| 2006 Defense
Histograms & Top 10 Big Run Pass Plays I had some time on my hands over the Memorial Day weekend and wanted to look at some more of the raw data from the 2006 season and the Trojan defense. Here is some of the data organized into different categories comparing the first half to the second half. In general the numbers were pretty consistent although there were some areas of improvement.
The average number of sacks per game picked up nicely the last half of the season as well as the rush defense. Part of that however is just the selection of the mid-point for comparison as there were 6 sacks in game seven against Oregon State. Moving that over one game makes the comparison 2.2 versus 3.3 which is still a nice improvement of a sack per game. USC increased the frequency of blitzing in the second half of the season quite considerably. As a side effect of this however, pass defense looked worse in terms of average yards surrendered. Also keep in mind though the Trojans also faced stronger passing attacks the second half of the season according to the NCAA. Overall however most other numbers that I checked in general were fairly consistent. I was also curious how the season would look if I put the defensive yards allowed data (both run and pass) into a histogram and made some high level comparisons. At the high level this first chart below is what the histogram would look like for the entire season for USC on defense. In sum total it tells you that USC did a good job of holding the offense to gains of 3 yards or less on most plays (4+27+293+166 = 490 plays). (Note: The reason the -5 yards to 0 column spikes so high is because it includes all the incomplete passes in this category). ![]() Conversely this also means that 342 times (109+99+61+26+25+22) during the season the defense gave up more than three yards per play. In the extreme cases there were 47 instances (25+22) where USC gave up more than 21 yards on either a pass or run play. I don't have the ability to easily match up the database of plays to the actual video of the files but the USC coaching staff does I can assure you. I have to do it manually which is time intensive and cumbersome to say the least. At the end of this post I will list the top 10 run and pass plays yardage wise against USC in 2006 in case you are curious. What coaching staffs do during this time of the year in addition to recruiting is to look at video cut ups of certain plays and situations for a more detailed review. They'll spend a fair amount of time studying what did work and why. Also for improvement they will take a hard look at what did not work (i.e. gave up big yards) and why as well. In problem solving this is just simple application of the Pareto Principle or 80/20 rule and focusing on what matters the most. During the first six games of the year the USC defense allowed an average of 282.3 yards per game. In the last seven they allowed 307.4 yards per game. Sometimes "averages" distort things however and it is more useful to look at the events in the form of a histogram. Here is the season broken down in half just using the mid-point of the data somewhere in the Oregon State game in order to make things roughly even in terms of plays (Disclaimer: The data is off by a couple of plays and the database is missing a half dozen plays due to TV recording errors etc. In general it matches the data the USC athletic department submits to the NCAA.) ![]() In this case the averages comparison is not distorting things too much. The shape of the histogram is not all that different for the first half versus the second half of the season. The defense produced 16 plays of -5 or greater yards in the first half and 15 in the second half. There were 142 plays of zero to -4 yards allowed in the first half and 151 such plays in the second half. On the other end of the spectrum there were 40 plays of greater than 16 yards allowed in the first half and 33 allowed in the second half. All in all it is pretty similar which is what I expected at the aggregate level. I was curious to see if this held up as well once sorted out in terms of run versus pass. Of course if can be argued that USC faced a tougher schedule during the second half of the year (Oregon, Cal, ND, Michigan, etc.) over the first half (Arkansas, Nebraska, Arizona State, Washington State, etc.). Jeff Sagarin's system puts the first half opponent strength at an average of 35.8 and the second half at 32.1 for comparison. Looking at the NCAA statistical rankings it is fair to point out that the Trojans faced higher ranked passing offenses in the latter part of the 2006 season as well. This next chart below shows just the rush defense overall in 2006 in terms of yards allowed. Since the NCAA stats include sacks as run plays that data is included here as well as tackles for losses. In total the USC defense produced 109 plays that stopped the offense from gaining positive yardage. This is an impressive number but in comparison the 2002-2004 defenses all produced a greater number of negative yardage plays (e.g. sacks + tackles for losses combined were 139 in 2002, 181 in 2003, and 170 in 2004) The rush defense also allowed 31 plays of 11 yards or greater. The longest examples were the 59 yard scamper by Brady Quinn of Notre Dame and the 46 yard reverse by Marlon Wood of Washington. ![]() I broke this up by half as well to see if there was any dramatic improvement between the first half of the season and the second half. The following chart depicts that comparison. ![]() As you can see the general shape of the curve is the same in terms of yards allowed. On average the run defense was stouter during the second half though allowing only 2.6 yards per carry. In the first half it was 3.7 yards per carry. Some of that difference might still be due to sack yardage and not necessarily true run defense. With more sacks in the second half of the season the net run yards went down due to the way the NCAA keeps its statistics. Regardless I still think the run defense was better in the second half of the season. The histogram shows in greater detail some of the specifics lurking behind those numbers. For example in the first half there were 50 plays of negative run yardage (including sacks) produced. In the second half there were 59 such plays. In the first half there were 7 run plays of +16 yards allowed. In the second half there were 7 as well. Still I believe however the run defense did improve in the second half of the season due to better gap discipline and pursuit by the defense. USC's run defense versus Michigan was the best example of that type of play all year long and hopefully a sign of things to come in 2007. Here are a similar set of charts that just look at the pass yardage allowed for the 2006 season. The number spikes as mentioned on the -4 to 0 yardage plays since this is where all the incomplete pass plays (0 yards) show up in the data. ![]() Overall USC allowed 204.8 yards per game passing over the course of 2006. The average completion allowed was 10.9 yards, however the histogram above gives a better picture of that distribution. Here is what happens below if you break it up by first half of the season versus second half. ![]() Here you can see that again the shape of the curves are remarkably similar in terms of pass defense. The average passing yards allowed in the first half versus the second half is 182.2 versus 224.1 for comparison. Above you can see the distribution behind that average. On the right hand side of the chart there were 50 pass plays of greater than 11 yards allowed in the first half of the season. In the second half that number was about the same at 53. The data is interesting to look at but what needs to be done next is to study the actual plays that resulted in either big gains or repeated yardage gainers for the opponent and figure out the "why".. Some fans I listed too often like to blame coaches and players or the "Who" aspect of the game when things go wrong. That has always just struck me as just a convenient form of scapegoating and venting of our emotions. Coaches are still stuck in the end with figuring out the "why something happened" and the coaching of "how to fix it" part of the equation. In this respect they are like teachers and problem solvers. Figuring out the why and how component leads to better instruction in technique for the players and the transfer of both awareness and skill in player development. In some cases of course it does mean starting one player over another or using players differently in certain situations like 3rd and short run plays versus 3rd down and long pass plays. We will see in 2007 if the Trojans can move the defense to the next level and rank among the 2003 and 2004 squads of the Pete Carroll era. For those interested here is a list of the top ten big plays given up by USC in the rushing and passing category. Send your insights to Heritage Hall! List of the top 10 run plays surrendered by USC in 2006.
List of the top 10 pass plays surrendered by USC in 2006.
|