Varsity Numbers

For students majoring in breakdown of college football play-by-play

VN: The Fun Has Begun

by Bill Connelly

Heading out from the bowels of Faurot Field late Saturday night, I got a text message from a good friend of mine that simply said, "I f---ing love Les Miles." I immediately called to find out what the hell the Mad Hatter did this time. Miles' latest magic act resulted in another win for LSU -- his win percentage in games decided by a touchdown or less is approaching .750. It also resulted in great attempted story-telling from my friend (he kept remembering other insane details and having to start over) and this great passage from Bruce Feldman:

How do you anticipate what Miles -- or this other-worldly power that is now driving him in a stunning way -- is going to come up with? [Urban] Meyer must have been thinking, "Oh, crap, I'm stuck in a Les Miles Moment. This is not good. This can't be good."

Pitted against Miles in moments like this must take on a surreal tone at this point. It's like realizing The Mush from "A Bronx Tale" is on your sidelines as the clock ticks down.

Between the No. 1 team losing in Week 6 and Miles seeming to benefit from any chaos surrounding him, it's beginning to look more and more like 2007. Fans of such teams as Illinois and Missouri (both of whom had memorable 2007 campaigns, to say the least) likely welcome such comparisons; fans of Alabama are probably less pleased. Regardless, it's time to buckle up, because the top dog has been taken down a few notches, the No. 1 team in next week's BCS standings might very well be from the WAC, and we've got what is likely another mess of upsets on the horizon.

Box Score of the Week

Alabama lost its first regular season game since Thanksgiving of (you guessed it) 2007. We basically have to give them the Box Score treatment, don't we?

No. 19 South Carolina 35, No. 1 Alabama 21

As Steve Spurrier said in his giddy post-game interview with Chris Fowler and the Gameday crew after Saturday's big win, it's one thing to knock off the No. 1 team; it's another to straight up beat them. People had almost forgotten that it was possible for Alabama to lose at this point, and the Gamecocks beat them by two touchdowns. Great win for Spurrier and the South Carolina program.


S. Caro.

Close % 100.0%
Field Position % 56.3% 35.1%
Leverage % 71.9% 75.4%
EqPts 20.6 25.1
Close Success Rate 45.3% 49.1%
Close PPP 0.32 0.44
Close S&P 0.774 0.931
EqPts 3.8 10.6
Close Success Rate 45.5% 44.4%
Close PPP 0.17 0.29
Close S&P 0.626 0.738
Line Yards/carry 2.86 1.79
EqPts 16.8 14.5
Close Success Rate 45.2% 57.1%
Close PPP 0.40 0.69
Close S&P 0.853 1.262
SD/PD Sack Rate 15.4% / 18.8% 10.0% / 0.0%
Success Rate 52.2% 48.8%
PPP 0.24 0.46
S&P 0.761 0.946
Success Rate 27.8% 50.0%
PPP 0.53 0.39
S&P 0.809 0.888
Number 1 1
Turnover Pts 4.7 4.2
Turnover Pts Margin -0.5 +0.5
Q1 S&P 0.748 1.396
Q2 S&P 0.773 0.726
Q3 S&P 0.632 0.777
Q4 S&P 0.910 0.805
1st Down S&P 0.754 1.029
2nd Down S&P 0.857 0.843
3rd Down S&P 0.681 0.766
Projected Pt. Margin +5
Actual Pt. Margin +14

So what is the recipe for defeating Alabama? Play really well. Explode in the first quarter. Pester Greg McElroy all game. Have a ridiculous receiver like Alshon Jeffery on your side. Easy, right?

Truth Nos. 1 and 2 ("The old adage of 'You can't win a game in the first quarter/half, but you can lose it,' rings true" and "Big plays win games") played a gigantic factor in South Carolina's win. (Truths No. 3 and 4, not so much.) After the Tide opened the game with a field goal, South Carolina unleashed drives of 63, 62 and 43 yards to go up 21-3. Despite a few iffy moments (Stephen Garcia tossed a badly snapped ball out of the back of the end zone for a safety even though he probably still had time to make a play), the Gamecocks' lead was never less than seven, and with the game on the line, it was the Tide making the mistakes.

The early lead significantly affected both teams' play-calling. Heading into Saturday's game, South Carolina ran 65 percent of the time on standard downs (32nd in the country) and 31 percent of the time on passing downs (79th). They have focused on the run as much as possible (and with Marcus Lattimore, why wouldn't you?), but in passing situations, Spurrier has, for better or worse, attempted to trust quarterback Stephen Garcia to make a play. Against Alabama, the Gamecocks ran more frequently than normal on standard downs (77%), likely because wanted to work the clock whenever possible. But at the same time, they ran just 21 percent of the time over the course of 14 passing downs. Instead of playing it safe with the lead, Spurrier took chances. Four of Alshon Jeffery's seven catches came on passing downs, and three gained first downs on second- and third-and-long situations. Despite the love-hate relationship Spurrier has had with the enigmatic Garcia, he put his faith in his quarterback to make plays. Garcia did that, with a huge assist from Jeffery.

