Occasionally when I am reading some material pertaining to football I'll come across a quote that advocates a particular school of thought for offense or defense. A good example of such a comment was made by Tom Flores former Oakland Raider head coach in his excellent book Coaching Football. In the text of the new 2nd edition on page 15 in the chapter entitled "Theories of Winning" he and co-author Bob O'Connor note that "In the college game, the top 10 running teams in the country generally average 80 percent wins, but the top ten passing teams generally average only 50 percent wins". It is an interesting statement that I have seen made by other coaches as well over the years such as Lou Holtz and former USC coach John Robinson. I have no doubt that the statement was probably true at one point in time during the 1970's or 1980's for example.
Conversely if you read internet chat boards where passionate fans hang out and discuss their team the exact opposite view point is often held. In other words the key to success is to air the ball out and throw it more down field. Running the ball is viewed as a necessary evil it almost seems by some fans. In USC's case in 2006 some fans got downright upset when the the Trojans kept trying to run the ball late in the season with limited success. It later turned out that it was Head Coach Pete Carroll who did not want to abandon the running game although the offensive coordinator as usual became the scapegoat in the media and public. Pete later commented that he was concerned about losing the "balance" in the offense by abandoning the running game.
So which side is right? Do you win more games on average with a strong running attack or a strong passing attack? And how have the BCS national champions in this decade actually performed in this regard? A quick check of the data at the NCAA statistical web site side yields some surprising insights. For the sake of simple comparison let's start by looking at the average winning percentage of the Top 20, Top 10, and Top 5 teams in both passing and rushing yardage. Then for fun let's look at where the teams in the BCS bowl game this past decade have actually ranked as well for comparison.
Let's start with the Top 10 Rushing Teams in the NCAA since 2001. Here is a trend chart that shows the average winning percentage by year.
Despite the Tom Flores comments above (at least for the time period specified) there is no indication that the top rushing teams in the NCAA average anywhere near 80% wins per season. As I mentioned previously I suspect this may have been true at one point in time during the 1970's or 1980's during the wishbone era for example but it has long since ceased to be an accurate statement. In general the top 20 rushing teams have averaged 62% wins, the top 10 teams have average 63.2% wins an the top five rushing teams only slightly better than a 63.8% winning percentage. The leading rushing team in the country since 2002 has been the United States Naval Academy under Head Coach Paul Johnson. Their record is a combined 37 Wins and 25 Losses for a 59.6% winning margin during this period.
So are the "air it out" fans right? Should teams be throwing more often and not worrying about the rush attack as much? Well as ESPN's Lee Corso often emphatically puts it, "Not so fast my friends!". Here is a similar chart showing the average winning percentage of the Top 20, Top 10, and Top 5 passing teams in NCAA Division I since the 2001 season.
As you can see the basic trend is fairly similar for the Top Passing teams in the country. The Top 20 passing teams average a winning percentage of just under 61% or slightly worse than the top rushing teams. Distilling the group down however to the Top 10 improves the winning percentage to 65.6% and taking it down to the Top 5 improves it to 66.1%. So at the broader level the numbers are roughly the same although the top passing teams in the narrower (i.e. Top 5) sense have won slightly more than the top rushing teams. The difference is pretty small however.
The poster children for airing it out these days are the University of Hawaii Warriors under June Junes and the Texas Tech Red Raiders under Mike Leach. Annually these two teams duke it out for the most pass happy squad in the nation and finish either first or second in the NCAA passing statistics. How have they performed in terms of winning? The Warriors come it at a 65.8% winning margin and the Texas Tech squad at 65.3%. This is better than in the recent past for both programs and is no doubt very exciting for the fans to watch. Texas Tech rallied from 31 points down for example in the 3rd quarter of the 2006 Insight Bowl to defeat the Minnesota Gophers and help cost coach Glenn Mason his job. An exciting offense such as this style also gets lots of highlights on Sports Center every week (but you don't see Navy very much eh?). But in reality the high octane passing game alone does not appear to generate a higher winning margin even among the best practitioners of this tactic. (Note: It still may be the right tactic for certain teams to employ...without such an offense these teams of course might have an even lower winning percentage!)
So what does work? I have opinions on the matter but it seems there is no shortage of those on this particular topic. Let's instead just examine what the top teams in the NCAA have done in terms of rush versus pass ranking. For starters let's look at the teams that have been in the BCS championship game since the year 2000. Here is a table of those teams and their respective performance in the aforementioned categories.
Interestingly as you can see there is some variation as always in the data. Eight times a team with a higher ranked pass attack has made it to the final game. Seven times a team with a higher ranked rush attack made it in. A #1 ranked passing team (Florida State in 2000) and a #1 ranked rushing team (Nebraska in 2001) both made it to the BCS title game. Interestingly both lost. Very few teams have been in the top 10 at all in either the rushing or passing statistics and made it to the title game. The big exception other than the two teams noted above was in 2005 when the Texas Longhorns and the USC Trojans met for the BCS title. Painfully for Trojan fans the #2 ranked rushing team beat the Trojans (#6 ranked rushing attack and #5 ranked passing attack). Of course other factors not discussed in this post more strongly correlate to winning (defense, turnovers, +4 plays on first down, red zone efficiency, and big plays, etc.)
