2012 Comparative Stats Post UCLA

USC sits at 7-4 with its starting QB out for two to eight weeks.  It is quite obvious that things have not gone as expected for USC.  With things potentially looking somewhat grim for the season finale against #1 ranked arch-rival Notre Dame, some fans have already started calling for Kiffin’s head.  However, I think this would be a good time to take a step back and try to take an objective look at exactly how good or bad this team is.  Let us try to take emotion out of the picture and just take a look at some comparative stats for USC.

What are Comparative Stats?

Comparative Stats look at a team’s production compared to what the opponent allows on average.  For instance, if USC were to gain only 250 yards of total offense against a mythical opponent, one might think “wow, USC’s offense is really bad.”  However, if this mythical opponent allows only 150 yards of total offense on average throughout a season, then suddenly the tune changes.  In this situation, USC has gained 100 yards more than what the opponent typically allowed, which shows a stronger USC offensive performance.

By using this over a number of games, you can see various trends on how good or bad a team is in a particular area.  A strong offense will consistently gain more yards than their opponents’ generally allow in various categories.  On the flip side, a poor offense will fall below the averages for comparative stats.

Please note that this is only part of the overall picture of football.  Winning in comparative stats in a particular game can be completely negated by committing turnovers, etc.  Using comparative stats is just one tool we can use to try to see how good a portion of a team is.

Please note that I have done something a little different with the Oregon game since I think they are a statistical anomaly.  The reason I say this is because the vast majority of their games have been over by halftime.  Oregon averages 32 points by half, which is greater than all but three of our opponents’ full game scoring average (Arizona, ASU, and UCLA average greater than 32 points per game).  So how do you accurately do comparative stats when Oregon is playing its 3rd and 4th stringers by the 4th quarter in most of their games while they are still playing their 1st stringers all game against USC?  My solution was to include Oregon’s comparative stats which only look at the first half.  This should normalize the stats based on when the first stringers were in.  I also left the full game stats in there for comparison because first half stats also don’t tell the whole picture.

So with that in mind, let us look at some comparative stats for USC’s offense.

Offensive Comparative Stats

First let us look at the comparative stats numbers for total offense.  The chart below shows USC’s total offense numbers in cardinal against the average total defense numbers of the particular opponent’s season in gold.  Since you’ll want to gain more yards than the opposing defense allows, in this case we want the cardinal bar to be higher than the gold one.

So what does this graph show us?  USC gained more yards of total offense than the opposing defense typically allowed in every game except for Stanford and Colorado.  USC averages gaining 66.9 more yards per game than their opponents allow.

Now let us take a different look at the same data.  This line chart looks at the margin between the opponent’s average total defense and USC total offense in the matchup.  So you take the gold bar and subtract the red bar to get this result.  In these types of graphs, I will have a yellow line for the Oregon game which shows what the margin is considering only the first half.  For this first line graph, the yellow line is almost exactly at the same data point as the red one, so it is pretty much not visible.  The dotted line shows the trend of the overall graph, which is useful to see how USC is doing as the season progresses.  For this graph you would want any data point in the positive side of the y-axis.

As you can see here, USC was consistently over the positive line other than Stanford and Colorado as noted earlier.  It is extremely positive to see the trendline as a fairly good positive slope.  However, it is a bit disconcerting that the last two games have been a drop off from earlier, but that could be just that the Oregon game was an anomaly.

Taking it a step further, I looked at the yards per offensive play comparative stats.  I won’t post the graphs here as it is fairly similar in looks to the total offense play.  I will note that USC averaged 1.2 yards more per play than defenses typically allowed.  This includes gaining 2.3 yards more per play than Arizona typically allowed.

Now let us take a look at rushing offense.

As you can see here, the rushing offense has been much less consistent.  On average, USC gains 3.4 fewer yards per game rushing than the defense typically allows.  The biggest offenders were the Hawaii and Colorado games, both of which we were held over a hundred yards under what the defense averages.  However, given the nature of those opponents, that may have just had to do with the play calling of the specific game (since we had great passing numbers for both of those).  Still, there were a number of games which we were held below average: namely Stanford and Arizona.  Oregon is tricky to call since we were above average for the first half but below average for the game.  The overall trend for rushing is positive, but not by much.

