 35 / bottom 30 this week), and will slowly expand the list as the season goes on. The reason I do this is that the teams at the very top and very bottom have largely separated themselves by now, while the teams on the next tier can largely be jumbled together.

4) Again this year I'm tracking the "Compu-Picks Curse" a bit more carefully. Below is the list of the teams that the system thought overrated each week (* means a bye/AA game or a game against someone else the model didn't like). So far teams have been exposed in nine of thirty-one potential games.

After week 7 (0-2): Florida, Notre Dame - no wins, but BYU came close at South Bend
After week 8 (3-2): Notre Dame, USC, Mississippi St, Georgia, and Ohio - Notre Dame had a very impressive win at Norman, Georgia beat a very good Florida team (though that game was UGLY), but USC lost to unranked Arizona, Ohio lost to 3-4 Miami(OH), and Miss St got annihilated at Bama.
After week 9 (0-6): Notre Dame, LSU, Georgia, Clemson, Ohio St, Louisville - Notre Dame came very close to losing to below-average Big East team, but for the purposes of this analysis, a win is a win. LSU lost... but if you think a close loss to the #1 team in the country is "getting exposed"... I'm sorry, but I just can't help you.
After week 10 (3-4): Georgia, Ohio St*, Nebraska, Louisiana Tech, Rutgers, Northwestern, Louisville, Toledo - 3 teams go down to unranked opponents (most notably Louisville getting smacked around by a now 5-5 Syracuse team), and Nebraksa and Louisiana Tech both came close to making it five.
After week 11 (2-3): Kansas St, Georgia*, Ohio St, Nebraska, Rutgers, Louisiana Tech, Louisville* - Louisiana Tech loses to unranked Utah St and Kansas St gets smoked at Baylor. Ohio St also came very close to losing at unranked Wisconsin, though Rutgers had a solid win at Cincy.
After week 12 (1-5): Notre Dame, Georgia, Ohio St, Nebraska, Louisville, Kent - Louisville gagged to a sub. 500 UConn team, but that was it for losses (though Nebraska, Notre Dame and Ohio St each got pushed pretty hard by unranked teams).

5) According to the BCS and all the human polls, Oregon should be ranked behind Georgia. This despite the Ducks playing a tougher schedule, being more dominant in their wins, having a MUCH closer loss (3 points in OT vs blowout loss) AND losing to a better opponent than Georgia (though it's fair to note Stanford was at home and South Carolina was away). This is utterly ridiculous.

6) Illinois still occupies the "worst AQ in the country" title, though it's truly neck and neck between them and Colorado.

7) Ohio State being ranked 4th by the AP poll is hilarious. They have everything wrong with Notre Dame's resume (way too many struggles against mediocre opponents) without any of the things that help bolster the Irish's resume, such as a big win at Oklahoma or a win (even a controversial home overtime win) over Stanford. Instead the Buckeyes played really well against Nebraska and othewise simply haven't managed to lose.
Legitimate top 5 teams obviously shouldn't struggle against mediocre to bad opponents, and yet Ohio St has done this more often than not, with single digit wins against Cal, Michigan St, Indiana, Purdue (in overtime), and now Wisconsin (though in fairness they're a bit better than mediocre). Not to mention far from convincing wins against UAB and Penn St. Honestly, this isn't even one of the ten best teams in the country, much less #4. "Lucky to be undefeated against a poor schedule" is not anything close to a top-tier resume.

8) The following teams are ranked materially higher by the model than the BCS: Alabama, Oregon, Texas A&M, Oregon St, Oklahoma St, USC.

In terms of power ratings, it's still easy to put Bama #1. They've been dominant in almost every game this year, and have done so against a very nasty schedule. Again, W/L doesn't rule over all, and their consistently high level of play is enough to elevate them here despite their loss.

After a dominating win at a very good Oregon St team, the Ducks have moved up from 5th to... 5th. While stickyness of rankings at this point of the year is generally reasonable, the Ducks have proven week in and week out they're one of the true elites of college football in 2012.
One of the great example of how Compu-Picks differs from the BCS is that Oregon played a very good Stanford team at home and lost a very close game, by just 3 points in overtime, and then followed that up with a dominating win in Corvallis. Should this fundamentally drop our estimate of the Ducks? Of course not. And yet in the past two weeks the Ducks slid from #2 in the BCS to #5, while in Compu-Picks their rating actually ROSE a bit (though not much). If we're truly looking for who the best teams are, isn't the Compu-Picks approach (giving credit rather than demerit for 1-1 against two very good opponents with the win a blowout and the loss an OT squeaker) more reasonable?