Meanwhile, the early deficit significantly altered Alabama's gameplan. This team typically runs 63 percent of the time on standard downs and 37 percent on passing downs; they will run whenever they can get away with it. Instead, they ran just 43 percent of the time on standard downs and 11 percent on passing downs. Playing catch-up hinders even offenses which are typically good at passing and making big plays. The extra passing allowed South Carolina to come at McElroy all night long, and even when Alabama did run, they didn't get very far. South Carolina allowed a reasonably high level of rushing efficiency (45 percent), and they allowed Alabama to run a high rate of plays on the Gamecocks' side of the 50, but they allowed few big plays, particularly on the ground. In the end they made Alabama run enough plays that eventually the Tide made uncharacteristic mistakes.

As much as anything else, this is evidenced by each team's S&P in the red zone. South Carolina 1.784, Alabama 0.542. They punched Alabama in the mouth early and kept punching all game long. As we will see below, their performance earned them a Top Five S&P+ ranking.

10 Notable Games

No. 3 Oregon 43, Washington State 23

EqPts: Oregon 30.9, Washington State 19.2
T/O Pts: Washington State +2.4
Oregon > Washington State +9.3

Washington State is poison to a team's S&P+ ranking. If you mess around for even a little bit, your ranking is going to plummet. Wazzu's Nico Grasu booted a 50-yard field goal to cut Oregon's lead to 22-17 late in the first half, and I began to wonder aloud about a Varsity Numbers anti-jinx. Yes, the Ducks eventually pulled away, but it took them a while (how much of it had to do with a funk caused by Kenjon Barner's frightening injury, we don't know), and it cost their already low S&P+ ranking to fall double digits.

No. 7 Nebraska 48, Kansas State 13

EqPts: Nebraska 39.9, Kansas State 15.9
T/O Pts: Nebraska +3.2
Nebraska > Kansas State +27.2

Taylor Martinez's play-fake defies physics. I am still trying to figure out how he is able to keep the ball in the running back's belly so long before pulling it back and running untouched into the secondary. It's the damnedest thing. Of course, sometimes physics fights back -- Nebraska has fumbled 18 times in five games, and that will bite them at a very inopportune time if it continues. It's high risk, high reward for Nebraska and the zone read.

No. 8 Auburn 37, Kentucky 34

EqPts: Auburn 32.7, Kentucky 23.8
T/O Pts: Auburn +0.6
Auburn > Kentucky +9.5

Auburn should have had this one in hand, but the Tigers allowed Kentucky to stick around for quite a while. But after the Wildcats tied the game with 7:31 left, the Tigers pulled off potentially the drive of the weekend. It took up 18 plays, 86 yards, and every remaining second in the game. Wes Byrum's field goal gave the Cardiac Tigers another tight win. They need to start sending out anxiety medication as part of the Auburn Alumni Association benefits.

No. 12 LSU 33, No. 14 Florida 29

EqPts: LSU 24.9, Florida 15.0
T/O Pts: LSU +1.9
LSU > Florida +11.8

A special teams touchdown allowed the seemingly outmanned Gators to stay close in this one, but the Les Miles Magic took over late. The S&P+ rankings are slow to warm on the undefeated Bayou Bengals (they currently rank behind four two-loss teams -- Miami, California, Illinois and Oregon State), but the numbers will come around if they keep winning ... no matter how they keep winning.

No. 23 Florida State 45, No. 13 Miami 17

EqPts: Florida State 32.8, Miami 18.5
T/O Pts: Florida State +7.8
Florida State > Miami +21.1

Perhaps the biggest surprise of the weekend is that the surging Seminoles didn't rise more than 12 spots in the S&P+ rankings, especially considering where Miami was ranked. Hopefully the new batch of FEI rankings will give them their due, as they have been playing ridiculously good defense ever since they got home from Norman after Week 2.

No. 17 Michigan State 34, No. 18 Michigan 17

EqPts: Little Brother 34.3, Big Brother 18.8
T/O Pts: Little Brother +13.6
Michigan State > Michigan +29.1

So much for the shootout we all expected. Only one offense brought its "A" game in Ann Arbor, and it wasn't the one led by the guy with the untied laces. Michigan State is now 6-0 for the first time since 1999, and the Spartans face the most surprisingly big game of the weekend -- a showdown against S&P+ darlings Illinois.

No. 16 Stanford 37, USC 35

EqPts: USC 33.5, Stanford 29.3
T/O Pts: USC +9.4
USC > Stanford +13.6

The award for Most Undeserving Win of the Week goes to Jim Harbaugh's Cardinal, who used some last-minute magic to get it done against USC even though the numbers didn't suggest that was much of a possibility. The performance actually bumped USC's ranking up two spots, but that doesn't mean much when accompanied by an "L."

Illinois 33, Penn State 13

EqPts: Illinois 24.2, Penn State 12.3
T/O Pts: Illinois +7.2
Illinois + 19.2

I'll have more on Illinois this Friday. The way Virginia flamed out the moment they tasted solid S&P+ rankings, I very much expected the same of Illinois. Instead, the surprising Illini (who have losses to only S&P+ No. 1 Ohio State and No. 11 Missouri, both by respectable margins) showed what commitment to your offensive identity can do. Illinois is playing reasonably mistake-free ball right now, and they are putting quarterback Nathan Scheelhaase in position to succeed. It is impressive, if surprising.

