Earlier I stated that Clemson's offense was one of the toughest units in the nation to judge. Well, the Clemson defense is one of the easiest-they have underperformed against subpar competition. While I enjoy reading STS's weekly take on the Clemson defense and despite giving up 48 points to NC State this week, the Clemson defense is actually improved this year. However, the improvement is a case of a really bad defense becoming a mediocre one.

Last week, Clemson's defense gave up two long touchdowns to NC State, a 77 yard TD pass and a 49 yard TD pass. This continued its trend of allowing big plays. I counted 16 plays of over 40 yards-by comparison South Carolina has only given up four. Additionally, five of those 16 are runs, which is surprising since most big plays are pass plays.

However, Clemson's defense does not fail at a greater than average rate. They allow 44% of pass plays and 41% of runs to be successful (59th and 54th in the nation respectively). It's just that when they fail, they fail spectacularly. I have not watched many Clemson games this year, but those numbers tell me that their linebackers and secondary miss tackles or take bad angles on runs and just straight-up bust coverages on passes. This is where South Carolina will miss Lattimore. He excels at finding a small hole for medium runs which would provide numerous opportunities for him to break a long run. If he gets the majority of carries, Kenny Miles will need to find a way to run through tackles this week. If he doesn't, I hope the coaching staff turns to Mike Davis similar to how they replaced Brandon Wilds with Kenny Miles early in the game last year.

Because of Clemson's schedule, it is difficult to compare South Carolina to any Clemson opponents. South Carolina has the biggest split between rush/pass offense of any team on Clemson's schedule. On a per play basis, the Gamecocks run the ball like Auburn or Duke, but pass better than Florida State or anyone Clemson has played this year. If you don't believe me, check out this stat: the Gamecocks average 3.7 yards per rush this year vs. 4.5 last year. I suspect that number is even worse now that Lattimore is out. However, South Carolina averages 8.5 yards per passing attempt this year vs. 7.4 last year.

Here is how South Carolina's offense stacks up versus the other teams Clemson has played. All numbers are my own based on play-by-play data.

As you can see, Clemson had quite a nice stretch on defense after the Georgia Tech game until this past week's disaster. However, you'll also notice that those past four offenses were pretty terrible, with only Duke coming close to mediocrity. The Gamecocks bring the best offense that Clemson has seen since Georgia Tech and the best passing game (on a per play basis) that they will see all season.

In many ways, Connor Shaw is the key to this game. He is both the best running and passing threat the Gamecocks have on offense. If he rediscovers is running magic after a couple of cortisone shots expect the offense to do well. But more importantly, the Connor Shaw against Tennessee and Arkansas who found open receivers against weak secondaries has to show up.

As a quick summary, here are my keys for the offense:

• Connor Shaw needs to make the throws to open receivers, and by open I mean receivers who have a step on the defender or are about to make a cut into an open part of the field
• Kenny Miles or Mike Davis need to take advantage of bad tackling or angles by the Clemson defense. Miles in particular needs to run through contact at the line to get into the secondary to exploit their poor play
  
• The Gamecocks need to take advantage of playaction passes-Clemson will be expecting many rushes but the Gamecocks need to do their damage through the air, just as they did against Georgia, Arkansas and Tennessee
• An effective Shaw executing zone-read runs or scrambling is a plus--if this happens I expect South Carolina to win by multiple scores

So how many points can we expect the Gamecocks to score? I went ahead and ran a regression with the points per drive versus the rush/pass stats and came up with 2.45 points per drive, or 27 points at 11 drives or 32 points at 13 drives. If you remember, earlier I discussed 26 to 31 points for the Clemson offense. Expect a close game-I'll have more to say in a quick final preview later this week.