Texas A&M clearly deserved a big bump from winning at Bama, and that's what they got. Unfortunately, they started out underrated, and that's still the case after their big bump. Nasty schedule, great results, and just 9th? That's just off.

The Washington loss was bad for Oregon St, but this team has still achieved a tremendous amount. Key wins at Arizona, BYU (convincingly) and UCLA, a home win vs ASU, and a solid showing (in defeat) at Stanford... there's a whole lot more heft to the resume than they're getting credit for.

Oklahoma St has faced a really hard schedule and has done pretty well against it. Their poor BCS ranking is another good example of the system strongly devaluing schedule strength.

USC may be one the most blatantly underrated teams in the country. 7-5 sounds bad, but three of their losses were fairly close losses to very good Oregon, Notre Dame and Stanford teams, a fourth was a close game on the road to another top 25 team (UCLA), and a fifth was another close game on the road to a team not far off the top 25 (Arizona). They also won at Washington (Oregon St and Stanford couldn't do that), won at Syracuse (who may split the Big East title) and beat up a top 25 ASU team. They're basically being punished for playing a rough schedule, for having a lot of breaks go against them (0-3 in games decided by single digits strongly suggests bad luck) and for being a disapointment (after starting preseason #1). When it comes to team quality, they're on the top 25 teams in the country. I mean seriously, does ANYONE think that teams like Northwestern, Kent or Northern Illinois would even be able to compete against USC?

9) The following teams are ranked materially lower by the model than the BCS: Notre Dame, Georgia, Ohio St (AP Rank 4th), UCLA, Nebraska, Northwestern, Northern Illinois, Kent.

Compu-Picks was impressed by Notre Dame's win at USC, and the Irish have moved up accordingly, but they're simply not the best team in the country. They've won all their games, but a team that struggled against Purdue, Michigan, Stanford, BYU AND Pitt isn't truly a top 2 team. Their schedule is weak compared to the other top undefeated teams, they haven't been anywhere near as dominant, and they're pretty clearly lucky to be undefeated.

Georgia almost splits into two teams. "Good Georgia" beat up Mizzou, Vanderbilt, Ole Miss, Auburn, Georgia Tech and beat Florida. They're a legitimate top 5 team and national title contender. However, there's another team to consider. "Bad Georgia" struggled against Buffalo, barely held off Tennessee and Kentucky, and got blown out by South Carolina. "Bad Georgia" has no business being in the top 25, much less top 10, much less top 5.
And it's interesting that Georgia's ranking of 5th basically ignores "Bad Georgia." Four out of their eleven 1-A performances this year (Compu-Picks ignores AA games) showed a team that was a total fraud to be in the top ten, much less top five, and yet here they are, as if those bad games never happened. Very strange.

Ohio St was covered above. But it's still ridiculous to see them at the 4 spot in the AP poll.

Five spots in the lower end of the top 25 usually isn't much, but in this case there's still a really big ratings gap between where Compu-Picks has UCLA (21st, rating of 0.35) and the BCS 16 spot (Compu-Picks rates the #16 team as Texas, with a 0.46 rating). There's nothing particularly bad about their resume, but there just isn't enough heft to justify the 16 spot. They were great against Arizona and strong against USC, but there really isn't that much else. And that Cal loss was simply AWFUL. Take away that loss and 16th looks more reasonable, but Cal is a total mess and UCLA got annihilated by them.

Nebraska has two losses (one a blowout), barely held off Wisconsin, Michigan St, Northwestern and Iowa, and their shining moments were a blowout over Arkansas St and winning by 14 against Michigan. They're not a top 20 team much less top 15.