(Meanwhile, Penn State may miss a bowl for the first time since 2004.)

Notre Dame 23, Pittsburgh 17

EqPts: Notre Dame 18.8, Pittsburgh 18.8
T/O Pts: Notre Dame +9.3
Notre Dame > Pittsburgh +9.3

With Notre Dame getting back to .500, Pittsburgh now assumes the "Best Team With Losing Record" throne. Turnovers did in the Panthers, as the two offenses produced virtually identical output.

Georgia 41, Tennessee 14

EqPts: Georgia 30.8, Tennessee 16.3
T/O Pts: Georgia +12.1
Georgia > Tennessee +26.6

This one was predictable. Georgia was angry and embarrassed following its loss at Colorado; Tennessee was just down after getting Miles'd in Baton Rouge. Early turnovers went Georgia's way, and they coasted.


Best Offensive S&P of the Week

1. Cincinnati (vs. Miami-OH): 2.191
2. TCU (vs. Wyoming): 1.828
3. Louisville (vs. Memphis): 1.827
4. West Virginia (vs. UNLV): 1.678
5. Boise State (vs. Toledo): 1.473

Cincinnati led 45-3 at halftime. By the way, the Big East is an amazing mess right now. Only two teams have lost fewer than twice, and one of those two is Syracuse. Meanwhile, Pittsburgh (at 2-3) is still talented, and Cincinnati (also at 2-3) has now played its best two games of the season back-to-back. Unless West Virginia runs away with the title, there is a solid chance that the Big East champion is 7-5 this season.

Worst Offensive S&P of the Week

1. Ball State (vs. Western Michigan): -0.187
2. Indiana (vs. Ohio State): 0.088
3. UAB (vs. Central Florida): 0.242
4. UNLV (vs. West Virginia): 0.329
5. UCLA (vs. California): 0.348

Bonus: 6. South Florida (vs. Syracuse): 0.373

Syracuse has the No. 98 defense in the country according to S&P+, and they shut South Florida down to an embarrassing degree.

Best Rushing S&P of the Week

1. Cincinnati (vs. Miami-OH): 1.892
2. West Virginia (vs. UNLV): 1.849
3. Boise State (vs. Toledo): 1.498
4. Louisville (vs. Memphis): 1.431
5. Nebraska (vs. Kansas State): 1.348

Four teams ran the ball better than Nebraska this week? Probably not, but realize that S&P only takes into account plays that took place when the game was deemed "close." Nebraska put the Kansas State game out of reach fast enough that a lot of its big runs came with a scoring margin far from close. Poor Bill Snyder's team simply has no speed.

Worst Rushing S&P of the Week

1. Indiana (vs. Ohio State): -0.007
2. Ball State (vs. Western Michigan): -0.004
3. Tennessee (vs. Georgia): 0.206
4. UCLA (vs. California): 0.236
5. UAB (vs. Central Florida): 0.267

In consecutive games, UCLA has killed Houston, killed Texas, flirted with doom against Washington State, and gotten mauled by California. You figure them out, because I give up. UCLA is a precautionary tale for anybody thinking about jumping on the Illinois bandwagon. Two weeks does not equate to a full season of strong play.

(Wait, nobody is actually considering jumping on the Illini bandwagon? Only me? Forget I said anything.)

Best Passing S&P of the Week

1. Air Force (vs. Colorado State): 2.934
2. Cincinnati (vs. Miami-OH): 2.789
3. TCU (vs. Wyoming): 2.554
4. Louisville (vs. Memphis): 2.223
5. Utah (vs. Iowa State): 1.544

It warrants mentioning that the two top teams on this list threw four and five passes, respectively, when the game was considered "close." Meanwhile, it also warrants mentioning just how much of a whooping Utah put on Iowa State on Saturday. I was curious how they would stand up to an odd trip and a salty Cyclones team. They passed the test.

Worst Passing S&P of the Week

1. Ball State (vs. Western Michigan): -0.370
2. Indiana (vs. Ohio State): 0.135
3. Akron (vs. Kent State): 0.224
4. UAB (vs. Central Florida): 0.226
5. UNLV (vs. West Virginia): 0.228

Bonus: 6. Purdue (vs. Northwestern): 0.247

Purdue was 2-2 and fresh off of a loss to Toledo and uninspiring wins over Ball State and Western Illinois. They took on undefeated Northwestern and won, getting outgained by 110 yards, splitting the turnover battle, and going 6-for-18 passing for 47 yards. Yes, they rushed for 232 yards (they got 132 from Rob Henry), but ... some games make no sense.