Oddly on average the teams that made it to the BCS title game in the past decade have only ranked 31st in rushing and 32nd in passing. Here are histograms of the data for the categories as well if you prefer to see the variation rather than averages.
The most common performance characteristic for the BCS teams is to rank about 20th to 30th in passing and generate over 260 yards passing. The two teams that ranked the highest in passing yardage (Florida State in 2000, and Oklahoma in 2001) however both lost.
The most common performance characteristic in terms of rushing yardage is to rank about between 30th to 40th in the nation and generate between 175 to 200 yards per game on the ground. The highest yardage rushing team was Nebraska who lost in 2001 to Miami followed by Texas who defeated USC in 2005.
Here is just the table of winners in case you are curious how that spread looks.
The numbers don't change much in reality for the winners. The average rush rank is 34th in the nation and 185 yards. The average pass rank is 34th and 248 yards per game. The total average offense is 185.4 + 248.4 = 433.8 yards per game for the winners. (Note: In 2006 offenses were down about 10% or 40 yards per game due to changes in clock rules that shorted games. Florida only averaged 397 yards per game for the season or roughly about what USC averaged).
The balanced offensive nature of the champions is no coincidence in my opinion. Almost every top coach in NCAA mentions this at some point if you listen closely. Regarding the USC offense Pete Carroll regularly states, "We are a balanced pro style offensive employing multiple formations, motion, and personnel groupings". Urban Meyer at Florida states a similar opinion in a different way (and with a different style of play). "We want to force the defense to defend the whole field," says Meyer. "All things being equal, anybody can stop the run and anybody can stop the pass. But can you stop both?" In his 2004 AFCA speech Meyer also notes, "People think we are some type of star wars type offense. Nothing could be further from the truth. We spread the field with multiple WR's in order to make it easier to run the football. We strive for an overall balanced capability in terms of attack." Note that this does not mean any team will always run throw the same amount or rack up the same amount yards in any given game...Coaches like Carroll and Meyer want this inherent capability so defenses can't cheat one way or the other on any given play.
Why does the balance help? As Meyer notes a good defense can easily take away either the run or the pass but it can rarely take away both. Want to stop the run? Simple - put 8 in the box to stop the ground attack. Want to stop the pass? For starters always put 5 men (or more) back in pass protection to stop the pass. Applying pressure on the QB helps perhaps even more. The trouble is to do all of this at the same time means that about 13 players need to be people on the field for the defense. Since that can not be done legally done teams have to use their 11 players in the best position possible each down and distance depending upon the situation and the opponents tendency. Good offenses in turn get around this defensive adjustment by being able to pass or run equally well (i.e. balanced) often changing the play call at the line of scrimmage with an audible by the QB.
So what does this mean for the 2007 USC offense? A quick look at last years rush and pass statistics in comparison to that of recent BCS title game winners gives some insight.
As you can see the USC rush attack paled in comparison to the average of the previous champions to the tune of almost 60 yards per game on average. Conversely the passing attack was 15 yards superior on average. I have no doubt that Pete Carroll and staff know this in incredible detail beyond the simple mathematical averages expressed here. Yards per carry by situation (e.g. 2nd and 7 run plays) as well as play against different defenses (e.g. I Formation runs versus 4-3 Over defensive fronts) are all things the team tracks. Interestingly USC started the 2006 season quite well in terms of balance averaging 178 yards rushing and 239 yards passing after the first three games. Then the 3rd fullback (Freshman Stanley Havili) was injured following the loss of Brandon Hancock and Ryan Powdrell. After that point the run game trended downwards game after game until the final two games of the season where only 50-60 yards net rushing could be mustered versus UCLA and Michigan's respective defenses. The passing game however improved nicely.
So one key to improving the Trojan offense in my opinion is a more effective run game in 2006. The return of Jeff Byers and insertion of Matt Spanos at center will probably have some impact. Another year of All American stalwart play by Sam Baker at left tackle as well as a more experienced Chilo Rachal and Drew Radovich in the line up on the right side will benefit also. The rest of course depends upon defense, special teams, turnovers, and other key parts of the game. My guess is still that most fans and the media will put more emphasis on the quarterback and WR play this upcoming season (and it will be vital as well for an undefeated season to occur) since it is more visible. And lets face it that is what makes the sports highlights on ESPN. I have little doubt however than USC can average 250 yards per game (or more) passing in 2007. I don't even think that will be much of a challenge in reality given the returning quarterback experience factor and skills of the players on the field. The bigger question in my mind is can the squad average something closer to 185 yards per game rushing? If the Trojans can perform closer to this level for the season that will be very good news indeed if previous history is any indicator of success. Somehow I bet Pete Carroll knows that equation all too well.