Advanced stats have been a growing part of looking at football and we even saw it this past weekend when people started arguing about the probabilities of the Falcons converting on the 4th and short, contrary to common football opinion that they should have punted. A lot of this can be due to the rise of advanced statistics in baseball and a passion of quantifying what people see on the field. Unlike baseball, football is a sport where you can’t really break down an individual’s value or their impact on a game. Baseball is easily quantifiable and there’s actually a statistic that states how many wins a player is worth. Moneyball is one my favorite books and I have taken great interest in football statistics and the perspective it creates for fans. While advanced stats can’t necessarily be completely trusted in a sport like football at face value they can be used to show trends in teams’ successes, failures, and what their weaknesses are. In a way they reaffirm what we are seeing on the field or make us question what is actually happening on the field by eliminating confounding variables. This is my first post as an official contributor to ATQ, you can find more of my writing here, but I'm looking forward to sharing this passion on a new platform. After the jump I’ll look at what the numbers say about the USC – Oregon matchup.
Let’s first address the FEI stat as it has caused a lot of arguments amongst people because at one point according to FEI Washington had a more effective offense than Oregon according to FEI. FEI looks at the possession efficiency of each team when compared against a baseline number that is made from looking at every possession in college football. I’m oversimplifying this but think about FEI as a value based system on a per drive basis, while S&P+, which is friendlier to Oregon measures on a per play basis. Oregon is a really hit or miss offense. Often times Oregon will score in 8 plays or they won’t get a first down. The Ducks rarely stall after they get a first down.
USC ranks 12th in offensive FEI while Oregon ranks 16th, with scores of .434 and .409 respectively. There are a few main stats that separate the Trojans from the Ducks in the FEI department. First, Oregon gets a first down or a touchdown in only 74% of their drives when compared to USC, which gets at least a first down or touchdown 81% of the time. Second, the biggest gap occurs in the methodical drive category, where Oregon has a drive of ten plays or more only 10.4% of the time while USC has drives of ten or more plays 21% of the time. This is fine with me as we tend to hold the view that we don’t care how we score as long as we score. My guess as to why methodical drives are important is because it keeps the ball out of the hands of the other offense as FEI looks at the efficiency to which a team can score points but also how efficient they are at stopping opponents from scoring. The OSOS Pvs and OSOS Fut also kill the Ducks. These two stats measure the probably of an elite offense would have an above average score against the competition the Ducks have faced in the previous weeks and the future weeks. Basically it adjusts for strength of schedule taking in to account games that haven’t been played yet.
Oregon’s defense is actually a huge strength when looked at from an FEI perspective. They rank ninth in the country with a score of -.576 and USC is ranked 37th with a score of -.202. The Trojan defense is good at stopping explosion plays (ranked 28th) but not at stopping methodical drives (55th). The USC defense though gives up first downs or touchdowns 67.3% of the time. When matched up with an offense like Oregon a defense really can’t afford to let Oregon’s offense start snowballing down the field and pick up momentum. This is going to be a key battle in the game this weekend. Oregon’s defense hasn’t benefited from the same number of turnovers as they did last year, but they have been effective. They rank 80th in the country in methodical drives, showing that they have troubles getting off the field. This could be attributed to the bend and don’t break defense that some fans believe we’ve adopted. On the bright side Oregon’s speed is putting opposing offenses in to situations where they have to run a lot of plays to get a score. This doesn’t necessarily match up well with the Trojans as USC is good at converting methodical drives in to points.
S&P+ is a much nicer statistic to our Ducks as it measures effectiveness on a per play basis. Oregon is very successful on the success metrics that Football Outsiders uses (getting 50% of the yards you need to each down, i.e. getting 5 yards on first, 3 yards on second, and converting on third or fourth down). After beating Stanford last week Oregon jumped Boise State to the third in the country with a score of 277.7. USC is sitting pretty at 24th with a score of 228.6.
Oregon’s offense is second best in the country after Wisconsin in terms of S&P+ and USC is 20th. Oregon excels in all facets of the offense. They rank 3rd in rushing, 5th in passing, and are the 2nd best team in the country on standard downs (1st and 10, 2nd and 5, 3rd and 2). Oregon is very balanced on offense and if they aren’t forced in to having to pass they are one of the most dangerous teams to defend. When in passing downs the Ducks are above average but not fantastic with a ranking of 27th. Duck fans know that the receivers have been struggling this season to get open, and when the defense knows that the Ducks are going to pass they can drop more players in to coverage. This has led to Oregon running the ball in passing situations 49.3% of the time, either to try and convert a makeable fourth down or because they think their running game can get a bigger chunk of yards than passing.