Full rankings here:

S&P+ Top 25 (After Six Weeks)
Rk Team Record S&P+ Last
Diff Off.
Rk Def.
1 Ohio State 6-0 311.6 2 +1 126.5 14 145.4 3
2 Alabama 5-1 275.3 1 -1 171.5 1 117.1 18
3 Miami-FL 3-2 272.3 4 +1 126.3 12 147.4 2
4 South Carolina 4-1 266.4 11 +7 129.0 8 118.0 11
5 Nebraska 5-0 266.2 12 +7 103.1 49 118.1 13
6 Boise State 5-0 264.6 7 +1 119.5 27 123.8 7
7 Arizona 4-1 262.9 3 -4 124.8 17 151.0 1
8 California 3-2 257.8 13 +5 131.3 6 117.7 9
9 Auburn 6-0 255.6 10 +1 134.4 3 108.9 33
10 Illinois 3-2 254.5 16 +6 125.9 16 123.5 6
S&P+ Top 25 (After Six Weeks)
Rk Team Record S&P+ Last
Diff Off.
Rk Def.
11 Missouri 5-0 254.3 8 -3 129.3 5 114.8 17
12 TCU 6-0 252.6 9 -3 115.8 32 118.1 12
13 Oregon State 3-2 251.7 28 +15 128.4 13 113.5 20
14 LSU 6-0 250.8 19 +5 112.4 37 112.0 25
15 Virginia Tech 4-2 248.9 17 +2 114.4 34 111.8 21
16 Iowa 4-1 246.6 6 -10 131.5 4 112.5 24
17 Florida State 5-1 246.1 29 +12 117.9 24 97.0 60
18 Michigan State 6-0 244.9 23 +5 120.1 21 101.4 49
19 Stanford 5-1 244.5 14 -5 136.1 2 108.5 34
20 Oklahoma 5-0 243.5 20 +0 117.2 29 107.4 36
S&P+ Top 25 (After Six Weeks)
Rk Team Record S&P+ Last
Diff Off.
Rk Def.
21 Florida 4-2 239.2 5 -16 119.8 22 134.3 4
22 Notre Dame 3-3 237.6 25 +3 104.1 48 129.0 5
23 Arkansas 4-1 236.1 24 +1 107.4 40 115.6 15
24 Utah 5-0 234.0 32 +8 106.3 45 101.1 52
25 USC 4-2 230.5 27 +2 111.4 35 114.1 19

Biggest S&P+ Movers of the Week


Syracuse (18 spots, from 88th to 70th). The Orange are still a few steps away from respectability, but they're also only two wins away from bowl eligibility after their surprising 13-9 win at South Florida.

Colorado (17 spots, from 58th to 41st). Colorado gets shut out, 26-0, and moves up? How? Two reasons: 1) Their strength of schedule got a strong bump from California whipping UCLA, Hawaii thumping Fresno State (in Fresno) and Georgia killing Tennessee; and 2) Penalties don't play a huge role in the S&P+ rankings (season-wide, they do not have any correlation to win percentage), and penalties murdered them in Columbia on Saturday. They committed several unforced penalties (delays of game, illegal formations, false starts) in Missouri territory, but the play-by-play stats don't see that.

Oregon State (15 spots, from 28th to 13th). They racked up a great win against a tough, S&P+ friendly Arizona team on Saturday, though it came with a cost: receiver James Rodgers was lost for the season with a knee injury after an awkward tackle.

West Virginia (13 spots, from 39th to 26th). Further proof that you can move up in the rankings despite iffy competition if you completely dominate.

Other Rises: N.C. State (51st to 38th), Rutgers (85th to 72nd), Florida State (29th to 17th), Indiana (97th to 85th), UL-Lafayette (113th to 101st), East Carolina (61st to 50th), Arkansas State (115th to 104th).


Penn State (20 spots, from 34th to 54th). A true freshman quarterback, a shaky offensive line, and a multitude of defensive injuries are making this a forgettable season in Happy Valley.

South Florida (19 spots, from 54th to 73rd). B.J. Daniels is such a frustrating quarterback. Twice a game, he'll make a play that knocks your socks off; it will distract you from the other 60-plus plays that are anything but impressive.

Virginia (18 spots, from 35th to 53rd). September's S&P+ darlings are not responding well to their tiny taste of success ... nor where they supposed to.

Florida (16 spots, from fifth to 21st). It's interesting watching how the rankings change following certain games. Sometimes, neither team's ranking changes at all. Sometimes the winner rises dramatically while the loser remains stable. Sometimes it's the direct opposite. LSU rose five spots after beating Florida in The Swamp, but the S&P+ ratings suggest Florida's poor play was more of a cause of defeat than LSU's quality play.

Other Tumbles: Middle Tennessee (78th to 97th), Utah State (77th to 96th), Miami (Ohio) (93rd to 111th), Tennessee (55th to 71st), Oregon (15th to 30th), Houston (40th to 52nd), UCLA (21st to 33rd).

Favorite Moment of the Weekend

Not to turn Outsiders into "All Miles, All the Time" territory, but ... my favorite moment of the weekend just has to be Les Miles' response to Holly Rowe's question about his "Mad Hatter" nickname. His "I understand it's the hat I wear, but there's nothing mad underneath it" response was equal parts embarrassed, proud, happy, and angry. It summed up everything that is both likable and befuddling about the man who is currently the most popular coach in America, whether LSU fans like it or not.


27 comments, Last at 15 Oct 2010, 12:16pm

1 Re: VN: The Fun Has Begun

by CuseFanInSoCal // Oct 12, 2010 - 12:47pm

Although chaos may allow the Orange into a bowl at 6-6 with two FCS wins, normal bowl eligibility rules say only one counts, so we need three more wins to get to a bowl. But with most of the Big East and BC pretty bad (and coming to the Dome) this year, that should be doable.

