USC was also among the nation's worst at converting on 3rd down - 110th after converting for yards needed for the first down. Consider this - Texas A&M was 50% more likely to convert on 3rd down and get at least four more shots at the end zone than USC. One consequence cutting drives short by turnovers and failing to convert on 3rd down was short drives - 5.39 plays per possession, 16th fewest in the country. USC still managed 34 yards and 2.3 points per possession, decent but not great numbers, but the offense would have been downright dynamic if they had converted on 3rd downs and avoided turnovers.
Outside of the interceptions, the USC passing game was a strength - top 10 to top 30 in all major categories after adjusting for competition, 18th most efficient overall - even if it was a disappointment relative to the individual talent level on the field. The running game was a different story. Despite averaging 5.5 yards per carry and an explosive play on 2.9% of runs, the numbers put the USC run game at 83rd in overall efficiency. The high fumble rate was half the problem, and USC runners lost yards 20% of the time. Inconsistency in the running game (they flip-flopped between big play and negative play) meant that it was unreliable in short yardage situations. USC ran the ball 58% of the time in those situations (94 teams ran more often). This undoubtedly contributed to their 3rd down woes.
A common narrative for USC is that they can't defend the spread. They did allow 7.1 yards per carry against Oregon, but all that tells us is that they struggled against Oregon, not that they struggled against the spread. They also allowed 5.5 and 5.3 yards per carry against Stanford and Notre Dame. The problem wasn't spread offenses but teams that knew how to run the ball. Against most opponents, USC was fairly good at preventing explosive running plays, but they were among the nation's bottom third at preventing runners from crossing the line of scrimmage. Especially against Stanford and Notre Dame, the line of scrimmage got pushed back a couple of yards before the runner caught up.
The struggle against the run may have been schematic. USC was fairly strong against the pass. They kept the completion percentage low and prevented explosive plays, but more impressive, they sacked the quarterback once for every nine pass attempts. The odd thing is that USC got to the quarterback more often on non-passing downs than passing downs. Whoever was bringing pressure on 1st down and 2nd and short wasn't in position to defend the run. This isn't an inherent weakness of the Tampa 2, but it is easy to see that it could become a problem.
The only positive I see on the USC offense is that Marqise Lee is really, really good. But he was really, really good last year, too. They will take a step back at quarterback, Silas Redd doesn't impress, and the receiving corps will be missing Robert Woods. I would normally argue that the turnovers should regress, but they turned the ball over more last year when Matt Barkley wasn't on the field, and he won't be on the field at all in 2013.
On defense, USC is moving to a five man front. The impact defensively will vary depending on whether it evolves more towards a 52 or a 34. The bigger question is how much of a headache will the Trojans suffer as they make the transition.
The saving grace could be the schedule. UCLA gets Washington, Oregon, Stanford and Cal. USC trades Oregon and Washington for the in-state land grant institutions. And they get Arizona State at home. If you don't think that remembers, just know that Texas A&M would have been playing Georgia for the SEC title while Alabama watched if Bama and A&M had traded Tennessee and Florida as SEC East opponents.
*We use EPA to estimate the point value of a turnover. We use the down, distance, spot and time remaining to estimate how many points a team can expect to score on that possession and their opponent will score on the next possession. The point impact of a play, then, is the difference in the expected points before and after the play.