Is Oklahoma-TCU or Baylor-Texas more likely to be an upset?

To both casual observers and those more statistically inclined, the Big 12 hierarchy through one month seems clear: there's No. 4 Oklahoma, then there's No. 7 Baylor, then there's everybody else.

West Virginia has played well against two of the four best teams in the country (OU and Alabama), No. 25 TCU has passed the tests it's been given, and both No. 21 Oklahoma State and No. 23 Kansas State created opportunities to beat Florida State and Auburn, respectively.

The Big 12 has quality and potential. But there are still two names at the top.

If that's going to change, then we will understand that within the next couple of weeks. The Red River Shootout looms (and before you scoff at the idea of Texas beating Oklahoma, remember that you scoffed heading into last year's game, when Texas won by 16), as does a TCU visit to Waco. But this week, both Oklahoma and Baylor must head to hostile territory; Baylor visits Austin, while Oklahoma goes to Fort Worth.

Between Oklahoma-TCU and Baylor-Texas, which home underdog is more likely to win?

(The short answer is, "the former." But you want to read the long answer, right?)

Oklahoma at TCU, 3:30 p.m. ET, FOX

Key stat No. 1: TCU is No. 1 in the country with a Havoc Rate of 25.9 percent. One of every four defensive plays has resulted in either a tackle for loss, forced fumble, or pass defensed. The defensive line's Havoc Rate is 12.7 percent, better than the entire defense for 21 FBS teams.

Key stat No. 2: Oklahoma's offense ranks 11th in Adj. Sack Rate, 25th in Adj. Line Yards, and 22nd in Stuff Rate (run stops behind the line).

Confused about any of these stats? Here's the glossary.

TCU outscored Samford, Minnesota, and SMU by a combined score of 134-21. Granted, Samford is FCS and SMU is absolutely terrible. But beating Minnesota, 30-7, while outgaining the Gophers by a 427-268 margin is impressive; Minnesota is 36th in the F/+ rankings.

Defense drove TCU's outstanding performance against Minnesota. Senior linebacker Paul Dawson had four tackles for loss, a forced fumble, and a break-up by himself. (Minnesota's entire defense: four tackles for loss, two forced fumbles, four passes defensed.) Safety Derrick Kindred added 0.5 tackles for loss and three passes defensed. Tackle Davion Pierson had 1.5 sacks. End Josh Carraway had 1.5 tackles for loss. TCU outflanked Minnesota on every level of its defense. The Frogs entered the season with one of the most experienced, exciting secondaries in the country, but the front six has been untouchable.

Basically, TCU stops the run on standard downs and dominates the pass on passing downs. The 4-2-5 defense is difficult to perfect -- few outside of TCU have mastered it -- but if you pull it off, it is the perfect defense for stopping spread offenses, and it still maintains enough size and discipline to stop strong running games as well.

Oklahoma has moved the ball just fine, but this game represents the stiffest test for both the OU offense and the TCU D.

The key for Oklahoma might be keeping TCU off-balance. Samaje Perine had an incredible night against West Virginia two weeks ago, rushing 34 times for 242 yards and four scores, proving that OU can run the ball without injured starter Keith Ford (who is expected to miss the TCU game). But if he and Alex Ross aren't able to establish the same level of dominance against a much better defense, the Sooners might have to resort to its quick passing game to stay ahead of schedule. Passing downs are death against the Horned Frogs.

Quarterback Trevor Knight has been solid but not spectacular -- 59 percent completion rate, 132.8 passer rating, four touchdowns, three interceptions -- but his ball distribution has been unique. Knowing that Sterling Shepard is the only proven receiver, Knight has tried to give him plenty of standard-downs targets (when the defense still has to account for a good running game) and distribute the ball as evenly as possible on passing downs.

Receiver	Standard Downs	Passing Downs
Sterling Shepard	24 targets, 15 catches, 297 yards (12.4/target)	15 targets, 8 catches, 139 yards (9.3)
Durron Neal	18 targets, 11 catches, 171 yards (9.5)	10 targets, 6 catches, 40 yards (4.0)
K.J. Young	7 targets, 7 catches, 63 yards (9.0)	12 targets, 7 catches, 104 yards (8.7)

Tight end Blake Bell got involved against Tennessee, catching three passes for 52 yards, but for the most part OU's passing game has consisted of the three wideouts above. They will face a significant test from the TCU secondary, and the trick might be getting Shepard the ball on first-and-10 and attempting a higher than normal percentage of rushes on second- or third-and-long.