Login or Register to post comments

2 Re: VN: The Fun Has Begun

by Joseph // Oct 12, 2010 - 12:57pm

Of course, if Miles could recruit a decent QB (not his ranking or something like that, but his playing ability), most of these last-second-crazy-wild endings wouldn't be necessary.

Login or Register to post comments

4 Re: VN: The Fun Has Begun

by lionsbob // Oct 12, 2010 - 1:21pm

Of course, those must be so easy to find.

Part of the problem was of course the heir apparent (and SEC championship MVP) Ryan Perriloux decided to keep himself constantly in the spotlight...for the wrong reasons.

That left LSU starting either a walk-on from Harvard (Andrew Hatch) or a freshman QB (Jarrett Lee).

They had Oregon's current QB, Darron Thomas, but he obviously changed his mind and went to Oregon.

They went after a QB this past year who was also a million dollar worthy baseball player (Zac Lee).

I think the only QB they have in this class is Phillip River's little brother.

Login or Register to post comments

5 Re: VN: The Fun Has Begun

by Rocco // Oct 12, 2010 - 2:28pm

I don't have the link in front of me, but there was a story about how Darron Thomas had verbally committed to LSU and wanted to play QB there, but LSU kept recruiting QBs and Thomas saw he was being listed on Rivals as an athlete. Before a game where Thomas and other recruits were in attendance, an assistant coach told Miles to call Thomas a QB and not an athlete. When Miles met Thomas at the game, he told Thomas something to the effect of "you're a great athlete and we want to get you on the field", at which point Thomas decided he wanted to go elsewhere. Whoops.

Login or Register to post comments

3 Re: VN: The Fun Has Begun

by mm // Oct 12, 2010 - 12:59pm

"Best Passing S&P of the Week
1. Air Force (vs. Colorado): 2.934"

Air Force played Colorado State.

The Pac 10 has dropped a bit from the top, but still has a good overall showing.

The Big 12 South continues to look unimpressive. #20 OU, #31 OSU, #40 A&M, #46 UT, #55 Baylor, #57 TT.

Login or Register to post comments

6 Re: VN: The Fun Has Begun

by Will // Oct 12, 2010 - 3:14pm

Heard this on the ESPNU podcast.

What happens in the (very unlikely) event that both Ohio State and Michigan State go undefeated (they don't play each other)? Could they face off in the BCS title game? Before the season, everyone said that Boise or TCU wouldn't make it over a 1-loss SEC champ or an undefeated Big Ten/Big 12/Pac-10 champ, but what if two teams from the Big Ten were to do it? Bizarro!


Login or Register to post comments

7 Re: VN: The Fun Has Begun

by Bill Connelly // Oct 12, 2010 - 3:44pm

Hadn't thought about that. If 2006 is any indication, poll voters will adjust their votes accordingly to make sure that doesn't happen.

Login or Register to post comments

11 Re: VN: The Fun Has Begun

by zlionsfan // Oct 12, 2010 - 7:34pm

2006 shouldn't be a factor. The issue there was that Michigan and Ohio State had already played; the issue this time would be that Michigan State and Ohio State don't play.

I don't disagree with your point, though. The next-highest AQ school would magically jump MSU (and any other non-AQ school) to create an "acceptable" matchup, because as we all know, the top two teams in the country couldn't possibly come from the same conference. (I'm not saying they are - MSU has benefited from a really light schedule so far, including Michigan - just making a general case.)

In fact, if somehow the voters disregarded the comments I'm sure BCS people would be making and caused a matchup between two conference rivals, the next year we'd see a change to the system that would prevent it from happening again.

Login or Register to post comments

12 Re: VN: The Fun Has Begun

by Will // Oct 13, 2010 - 12:17am

Well, next year, the Big Ten gets a championship game, so it couldn't happen. Not sure if the Big East plays a full round robin or not, but if it does, then this is the last year this scenario can play out (Big 12 will play full round robin).

Suddenly, I'm rooting for Michigan State! They should be favorite in every remaining game except @Iowa, although Illinois will be a big test this week. This could be fun :)


Login or Register to post comments

16 Re: VN: The Fun Has Begun

by Flounder // Oct 13, 2010 - 10:46am

I'll be in the stands cheering for MSU, but I'm still far from convinced, and expecting a final record of 9-3. And that will be a very, very solid season for my Spartans.

Login or Register to post comments

8 Miami . . .

by Matty D // Oct 12, 2010 - 4:27pm

Miami moving up after losing to FSU is a real head-scratcher. They looked pretty bad doing it, too.

Login or Register to post comments

9 Re: Miami . . .

by Bill Connelly // Oct 12, 2010 - 4:41pm

Yeah, all I could come up with was that they were so far ahead of the rest of the field last week (the top four teams were much, much higher than everybody else) that while their rating may have dropped, their ranking did not. I assume it'll work itself out in the end, but that was a bit odd.