To remove play calling from the picture, let us look at the rushing yards per carry charts.

From this chart, the Hawaii and Colorado games, while still in the negative, are much less impacted.  Hawaii, Colorado, and Stanford were all about one yard per carry below what the defense allows.  On average, USC gained one yard per carry over what the defenses allowed.

Now for passing offense.

On average, USC gained 80 yards more through the air than defenses typically allowed.  Passing was below average against Syracuse and Cal (both games which had great running outputs) and about even against Stanford and Washington.  The general trend is positive, but again it is disconcerting that ASU and UCLA were both significantly lower outputs than the two previous games: Arizona and Oregon.

Now for completion percentage.

Completion percentage has been off and on.  Generally it is on the positive side of things, but not by much.  It also dips down negative in a number of games.  USC averaged about 4% better than typical on completion percentage, which pretty much just means right at average.  Given that our passing offense averaged more yards than typical, this implies that our yards per attempt is higher than average.

Now let us look at 3rd down conversions.

As you can see here, USC has improved greatly in their 3rd down conversion rate when compared to their opponent’s average.  USC generally was below average for the first seven games with the general breaking point happening at the Arizona game.  I believe much of this has to do with penalties.  Penalty yardage makes it much harder to convert first downs, especially if you get penalties for more than five yards (as you can see in my penalty analysis here).  As a summary, USC converts on average of 75% of its series of downs when there are no penalties but only 50% if there is a penalty and only 31% if there is a penalty of 10 or more yards.  While the number of penalties were high for Syracuse and Utah, the yards per penalty were much lower in those games, which helped us keep conversion rates up.  Ever since the Arizona game, our penalty numbers have been much more respectable and that directly leads to a much more respectable 3rd down conversion rate compared to what the opponent allows.  Overall, USC averages 2.5% higher 3rd down conversion rate than the defense typically allows.  However, if you only look at the last three games since our improvement, that number is 16.8% higher than typically allowed.

Now let us take a look at turnovers lost.  Note that a higher number of turnovers lost is bad, so in this case a negative margin is better.

As you can see here, USC has been higher than average for turnovers, which means they are having a below average performance.  This season, USC has averaged 0.7 more turnovers than the opposing defense typically allows.  USC has had a higher than average number of turnovers in every game except for three: Hawaii, Syracuse, and Washington.  Oregon was also pretty much even with their average in regards with the entire game, but positive when only looking at the first half.  As you can also see, the trendline is on a fairly sharp incline.  This is quite disconcerting as turnovers have been a large culprit in all of USC’s losses this season.

Now let us look at sacks allowed.  Again, a positive margin is a bad thing in this category.

In general, USC has done very well in protecting the quarterback.  Overall, USC has allowed one fewer sacks than the opposing defense typically gains.  Two games allowed an above average number of sacks allowed: Stanford and Arizona.  Two more games were at average: Syracuse and Washington.  Overall, the trend has been in the negative direction, showing improvement as the offensive line gels and as our younger players get better.

Finally, let us look at scoring offense.  Note that we are back to positive being a good thing.

Generally, USC has been in the positive in terms of scoring offense.  The only games we were below what the opposing defense averages is Stanford and Cal.  Other than that, USC has done particularly well and averages 7.5 more points than the opponent typically allows.  This trend is also increasing over the course of the season.

Defensive Comparative Stats

Now let us look at the defensive side of the ball, starting with total defense.  Keep in mind that negative numbers are better in terms of defensive yards allowed.

As you can see here, USC has been fairly average on defense.  USC averages allowing only 6.8 yards below what the opposing offense typically allows.  The general trend shows the defense improving, but not by much.  There have been games which USC has done better than average (Cal, ASU, and UCLA), while there were others which have been worse (Colorado, Arizona, Oregon).