Key stat No. 3: TCU's offense is No. 1 in the country in raw Success Rate and No. 67 in raw IsoPPP (which measures the magnitude of successful plays).

Key stat No. 4: Oklahoma's defense is No. 5 in the country in raw Success Rate and No. 76 in raw IsoPPP.

That TCU has a good defense isn't a surprise. The Horned Frogs always have a good defense. They went 4-8 last year but still had the No. 12 D in the country, according to Def. F/+.

No, the problem for the Frogs has unfolded when the defense isn't on the field. They ranked 94th in Off. F/+ last season, marred by a chicken-and-egg effect: inconsistent quarterback play and no run game.

This offseason, Gary Patterson brought in two new offensive co-coordinators: Texas Tech's Sonny Cumbie and Houston's Doug Meacham. They were tasked with restoring some semblance of efficiency in the TCU offense, and through three games, they've more than done that. The Horned Frogs' 57.4 percent success rate is the best in the country through three games (only four other schools are above 54.2 percent: Alabama, Michigan State, NC State, and Texas A&M), and even if you adjust for opponent, that's still an awesome average. And they produced a 50 percent success rate against a Minnesota defense that currently ranks 12th in Def. S&P+ and 26th in success rate.

That TCU has produced such efficiency with quarterback Trevone Boykin in a pass-first offense is stunning. Boykin has been a starter for large portions of the last two seasons, but he entered 2014 with a 58 percent completion rate and 17 interceptions off-setting 22 touchdowns. His numbers this year: 64 percent completion rate, eight touchdowns, one pick. After a career of 120s passer ratings, he's at 142.7. He held off Texas A&M transfer Matt Joeckel in fall camp (which surprised many, including me) and has thrived. He's on pace for 3,500+ passing yards and 700+ rushing yards.

Boykin's spreading the ball around incredibly evenly. He's targeted five receivers between 14 and 22 times. Three have catch rates of at least 72 percent, and one (Kolby Listenbee) is averaging an all-or-nothing 18.8 yards per catch.

Of course, Oklahoma might have the best defense the Frogs see all year. The Sooner defense has been every bit as efficient as the TCU offense, and the Sooners have already faced two competent offenses, Tennessee and West Virginia. The Sooners are holding up against the run and making all sorts of havoc plays with linebackers and safeties.

If the TCU 4-2-5 isn't the perfect defense for stopping a spread-'em-out attack, OU's hybridized 3-4 might be. Outside linebackers Eric Striker and Geneo Grissom have combined for seven tackles for loss and five passes defensed in four games, strong safety Quentin Hayes has three sacks, and linemen Chuka Ndulue and Jordan Phillips are giving the OU line a play-making presence it hasn't had. The Sooners rank fifth in Def. S&P+, masterfully leveraging opponents into passing downs, swarming the ball, and creating harried quarterbacks.

If Boykin's numbers are still as strong after this one, he'll have earned those stats. Boykin was only decent with his arm against Minnesota -- he rushed 12 times for 92 yards but completed only 27 of 46 passes at 9.6 yards per completion -- and we'll see if TCU can maintain that efficiency. Oklahoma's been pretty leaky in the big-play department, but aside from the occasional toss to Listenbee, TCU doesn't make many big plays. The Frogs will have to stay on schedule against a mean, fast Sooner defense.

Baylor at Texas, 3:30 p.m. ET, ABC

Key stat No. 1: Baylor's defense has allowed a 99.6 passer rating, 49.6 percent completion rate, 10.7 yards per completion, four touchdowns, and three interceptions, and it's produced a 12.2 percent sack rate (16.4 percent in non-garbage time, best in the country).

Key stat No. 2: In Baylor's last nine losses, the Bears have allowed a passer rating of at least 146.0. If you can pass on Baylor, you can keep up with Baylor.

Key stat No. 3: In six appearances, Texas quarterback Tyrone Swoopes has produced a rating of better than 130.0 just once: 138.4 against UCLA.

Texas' defense has allowed just 4.1 yards per play and allowed just 27 combined points in three games against teams not named BYU. The Longhorns are active up front and steady in the back, currently seventh in IsoPPP and ninth in unadjusted Passing S&P. It is fair to say that the Longhorns will have more success against Baylor's offense than SMU (45 points, 5.9 yards per play), Northwestern State (70, 9.2), Buffalo (63, 8.6) or Iowa State (49, 6.2) did.