Login or Register to post comments

10 Re: VN: The Fun Has Begun

by Kal // Oct 12, 2010 - 7:27pm

Poor Oregon. I'm not that worried that their S&P is so far down compared to teams they beat, but it's sad to see them at #2 in the polls but #30 here. Time will tell, and playing Cal, Arizona and USC should let us figure out their real mettle. (UCLA will of course be completely useless as a gauge because it's, well, UCLA)

Login or Register to post comments

13 Oregon at #30

by cfn_ms // Oct 13, 2010 - 12:39am

just looks like a model bust to me. Something looks seriously off to me. If I had to guess, I think it's that the model is biased towards punishing teams way too much for playing one or two totally hideous teams (and UO has played 2 - UNM and a AA team, plus a pretty to very bad Wazzu team), which is probably also the source of it underrating Boise the last couple years.

Login or Register to post comments

14 Re: Oregon at #30

by Will // Oct 13, 2010 - 2:29am

I think the struggles in Pac-10 play hurts them - S&P gives you credit when you hammer bad teams. They struggled for long portions of #47 Arizona State and #99 Washington State, which I imagine severely hurts the rankings.

They still have three top 20 teams left in Pac 10 play (Arizona, Cal, OSU) - plenty of time to make up a lot of ground before the season ends.


Login or Register to post comments

15 Re: Oregon at #30

by Kal // Oct 13, 2010 - 10:08am

What I don't get is why those are 'struggles' though. Arizona State played Wisconsin close, OSU close and beat Washington. ASU is clearly not a bad team; they have turnover issues, but otherwise is good. WSU isn't good at all, but that's still a huge penalty.

Put it another way: Oregon lost significantly more ranking for playing WSU than they gained for playing Stanford. That seems flawed, especially given that they beat Stanford so heavily.

But as you say - it'll work itself out in the end.

Login or Register to post comments

17 Re: Oregon at #30

by cfn_ms // Oct 13, 2010 - 12:25pm

Actually, it won't "work itself out in the end." If the model and its assumptions has created a substantial bias against Oregon (which I think is fairly self-evident at this point), then that bias is going to remain. It will get weighted less and less as there are more games played that factor into the resume, but it's still there.

And a bias significant enough to rank #30 a team whose resume is no-question top 10 and by most estimates top 5 is clearly material. Of course, this isn't a BCS computer so it doesn't make a huge difference what its opinion is, but it's the sort of result that goes so far against the grain that it pretty much demands either a response of "yes, the model is broken, and I'm going to investigate why for next year's version" or a meaningful defense of it (and a "well they struggled a bit too much against Wazzu, sucks for them" type blurb doesn't cut it). Unless the model's error range is REALLY large (which I doubt), then it's not a result where it can be brushed off as being within a reasonable error range.

There should have been 2 or 3 paragraphs in this article dealing with this issue. Unless the FO perspective is "well, we do well for most teams' ratings, and we're cool with having a few numbers that look totally wrong", it just can't be brushed aside.

If I were Bill and/or Brian, one thing I'd do is every week, take the AP top 25 (or the BCS top 25 when it comes out), compare it to the model list, find the biggest one or two differences, and talk about them (it'd also be fine to just do this for the combined F/+ instead of the two component parts).

Coming up with a meaningful explanation that convinces people the model is correct is something which adds real value to the work done (plus they can probably cut and paste some of the work from week to week, since chances are some of the differences will remain for a number of weeks at a time). Although, to be fair, I'm biased on this issue, since that's one of the things I do myself, so maybe it's not fair to demand that others do them same (though I do think it's a really good idea).

Login or Register to post comments

18 Re: Oregon at #30

by Bill Connelly // Oct 13, 2010 - 4:43pm

"If I were Bill and/or Brian, one thing I'd do is every week, take the AP top 25 (or the BCS top 25 when it comes out), compare it to the model list, find the biggest one or two differences, and talk about them (it'd also be fine to just do this for the combined F/+ instead of the two component parts)."

I've been doing exactly this each Friday for the last two weeks.

I think we know who Friday's "What the..." team of the week is going to be, huh?

And while I would personally vote for Oregon in the Top Five right now ... or even No. 1 ... saying that Oregon is "a team whose resume is no-question top 10 and by most estimates top 5," and any model that doesn't agree is complete b.s. (which is, for all intents and purposes, exactly what you just said), is ridiculous. We're six weeks into the season. Here's who Oregon has beaten:

New Mexico ... who is most likely the worst team in FBS.
Tennessee ... who just got drubbed almost as badly (48-13) by a 2-4 team.
Portland State ... who is Portland State.
Arizona State ... who is currently 1-3 versus FBS competition.
Stanford ... who just tried to lose at home to USC (who is borderline Top 25).
Washington State ... who is most likely the worst BCS conference team.

Do I think they're the 30th-best team in the country? No. I think they're much, much better than that (Fremeau's ratings agree), and I think it will, indeed, work itself out in the end. In fact, their remaining opponents currently rank 33rd, 25th, 48th, 8th, 7th and 13th in S&P+. Their strength of schedule is about to improve exponentially, and their rating will follow ... as long as they keep winning, of course.

In the meantime, to suggest that Oregon's low ranking, against the schedule they've played so far (when the wins over ASU and WSU were less than impressive), proves something damning against the system itself, is insulting and premature. If they win out, they'll be at or near the top of the rankings. For now, they're only evaluated for who they have played ... and they've barely played anybody. If they're undefeated in early-December, and they're still No. 25 or something, we'll talk. Actually, we won't talk, because I'll be too busy fixing my system. But how about we see how this plays out first?

