Cardinal Numbers: Hitting the Wall?

For those of us obsessed with the collection, analysis, interpretation and presentation of data, Terry Johnson attempts to put the Cardinal's current offensive malaise in context. Read on for more, including some credible data suggesting the Stanford defense has made remarkable improvement despite the distressing presence of familiar results on the scoreboard.

Coming into this season, Stanford desperately needed a major improvement in its offense. Restoring the offense to legitimate Pac-10 levels was one of the most important challenges for new coach Jim Harbaugh.

Through the first three games, Harbaugh's work with the offense got generally good reviews. Stanford was gaining 435 yards per game and scoring 28 points per game after three games. The look and feel of the offense was better. But on Saturday, after three decent games, the Stanford offense took a big step backwards. Against Arizona State, Stanford managed only 235 yards of total offense and scored just three points.

What happened? Has the offense hit the wall?

We've seen that happen before. In each of our two worst seasons in recent years, Stanford's offense managed to stay more or less afloat for the first three games. But in each of those seasons, the offense hit the wall in the fourth game and never recovered.

It happened in Buddy Teevens' first season, 2002, when Stanford went 2-9. Stanford's offense scored 24 or more points in each of the first three games and went over 400 yards in two of the first three games. Stanford didn't reach those levels again for the rest of the year:

2002 Offense
 First Three
Total Offense/Game430267
Passing Yards/Game243154
Rushing Yards/Game187113
First Downs/Game2017

It happened again last year. In each of the first three games, Stanford had over 340 yards and over 20 first downs. Again, the offense hit the wall in the fourth game. Stanford didn't come close to those levels for the rest of the year:

2006 Offense
 First Three
Total Offense/Game362189
Passing Yards/Game242142
Rushing Yards/Game12047
First Downs/Game2211

So, twice in the recent past, the offense in the first three games was significantly better than it was for the rest of the season. Those offenses did not find their natural, abysmal level until hitting the wall in the fourth game of the season. How did that happen? Why would an ineffective offense be able to muddle along for three games before the bottom dropped out? There are several factors that perhaps could be at work.

First, three games are enough for other teams to scout an offense and figure out what it's doing. It gets harder to fool people after they've seen some film.

Second, injuries start to crop up after a few games, exposing areas in which there is a lack of depth. It's probably not a coincidence that we had significant injuries on offense in both 2002 and 2006.

Third, the early-season statistics benefit from playing our weakest opponent, San Jose State, early in the year – though that explains just one game, and is not a full explanation for the drop-off in performance we've seen starting with the fourth game.

Finally, our coaches (Teevens in 2002 and Harris in 2006) may have implemented their best ideas early in the season. Once they had to adapt and try something new, perhaps they weren't able to adjust, add new wrinkles, and so forth.

Last Saturday's game was the fourth game of the season, and the offensive performance dropped off drastically. Was that just a bump in the road, or has our offense once again hit that fourth-game wall? We do have some of the same factors this year that affected us in 2002 and 2006 – the opportunity for opponents to scout our offense, some key injuries, a big early-season win over San Jose State that may have made our offense look better than it is. But the real question is whether Harbaugh and his staff have the coaching ability to adapt and succeed, rather than sink into the abyss as Teevens and Harris did. I think so.

There is some real reason to hope that Harbaugh has the coaching ability to have the team improve as the season progresses, rather than hit the wall and collapse. In his first season at the University of San Diego, 2004, Harbaugh started out slowly, with a 1-3 record. In those first four games, his team scored 22 points per game and gained 340 yards per game. But Harbaugh's team really hit its stride in midseason. San Diego went 6-1 in the last seven games of the season. The team scored 44 points per game and gained 485 yards per game over those seven games. Of course, there are numerous reasons that the situation at San Diego isn't analogous to our situation, but at least we know that Harbaugh's offense didn't hit the wall in San Diego, for what that's worth.

We probably won't be able fully to judge our offensive performance until we get past the USC game and into the following stretch of games, when we will face six straight opponents that currently are unranked. That should give Harbaugh an opportunity to show what he can do with our offense.

At least one of the players thinks the Arizona State game was a bump in the road, not a stone wall. Team captain T.C. Ostrander , who unfortunately won't be available against USC after suffering a seizure, commented after last Saturday's game that the Stanford-ASU score was frustrating: "People are going to look at that and go 'Oh, there's Stanford again.' We're not like that, we're better than that." I think he's absolutely right. I really believe the offense (and T.C) will bounce back in the second half of the season.

Pass Protection

Last season, Stanford ranked last in the nation in sacks allowed – 119th out of 119 teams. Stanford allowed 50 sacks, or 4.2 sacks per game. So far this season, Stanford is showing little statistical improvement. Stanford ranks 112th in sacks allowed, having allowed 15 sacks in four games, or 3.8 sacks per game. Stanford is on a pace to allow 45 sacks. So, from the stats, it would appear that Stanford's pass protection woes from last season have continued.

However, in all fairness, I believe this is a misleading statistic. From what I see on the field, it appears to me that Stanford's pass protection has been considerably better this year. It seems to me that our quarterback has had more time to find a receiver. But why doesn't the apparent improvement in pass protection show up in the statistics? Three possible explanations occur to me.

First, Stanford is throwing the ball more this year. Last year, Stanford averaged 26.3 pass attempts per game. This year, Stanford is averaging 42.0 pass attempts per game. Stanford's pass attempts therefore have increased by about 60%, which increases the risk of a sack accordingly. Looking only at sacks per game ignores this difference.