It's hard to see what exactly Northwestern is doing in the top 25. Their season highlights were wins against Vandy and Michigan St, which while not bad wins are still nothing close to what ought to be at the head of a top 25 team's resume. They also lost by 11 at Penn St (a pretty bad loss for a supposedly top 25 team), lost to Nebraska (though it was very close), lost to Michigan (though again it was very close) only beat Syracuse by 1, only beat Minnesota by 8, only beat BC by 9... honestly, most of their performances pretty clearly rate as mediocre. This simply isn't a top 25 team.

It's no mystery why Compu-Picks thinks NIU is wildly overrated; their schedule (at least for 1-A games) is rated 124th out of 124. And while they have dominated this schedule, dead last is still dead last. Plus they lost to Iowa.

Kent was blown out by Kentucky. That alone makes top 25 for these guys ridiculous even if they'd played an anotherwise solid schedule (they haven't) and/or they'd dominated said schedule (they haven't).

10) This isn't directly to do with the list, but here are some fun lists of results:

@ Stanford 21, USC 14
@ Washington 17, Stanford 13
USC 24, @ Washington 14

@ Miami 44, NC St 37
@ NC St 17, Florida St 16
Florida St 33, @ Miami 20

@ South Carolina 35, Georgia 7
@ Florida 44, South Carolina 11
Georgia 17, Florida 9

LSU 24, @ Texas A&M 19
Alabama 21, @ LSU 17
Texas A&M 29, @ Alabama 24

If you try to apply "head to head is the only thing that matters" logic to this list, your head will explode. You can tease out certain information from these lists (USC and Florida St had closer losses than wins, therefore they get a bonus; Washington and Miami had both games at home, therefore they get a demerit, etc.), but what it really does is highlight that each of these results was JUST ONE GAME. To properly evaluate a team, you need to evaluate the whole resume, not pretend that a single result means everything and the rest almost nothing just because of head to head "logic". That's why Compu-Picks doesn't give ANY special consideration to head to head results. You are what your resume says you are. Period.

Technical notes about the lists:

1) Conference ratings are straight averages of all of the teams in the league. There is no "central averaging" (like Sagarin does), or over-weighting the top teams, or anything like that. Such approaches would yield different numbers, and could potentially change the order of some of the leagues.

2) Games against AA teams are not counted. There are many good arguments both for and against counting such games (see this link for an interesting analysis of the issue). I have elected not to count these results in the Compu-Picks model. As is the case almost every year, this means that one or two especially surprising AA upsets don't make it into the numbers, skewing the results to a fair degree for a couple of teams. I believe that this is a more than acceptable tradeoff given the substantial issues that counting AA games would create, but you are certainly welcome to disagree with my decision on this matter.

3) As mentioned here, the purpose of this system is to make picks, not to create a list used for rankings. As such, I evaluate the system solely on the basis of how good a job it does making picks. I do not evaluate the system on the basis of whether or not it agreed with AP polls, BCS rankings, the BCS computers, or any other such list out there. In fact, the system has a long and established history of being substantially different than those sources. I am fine with these differences. To be honest, I publish these lists because I find them interesting and thought-provoking, and because I believe it is a good thing to introduce an approach that doesn't simply regurgitate the same avenues of thinking as you can find in most places.

4) The system is noisy, especially earlier in the year. This is why I start with only the top and bottom few, and slowly expand the list. While I believe that the numbers are reasonable, I certainly accept that they're not perfect. If you believe that a specific team is over- or under-ranked, you may well be right. I bring this up because if you're going to criticize the system for being wrong about a team, I'd appreciate it if you explain why you think the system is substantially wrong, rather than just marginally so (if it's just one or two slots off, especially well before the end of the year, I'd consider that well within a reasonable error range).

There are a few important notes and caveats I need to make about this model:

1) Compu-Picks does not endorse implicitly or explicitly any form of illegal gambling. Compu-Picks is intended to be used for entertainment purposes only.

2) No guarantee or warranty is offered or implied by Compu-Picks for any information provided and/or predictions made.

2012 Compu-Picks Blog

Questions, comments or suggestions? Email me at cfn_ms@hotmail.com

Follow cfn_ms on Twitter

Related Stories
Nebraska and UW know all about close games
 -by BigRedReport.com  Nov 26, 2012
Todd Monken: Chelf Is Our Quarterback
 -by GoPokes.com  Nov 26, 2012
Bill Young: Looking For A Defensive Rebound
 -by GoPokes.com  Nov 26, 2012