Login or Register to post comments

19 Re: Oregon at #30

by cfn_ms // Oct 13, 2010 - 6:04pm

I refuse to let things play out, I want to complain NOW, DAMMIT!... ... ... OK, yeah, that's a fair point. Letting things play out is fair to a certain degree; it's not like I expect the model to keep Oregon low indefinitely.

However, the point I'm raising about Oregon is that 30th is WAY outside the boundaries of where almost anyone else has them. looking at: , the computer consensus is 3rd, and something like 90% have them top 10 (which is a lot, given the number of wacky, off-beat models on that list). The worst rating was 23rd, the 2nd worst was 20th, and everyone else was 16th or better. So having them 30th is an enormous outlier, so unusual that it pretty much demands to be addressed in some way (doesn't necessarily mean it's wrong, but it does need to be defended).

Personally, I think it's wrong because the S&P model over-penalizes teams for playing REALLY bad teams (which I've theorized in the past as well), even if they totally thrash them (if the UNM and AA games never happened, would Oregon's rating have been more like 20th? I would guess so [just a guess], and since those games were clearly ass-kickings, it really shouldn't be relevant whether those two opponents were bad or REALLY bad).

As far Oregon's resume goes, they were (judging by scores) far more dominant at Tennessee than 21st ranked Florida, much less 14th ranked LSU (who got them at home). Yes, Georgia also won handily, but overall that was an impressive performance, not a mediocre one.

They also won by 21 against 19th rated Stanford, who did struggle vs 25th ranked USC, but also lit up 22nd ranked Notre Dame (and I do wonder why that gap is so small), totally thrashed 78th ranked Wake, and won 35-0 at 33rd ranked UCLA. To be honest, I would consider that such an impressive win for Oregon that with nothing else of note (and the Tenn win was good too) I would need good reason not to have them in, or at the absolute least VERY close to, the top 10. If the model penalizes UO for being down early, that may be fair, but it needs to be highlighted if it's a material contributing factor.

As for ASU, they're 1-3 vs 1-A, but they also very nearly won AT Wisconsin (who lost to Michigan St and somehow is still rated above Oregon), and is a somewhat respectable 47th. If that's the big wart... yeah it's a wart, but not THAT big of one.

To take a different tack, why is Oregon rated well below the following teams:

Michigan St - seem to have an even worse schedule (played a AA team, a Sun Belt team, a MAC team, a mediocre ND team, and then Wisconsin and Michigan), or at best only a mildly more difficult schedule, and have been far less dominant than Oregon

Illinois - also played a AA team, barely beat NIU, did beat up Penn St, but also had TWO losses, though both were to teams S&P likes a lot

Florida - I would think Oregon wins the Tennessee comparison by a good amount. The Gators also lost twice, and LSU's only ranked 14th. MiamiOH is rated worse than Wazzu, so that's about a wash. Maybe it's because the model's killing Oregon for simply suiting up against UNM and AA Port St?

Miami - Got crushed by FSU, plus had Ohio St loss. Were dominant at Pitt, but I would think that's less impressing than UO's 21 point win against a better Stanford team. Won by 9 at Clemson, which strikes me as comparable to UO's win @ ASU (though S&P likes Clemson much more than I do, so maybe there's a meaningful gap there). Beat up a AA team, which is comparable to UO doing the same.

So if Pitt/Stan, Clem/ASU, and AA/AA are washes (or at least close), you're left with Miami's 12-pt Ohio St loss, and 28-pt FSU loss vs Oregon's 35-pt Tenn win, 72-0 UNM win, and 20-pt Wazzu win. I wouldn't go nuts if you had the two rated fairly close, since those were very good opponents Miami lost to and very crummy ones UO beat... but Miami is 3rd and Oregon is 30th! I mean come on, how does that make sense? I just for the life of me can't see it at all.

P.S. Total brain-fart about the AP comparisons on Friday. I completely forgot about them. Sorry about that.

P.P.S. My posts have probably been overly critical and hostile. I apologize for that. But this really does look like a model bust to me. "no-doubt top 10" is probably too much, but throwing them at #30... I simply can't see the justification for it. And to be honest, I think that early to mid-season numbers are a good time for this sort of analysis, since by the end of the year pretty much any model would have an undefeated AQ in the top 10, and the differences between models, especially for the top few teams, tend to be smaller. It's earlier in the year that things like these stand out, and that's why I think it's good to highlight them and talk about them.

P.P.P.S. As an aside, if it DOES turn out that the AA game was a major drag on Oregon's resume, despite the fact that they beat the hell out of them, I'd consider that to be a good argument against using AA games at all, since they have the potential to blast teams simply for playing them, no matter how ridiculously dominant the win was.

P.P.P.P.S. You don't actually have to answer this post, it was more me getting the laundry list of objections off my chest. Admittedly, I'm a Pac-10 guy, and I REALLY didn't like to see that number. So in all fairness, I myself have a bias in the matter.

Login or Register to post comments

24 Re: Oregon at #30

by mm // Oct 14, 2010 - 12:34pm

Only looking at point totals for a team that has a lot of possessions will be misleading. A small difference in quality can multiply into a big difference in points when there are a lot of drives. If they encounter a team better than they are, then those many possessions can turn just as quickly as a lot of points against them.

