CFB · UPSET-WATCH · PREVIEWS · WEEK-8 · SOUTH-CAROLINA · TENNESSEE
Upset Watch Week 8: Our Sims Take South Carolina Over No. 11 Tennessee, and Auburn-LSU Is a Coin Flip
By MVPeav · @MVPeav ·
Tennessee beat Alabama 26-10 last Saturday, sits at No. 11, and rolls into Columbia this week as the team everybody expects to win. My sims don't care about any of that. They have South Carolina winning this game 59 times out of 100.
That's the headliner, but it's not the only live dog on the Week 8 card. I ran all of this weekend's games 10,000 times each and ended up with these 6 catching my eye. Out of the 6 we came away with one outright flip, one coin flip on the Plains, and two games where my sims and ESPN's FPI are 20-plus points apart on the underdog.
Underdog here means the team the outside world expects to lose, by the rankings or by FPI. Sometimes that's a 3-3 team at home against a ranked one. Sometimes it's No. 7 Ole Miss on the road. The rankings don't get a vote in my sims. The matchup does.
South Carolina 59% vs Tennessee: The Sims Flip This One
Start with the only game on the slate where my sims take the underdog outright. South Carolina wins 59% of 10,000 sims against Tennessee. ESPN FPI gives the Gamecocks 40.1%. The typical sim lands 28-24 South Carolina, and half of them fall between a 7-point Vols win and a 13-point Gamecocks win.
The mechanism is a little backwards and I love it. Tennessee throws for 188.0 yards a game, 120th in FBS. South Carolina's pass defense is 104th, giving up 237.2 a game. You'd figure the bad secondary bails out the bad passing game, and in my sims it does. The Vols throw for 245 against this defense.
That's the problem for Tennessee, tho. The run game, the thing this offense actually lives on, gets held to 139 yards in my sims, right at what South Carolina allows. So Tennessee turns into a team that throws more and runs less, and it's doing that with a backup at quarterback. George MacIntyre is expected to make a third straight start in place of freshman Faizon Brandon, who hurt his right shoulder against Texas.
On the other side, South Carolina's offense is scoring 37.7 a game, 15th in FBS, and it comes off a bye with Shane Beamer expecting nearly everybody back. Quay'Sheed Scott is the only real question. This is a 3-3 team that lost to Kentucky 34-35 and Mississippi State 34-41 at home, so the record is uglier than the roster. It also went to Florida and won 38-19, which is the type of game where the talent that my sims keep seeing plays like they are meant to.
Tennessee allows 14.7 points a game, 11th in FBS, and that defense is the whole reason the Vols are 6-1. South Carolina's defense is 90th in points allowed, so Tennessee doesn't need 35 to win. My sims just don't think a backup-QB offense that can't run gets to 28 in Williams-Brice.
Auburn vs LSU: 49% Is a Coin Flip at Home
Auburn wins 49% of my sims against No. 18 LSU. FPI gives the Tigers 28.7%. The typical sim comes out LSU 25, Auburn 24, and half of them sit inside a window that runs from a 10-point LSU win on one end to a 10-point Auburn win on the other. That's a coin flip in a stadium that's going to be loud.
This is the same Auburn team that walked into Athens and beat Georgia 27-24, so the idea of them beating a ranked SEC team shouldn't shock anybody.
The way it happens in my sims is on money downs. LSU's offense has been converting 57% of its third downs. Against Auburn, that drops to 36.6%, and it's no coincidence that's exactly what Auburn's defense allows. The drives die before they get anywhere: LSU's red-zone trips go from about 5.3 a game to 3.8 in this matchup. Auburn's defense doesn't give up the cheap one either, 21st in FBS in explosive-play rate allowed at 10%.
Then you've got Byrum Brown, who leads Auburn in rushing with 459 yards and put 265 total yards and 2 touchdowns on Georgia. A quarterback who can run turns a 3-and-out into a first down, and LSU's defense has to account for it.
The trouble for Auburn is what Brown has to run into. LSU's defense is 1st in FBS in EPA per play at -0.17. Brown running into that is a different proposition than Brown running at Georgia. LSU's offense is 9th in success rate at 51% too, so the third-down number in my sims is asking Auburn to do something to a good offense that it hasn't had to do yet. Still, 49% at home is a flip, and I'll take a flip at Jordan-Hare over FPI's 28.7 every time.
Ole Miss at Texas: The Ranking Says Underdog, the Sims Say 42%
This is definitely a weird one. No. 7 Ole Miss is the underdog on the road at No. 19 Texas, and my sims mostly agree with that. My sims give Texas 58%, and the one right in the middle of the pile reads 27-24. But Ole Miss's 42% is a long way from FPI's 28.9%, and the gap is Texas's own offense.