But let's be honest: Baylor's going to score. The odds of Texas pulling an Oklahoma State and holding Baylor under 20 points are unlikely.

So the questions for this game are easy: can Texas keep up? Can the Longhorns put enough points on the board to give themselves a chance? And was I mistaken in calling Oklahoma's defense the best in the conference? Because nobody's moved the ball on Baylor.

Opponent-adjusted numbers are unstable just five weeks in, but the fact that Iowa State had a little bit of success against Baylor's defense and moved from 88th in Off. S&P+ to 34th in one week tells you how the Bears performed in their other three games. SMU, Northwestern State, and Buffalo combined for 27 points, 3.4 yards per play, and an absurd non-garbage time success rate of just 17.0 percent. (Iowa State's: a prolific-in-comparison 26 percent.) But they're SMU, Northwestern State, and Buffalo.

So what does Texas have to offer? How do the Longhorns intend to score the minimum of 24 to 28 points it will take to take down Baylor? I'm not sure I know the answer. The Longhorns would be happy to run the ball as frequently as possible, but they're not good at it; Johnathan Gray and Malcolm Brown have combined to average just 4.0 yards per carry behind a line that has suffered all sorts of personnel attrition. And while Swoopes seems to be improving, he'll need to step further out of his comfort zone against the aggressive Bears.

Against UCLA and Kansas, Swoopes had a 63 percent completion rate (9.6 yards per completion) with four touchdowns and no picks. He isn't doing a lot of damage, but he's also not making killer mistakes. (He's also been a bit lucky; Texas opponents have broken up 10 passes and intercepted only one David Ash pass, and on average, you would expect two to three picks among 11 passes defensed.)

Baylor defensive coordinator Phil Bennett has all but mastered the defense he brought to Waco. Last year, we saw a BU defense that swarmed and grabbed and hit hard, attempting to overwhelm with speed and meanness. The Bears had to replace five of last year's top seven defensive backs, plus linebacker/wrecking ball Eddie Lackey, so it was easy to assume a 2014 drop-off. Maybe that will take shape, but it hasn't yet.

The Bears are averaging 8.5 tackles for loss per game (sixth in the country) and a nation's-best 4.5 sacks per game. Big Shawn Oakman and tackle Beau Blackshear have combined for 7.5 sacks, while sophomore cornerback Xavien Howard has defensed eight passes (two picks, six breakups), sixth in the country. Linebacker Bryce Hager is doing what Bryce Hager does: clean up messes and hit you really hard. It's all worked. The Bears are preventing all big and small plays.

You can beat Baylor's defense; UCF and Oklahoma State did. But you have to have a quarterback who maintains his poise and skill-position players who are ready to scrap. Receiver John Harris is talking like someone itching for a fight -- "They're still Baylor ... we're still Texas." -- but Baylor will force him to back it up? Can he? And if he doesn't, who else does Texas have that might?

And does this even matter if Baylor's line is overwhelming Texas' line? Texas has been living off of turnovers -- how much of a positive turnover margin do the Longhorns need to have a shot?

***

F/+ Projections
Oklahoma by 6.0 (64 percent chance of winning)
Baylor by 16.4 (76 percent chance of winning)

Conference road games are never the easiest. If the home team gets a couple of turnovers and semi-easy scores and the home crowd gets rolling, you never know how a game will unfold. But by the numbers, Oklahoma would win two of three games at TCU and Baylor would win three of four against Texas. Texas will need impressive help from turnovers, but it's not hard to lay out the script of a TCU victory.

The Horned Frogs have played Oklahoma well under Patterson -- a 20-17 road loss in 2013, a 24-17 home loss in 2012, a 35-10 loss in 2008 (which was closer than just about anybody else came to slowing down Oklahoma that year), a 17-10 win in 2005 -- and should have the athletes to match up with Oklahoma's offense. There's a chance that the Sooner run game is too overpowered, but that's not a certainty.

If TCU can stay ahead of the chains and avoid mistakes (let poor drives end with punts, not turnovers), a win is within reach. This is a serious hurdle for Oklahoma; the Sooners can probably overcome a loss and still make the College Football Playoff, but a loss here would put pressure on them to win out.