To get a better idea of how Stanford is doing on pass protection, rather than looking at the number of sacks per game, we might want to look at the frequency of sacks relative to the number of pass attempts. When we run the numbers that way, the picture is decidedly different. Stanford allowed one sack for every 6.3 pass attempts last year. This year, Stanford is allowing just one sack for every 11.2 pass attempts. Those numbers tend to confirm what we've seen on the field: there has been clear improvement in the pass protection.

Second, the number of sacks doesn't depend entirely on the pass protection. The quarterback also has some responsibility for sacks. If the quarterback has a reasonable amount of time, but fails to get rid of the ball, the sack is his responsibility. Given enough time, he must find a receiver, scramble, or throw the ball away. Some quarterbacks are better than others at getting the ball away. I think our starting quarterback is responsible for some of the sacks this season because he didn't get the ball out of his hands quickly enough.

It's difficult to measure statistically the extent to which a quarterback is responsible for sacks. A quarterback's sack totals depend on his team's pass protection, scheme, and coaching, as well as the quarterback's own abilities. In Ostrander's case, the only reasonable point of comparison is Trent Edwards, the current rookie starter for the NFL's Buffalo Bills. Over the last three seasons, Ostrander and Edwards faced basically the same conditions. Each of them got substantial playing time, with the same pass protection, scheme, and coaching.

The statistics seem to indicate that Ostrander is more prone to sacks than Edwards was. Over the last three seasons (2004-2006), Ostrander took one sack for every 5.9 pass attempts, while Edwards took one sack for every 9.8 pass attempts. Ostrander therefore was sacked significantly more frequently than Edwards during that three-year period, although the difference between them lessened over time.

The statistics also suggest that Edwards had better ability to escape the rush and make positive yardage. In the last three seasons, Edwards had positive rushing yardage each year, with a total of 222 rushing yards. Ostrander had negative rushing yardage each of the last three years, with a total of negative 340 rushing yards. He has negative rushing yardage again this year. Edwards' lower sack frequency probably was related to his greater rushing yardage, because he probably was better at turning potential sacks into gains. This tends to suggest that some of the sacks this season may have been a result of Ostrander's inability to avoid sacks, rather than being solely due to the pass protection. Fortunately, this can be corrected. Ostrander is a smart kid who can learn to avoid sacks and he will.

Third, Stanford's "sacks allowed" statistics took a turn for the worse in the Arizona State game. That single game accounted for seven of Stanford's 15 sacks allowed this season, which had a big impact on Stanford's overall sacks allowed per game. There's some reason to think the Arizona State game may have been a little bit of an anomaly due to Stanford's injury situation. Chris Marinelli played injured, and Ben Muth replaced Allen Smith. Marinelli should get healthier, and the offensive line should become more effective as they become accustomed to the change in personnel.

Therefore, I believe Stanford's pass protection has improved this season, even if it doesn't show up in Stanford's sacks allowed per game.

The "Attack Defense"

To evaluate the success of Stanford's "attack 4-3 defense," I've been looking at turnovers created, sacks, and tackles for loss, in addition to the normal defensive statistics. At Western Michigan last season, Scott Shafer ran this scheme with considerable success. Last season, Western Michigan had 46 sacks (3.5 per game), which ranked 1st in the nation; made 6.9 tackles for loss per game, which ranked 22nd; and created 34 turnovers, which ranked 4th. Western Michigan was 11th in total defense, giving up 275 yards per game. That's the kind of performance you would want from an "attack defense."

Obviously, Stanford isn't there yet. But perhaps we can see some glimmer of what Shafer is trying to create:

Last year, Stanford had 1.2 sacks per game, which was last in the conference and 111th in the nation. So far this year, Stanford has 3.0 sacks per game, which is 3rd in the conference and 20th in the nation. Big improvement so far.

Last year, Stanford made 3.8 tackles for loss per game, which was last in the conference and 115th in the nation. So far this year, Stanford is making 8.25 tackles for loss per game, which is 2nd in the conference and 14th in the nation. Again, big improvement so far.

The area in which we haven't seen any improvement is turnovers. Last year, Stanford forced 15 turnovers (1.25 per game), which was 9th in the conference and 109th in the nation. This year, Stanford has forced just 5 turnovers (1.2 per game), which is last in the conference and 103rd in the nation. Stanford needs to do a better job of creating turnovers.

Random Numbers

Stanford's first four conference opponents this season were UCLA, Oregon, Arizona State, and USC. Those four opponents currently have a combined record of 17-2, and their current Sagarin rankings are 12th, 11th, 8th, and 2nd. Stanford didn't really get much of a chance to ease in to its conference schedule...

In the first four games, T.C. Ostrander has 1,065 passing yards, which is more than either Stanford quarterback had all of last season...

Stanford's team leader in sacks last season was Trevor Hooper, with 2.5 sacks. This year, Clinton Snyder already has 3.5 sacks...

Erik Lorig's eight tackles against Arizona State were the most by a Stanford defensive lineman this season...

In last week's game, the Arizona State offense got 30% of its yards (134 yards) and two of its three touchdowns on two big plays, a 62-yard pass play and a 72-yard running play...

Redshirt sophomore Tavita Pritchard will be the least-experienced Stanford quarterback to start a game since Kyle Matter started against Boston College in the 2002 season opener. At the time, Matter was a redshirt freshman who had never stepped on the field. He had a pretty good game, completing 17 of 27 for 244 yards, with 1 TD and 1 interception...

The Bootleg Top Stories