Ole Miss brings the No. 1 success-rate offense in FBS at 52%. Trinidad Chambliss has thrown for 1,676 yards and 14 touchdowns in six games, and this offense already beat LSU 32-24 at home. Texas's defense is the one unit that can slow that down: 1st in FBS in explosive-play rate allowed at 7%. In my sims the Rebels' yards per carry come out at 3.4, and their red-zone touchdown rate falls from 87.5% to 53.7%. Texas lets you drive. It doesn't let you finish.
And Texas does that without Colin Simmons, who leads the nation with 10.5 sacks and isn't expected to play with a high-ankle sprain. Lose the best pass rusher in the country and 58% starts to feel generous.
The reason the Rebels stay live is that Texas can't score either. The Longhorns' offense is 95th in EPA per play at +0.01 and sits at -3 in turnover margin. Ole Miss's defense is 102nd in EPA per play, so it's a bad offense against a bad defense, and somebody has to move the ball. This Texas team beat Ohio State 24-23 and then lost to Oklahoma 24-27 and Florida 24-27 (I'm not sure anymore which of those is the real Texas, tho the last two games are the fresher evidence). A 27-24 median with the No. 1 success-rate offense on the other sideline is a thin margin to lean on.
Iowa State at Arizona: Kill the Third Downs, Steal a Win
Iowa State wins 40% of my sims at No. 25 Arizona. FPI says 24.4%. The median is Arizona 24, Iowa State 20.
Arizona's passing game is the real thing, 303.8 yards a game, 8th in FBS, at 9.0 yards per attempt. The Wildcats convert 58.7% of their third downs, and that's how you keep a passing offense on the field. In my sims that number craters to 37.6% against Iowa State, because the Cyclones allow 32.8% on third down and that's one of the best marks you'll find. Arizona's passing yards fall to 235 in this matchup. Not shut down, just shortened.
That's the Iowa State path: make Arizona punt three or four more times than it's used to and keep it a one-score game into the fourth.
The catch is that the Cyclones have to score too, and this 3-3 offense is 115th in passing at 178.2 yards a game. The defense is also 121st in yards per carry allowed at 4.4, so if Arizona decides to run instead of throw, the third-down trick doesn't matter much. The 40% holds up, but it comes with a lot of "if Iowa State can get to 20."
Vanderbilt at Kentucky: Losing Streak, Live Dog Anyway
My sims still favor Kentucky, 61% to 39%. But 39% for a 3-4 Vanderbilt team on a four-game losing streak is the biggest gap from FPI on the whole card. ESPN gives the Commodores 14.6%, about 25 points lower. The median is Kentucky 28, Vanderbilt 24.
The case for Vandy is Kentucky's defense, which just isn't good. The Wildcats are 82nd in FBS in yards allowed at 377.0 a game and 106th against the pass at 249.7. No. 12 Kentucky got to No. 12 on a +11 turnover margin, 4th in FBS, and its wins over LSU and South Carolina were both 35-34. Two one-point wins and a turnover margin like that is the profile of a team that's been getting the bounces, and my sims don't hand out bounces.
The sims do give Kentucky one thing: its offense converts 30% of its third downs coming in and 40.7% in this game, because Vanderbilt's defense allows 44.6% and lets everybody look better. That's most of why Kentucky still wins 61 of 100.
Vandy is losing games for a reason, though. It's -5 in turnover margin, 119th in FBS, and it can't run, 119th in yards per carry at 3.6. A team that gives the ball away against a team that lives on taking it is a bad recipe. Live dog, not a flip.
Indiana at Michigan: The Big House Run Wall
Indiana is No. 6 and the favorite in my sims at 62%. Michigan's 38% at home is still more than double FPI's 17.8%, and the median score, Indiana 24, Michigan 20, is a lot closer than the rankings suggest.
Michigan's shot is the run defense. Indiana has been running at 6.1 yards a carry, and in my sims that drops to 3.7 against a Michigan front that allows 3.5. Take away the explosive run game and Indiana has to win it through the air, in Ann Arbor, in a game where my sims expect the Hoosiers to throw their first interception of the season. Zero picks through seven games is impressive. It's also a streak that ends on the road against a defense like this.
All that said, Indiana is No. 6 for a reason and still wins 62 of 100 even with the ground game smothered. A 4-2 Michigan team needs the run wall to hold and the turnover to come. In my sims both happen often enough to make this a game, not a coronation.
Bottom Line: Columbia Is the Spot
If you want one upset from this card, it's South Carolina over Tennessee, the only game where my sims flip the favorite. The letdown setup is huge, with Tennessee coming off the emotional rivalry win over Alabama and now going on the road with a backup quarterback against a team that has already shown once this year it has an upset in it. That's the exact spot teams lose.
Auburn at 49% is the tough one. Do they come out energized after that hard fought win over UGA or do they have a let down after drinking too much of the koolaid from that win. I wouldn't be surprised by either result, but I'd be surprised if it's not close.
Vanderbilt at 39% and Michigan at 38% are the games where my sims and ESPN disagree the most, and in both the sims still take the favorite. Live dogs, not upsets, until somebody holds Indiana under 4 yards a carry or Kentucky's luck runs out.