Now, I'm not saying that this rating is 'right', but both of FO's systems are designed to look at the average performance per play or per drive. There is a lot of merit in that idea. A hypothetical team that scores a touchdown every time they touch the ball, but takes 15 minutes per drive, will have far less of a point differential than Oregon. That doesn't mean that Oregon would beat the hypothetical team.

Login or Register to post comments

25 Re: Oregon at #30

by Bill Connelly // Oct 14, 2010 - 12:40pm

I'm obviously going to be addressing this in VN tomorrow, but one thing: you're annoyed by this as a Pac-10 guy, even though the Pac-10 currently has the No. 7, No. 8, No. 13, No. 19, No. 25 and No. 30 teams in the rankings? The only conference more well-represented than the Pac-10 right now is the SEC.

Login or Register to post comments

26 Re: Oregon at #30

by cfn_ms // Oct 14, 2010 - 1:32pm

I would say that there shouldn't be any league more represented than the Pac-10 this year (at least as a proportion; obviously only having 10 teams drops the # represented). IMO the league's OOC achievements speak for themselves, while no one else's has really come close.

P.S. Also worth noting that there aren't even any top 5 from the league, which seems odd to me. Oregon is the obvious objection, though I struggle to see why, for instance, Miami is rated above any of Oregon, Arizona, and Stanford (all three are 1 or fewer loss teams, all have beaten teams better than Miami's best victim - Pitt), much less all of them.

Login or Register to post comments

20 Re: Oregon at #30

by Kal // Oct 13, 2010 - 9:09pm

While I agree that Oregon hasn't had the toughest schedule, I think it's a bit disingenuous to look at the win-loss records of the teams they've beaten (ASU) and then say how bad those teams are. ASU played OSU to 3, Wisconsin to 1, and Oregon beating them by 11 is a horrible game? Similarly, beating Stanford by 21 was somehow not as informative as beating WSU by 20 and ASU by 11? That seems just odd to me.

And while I agree that there is more proof to be had in the system, looking at the other teams and who they've played isn't much more informative. Ohio State, Nebraska, Arizona, BSU...they all have blemishes of these natures. I guess it's hard for me to see a team that dominated bad competition aside from one game where they lost their starting QB and still managed a 20-point win on the road as worse than (for example) Ohio State losing their starting QB and managing an 11-point win against Illinois, but S&P sees it as quite damning.

Login or Register to post comments

21 Re: Oregon at #30

by Will // Oct 13, 2010 - 10:01pm

You are missing the fact that Illinois is a top 10 team, so an 11 point win over them on the road is indeed impressive. Now, Illinois could just as easily fall apart the rest of the season and take Ohio State down with them, but as of right now, it's a better win than anything Oregon has done in the S&P+ world.

Also, remember that Oregon was #15 last week, so the struggles against Wazzu really, really hurt because they are really, really bad. Look beyond the rankings though - Oregon is roughly 36 points behind South Carolina for #4, or roughly the same gap as between #1 Ohio State and #2 Alabama.


Login or Register to post comments

22 Re: Oregon at #30

by cfn_ms // Oct 13, 2010 - 11:04pm

"the fact that Illinois is a top 10 team". Uh-huh... riiiight... because 3-2 with two double-digit losses, a way too tough outing against NIU, a AA win, and only one solid showing, 33-13 at Penn St (who's 3-3 with a AA win and two MAC wins, plus three 20+ point losses), is a top 10 resume. Of course...

For the rest of us, an 11-pt win at Illinois is a nice win but hardly a massive resume-booster. That said, I'm sure reasonable people can argue for Ohio St over Oregon. I don't agree but I don't have much of a problem with it either.

If Will would like to take a stab at why Illinois ought to be rated well over Oregon, though, I'd be happy to hear it. Bonus points if he can explain how a 6-point home win over NIU (68th in S&P) is substantially less bad than a 20-point win at Wash St (99 in S&P, hardly the utter dregs of 1-A).

Login or Register to post comments

23 Re: Oregon at #30

by Will // Oct 14, 2010 - 2:13am

I'm not arguing it, S&P+ is - I'm just quoting it's numbers. The system is internally consistent, and I was just trying to show reasons why Oregon is rated so low in this system.

I personally don't put too much faith into any computer ranking until very, very late in the season. Up until then, there just isn't enough connectivity for me to take too seriously.


Login or Register to post comments

27 Re: VN: The Fun Has Begun

by southpaw2 // Oct 15, 2010 - 12:16pm

Nice shout on Les's post game comments. I need to check if that is up on youtube, his facial expressions were a blend of vindicated, embarrassed, contented and angry that you don't see very often, even on the gridiron.

Not to keep harping on and against the "truths", but the Gamecocks put Bama away with the punishing drives they put together in the second half, after Garcia's "interesting" call to give up the safety threatened to decimate their momentum. The way they marched the better part of eighty yards and converted multiple third and longs along the way absolutely took the life out of Bama's defense. I would agree completely that Spurrier's offensive gameplan was brilliant throughout the entire game, but it was his players' second half execution that really stood out to me. I think most people who watched the game would agree.

And, Alshon Jeffrey was a man playing against children.

Login or Register to post comments