USC ranks 37th in defensive S&P+ and ranks 48th in stopping the run, 27th in passing, 48th in defending standard downs, and 34th in defending passing downs. All of these bode well for Oregon. These teams are a fair amount apart from each other when Oregon is on offense and USC is on defense. The best part is that USC is not fantastic at stopping the pass when the Trojans know the other team is most likely going to pass. They are close to the middle of the country in defending the run, so it may be Oregon won’t even be in that many passing situations.
Oregon’s defense ranks 7th in S&P+ defense, ranking 27th in stopping the run. The defense has played fantastic the last two weeks in stopping the run so this may not be a huge problem but it is still worrisome if USC is able to run the ball. USC is above average at running, they’ve been great the last couple weeks, and if Oregon can’t stop the run to focus on the pass they could be in trouble. While some (including me) have stressed about the cornerback play and how much separation receivers get, the S&P+ numbers show that the Ducks are actually the fifth best team in defending passing plays. On standard downs Oregon ranks 5th and on passing downs they rank 16th. This may be contrary to we think of our secondary but it is a positive sign against a great Trojan passing attack. I’m not sure how much I trust the pass defense numbers because it has been frustrating to see our team get passed on so much. I wouldn’t be comfortable with this pass defense against the Trojans even though the stats say that the two units are much closer together than we think (see, don’t trust these all at face value).
USC comes in at 20th in the country in S&P+ on offense, ranking 30th in running, 25th in passing, 23rd on standard downs, and 38th in passing downs. Oregon does well in passing downs so if Oregon can hold the Trojan rushing attack to little production then Oregon can get off the field and lower their methodical FEI stat. Running back Marc Tyler has fully rehabilitated from a shoulder injury and the Trojan line has been opening up bigger holes in their last few weeks, allowing them to keep close with Stanford, and easily beat Washington and Colorado. I’m not sure these stats tell the whole story of the USC offense as USC is peaking right now and they struggled early on at the beginning of the season. There are maybe a few too many intangibles to have the game be accurately explained by these stats.
If you combine FEI and S&P+ you get a statistic called F/+, which is the best predictor of who wins. FEI and S&P+ both show strengths and weaknesses of teams but F/+ decides which strengths match up the best and outweigh weaknesses. Oregon is ranked as the fourth best team in the country according to F/+ rankings, a statistic that has around an 84 R^2. The 84 means that the F/+ model is a great predictor who is going to win and can explain around 84% of the games. Anything above 84 in an R^2 score means that the model may be created too specific to explain events in the past and therefore not a great predictor of future events. Oregon has a F/+ rating of +26.8%. USC comes in ranked 15th with a score of +14.5. For comparison, LSU ranks first in the country with +36.2% and Alabama is second with a score of 35.2%. Bill Connelly of Football Outsiders and Football Study hall used these F/+ rankings to predict an Oregon win by 14.7 points. I’m not sure how they get the predicted number because they don’t equal the difference between the F/+ scores, but it still creates around a two touchdown win for Oregon.
Keys to the game after looking at these stats (Basically the SparkNotes of the above information)
1. Oregon needs to be able to bottle the Trojan rushing attack.
While "stopping the run is key" may seem really obvious to anyone that knows football it is especially true in this game. USC has been surging on the ground recently and has enabled them to play up to the levels of Stanford on offense. I don’t think the Ducks can stop the Trojan passing attack, they can contain it though by putting the Trojans in situations where they need to pass, a facet of the game Oregon excels at defending. By putting the Trojans in passing downs Oregon can send crazy blitzes and either drop in to coverage or the send the house to try and confuse Matt Barkley.
2. Oregon needs to get turnovers on defense.
The Ducks have a low ranking in defending methodical drives and USC excels in running methodical drives. The Trojans can keep this game close by holding on to the ball and shorten the game, limiting the number of possessions Oregon gets, therefore reducing the opportunities Oregon has to score. Oregon had one of their most dominant performances in years last week when they beat Stanford, and a big part of the win was getting turnovers and shortening the field that offense need to drive. Turnovers cause the biggest swings in games, and if Oregon can get out to a quick lead then USC is going to have to pass and play quick offense to catch up, which hasn’t worked well for anybody this season when playing the Ducks.
3. Get successes running the football
Oregon is a strong team on offense, and the only weakness they have is when they get put in to situations where they have to throw the football and the other team knows they are going to throw the football. USC has a talented defensive front seven but they allow first downs and touchdowns on 67.3% of all possessions, and once Oregon gets in to a rhythm and starts moving the ball they are hard to stop. A recipe for USC to beat Oregon is to gang up on the run, get Oregon in to bad passing downs, and then to play the pass against Oregon’s receivers.
All stats used in this article can be found here and a big thanks goes to Bill C. at Football Study Hall.