Now let us look at rushing defense.

The rushing defense has been again very average.  The general trendline for this is positive, meaning our defense has gotten worse at defending the run.  Overall, USC allows 15.5 yards fewer than the opponents typically gain.  This is pretty close to average.  In terms of yards per carry, USC has held opponents to 0.04 yards below their average yards per carry, which is also pretty much average.

Now let us look at passing defense.

The passing defense started this season below average, but has had a general trend that shows it improving as the season goes on.  The ASU and UCLA game were both significantly in the negative margin.  Overall for the season, USC has allowed 9.9 more yards passing than the opposing offense typically gains, so they have been overall about average again.  The last couple games, however, give some hope.  If you look at the yards per pass attempt, it is about average again with USC allowing 0.12 more pass yards per attempt than the opponents typically average.

Now let us look at completion percentage.

USC allows a very slightly higher completion percentage than opponents typically average.  Overall, we’re looking at a 3.3% higher completion rate than opponents’ average.  The margin has been up and down all season and has been fairly inconsistent.  The general trendline shows the completion percentage defense getting slightly worse as the season goes on.

Now let us look at 3rd down defense.

The third down defense has been fairly average yet again.  USC typically allows a 3.0% higher 3rd down conversion rate than opponents typically average.  However, the general trend shows that the defense is improving in this regard.

Now let us look at turnovers gained.  Keep in mind that turnovers gained is something that positive is a better number for our defense.

USC is above average in forcing turnovers.  USC averages gaining 0.55 more turnovers than the opponent typically gives up.  However, the general trend shows that this is slowly getting worse.  Also note that a good number of our losses has the turnovers gained as negative: Arizona, Oregon, and UCLA were all below average.  Turnovers themselves are such a high variance play as you will typically only get a couple per game at best.  When combined with high variance on when the offense loses turnovers, it becomes a bit of dice roll with this USC team for turnover margin.  USC gains a lot of turnovers from the defense, but it also gives up a lot too.  This means that if the dice happen to fall the wrong way, as they have a couple times this season where we lost the turnover margin, we end up losing those games.  Stanford was a -1 turnover margin, Arizona was -4, Oregon was -2, and UCLA was -2.  These key turnovers dug USC into too deep a hole to recover from.

Now let us look at sacks gained.  Again, positive is better for this category.

USC has been above average in terms of gaining sacks.  On average, USC gets 1.0 more sacks than the opposing team typically allows.  There have been a couple games that we have been limited in the number of sacks we have gained: Stanford, Utah, and Arizona all held us to one sack below what they typically give up.  However, the general trend is significantly positive.  Applying pressure is one of the few bright spots in this USC defense.

Now finally we take a look at scoring defense.  With this stat, having a negative margin is better.

Overall USC has been fairly good at scoring defense.  Not surprising since Monte likes to play a “bend don’t break” scheme.  USC holds opponents to 4.2 points below their average.  This isn’t a great scoring defense, but it is a significant margin.  The trend shows this defense getting slightly worse as the season goes on.  However, Oregon is really the only team to significantly score more than their average.


Overall, the USC offense is definitely above average in most regards.  It consistently produces more yards and points.  The passing game has a jump on the running game.  However, the main issue that this offense has compared to previous seasons is the turnover rate.  USC has been much more turnover prone this season and the high variance nature of those turnovers has caused plenty of issues which lead to losses.  I was surprised to see that sacks allowed has been slightly above average.  I have felt that Barkley has been pressured so much more this season than last, which caused it to seem probably worse than it was in my head.  Overall, the line is still protecting Barkley decently well (as you might expect with 4 out of 5 starters returning), but they still could improve.  3rd down conversions must also continue to improve, but I suspect that penalties have much to do with that.

Defensively, we are a very average team.  I think that losing Isiah Wiley has hurt this team much more than we would like.  The defense has been average in almost all categories for comparative stats.  However, there are bright spots in turnovers gained, sacks, and scoring defense.  We, as fans, are definitely missing the well above average defenses of Pete Carroll, though.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s