[terry]

Well, except yards per play is the meta-statistic. Total yardage is meaningless without knowing how many attempts it took to get there. If you pump out 80 plays or whatever kal had, you're just going to get more yardage, scores, and first downs -- everything. It's just simple math.

An extreme example. Let's say I threw 100 times in a game, for a total of 300 yards. That, by traditional reckoning, is a pretty successful game -- but of course at 3 yards per attempt it's entirely possible that my hypothetical passer and offense never converted a first down.

The correct way to think about offense is on a per-play or per-drive basis. Because possessions are typically equal between teams (excepting onside kicks), the team that more efficiently uses possessions (and plays) will be the winning team. The particular raw totals that you achieve in getting there are just aesthetics and shouldn't be confused with actual effectiveness.

I agree with this. We are accustomed to evaluating offense by yards per game. But in the no-huddle spread offense era, yardage per game is becoming a poor metric. A team can increase its yardage simply by running plays quickly to increase the number of plays per game, rather than by being more effective.

D.ykes' offense is designed to run as many plays as possible. C'al ran 99 plays against Northwestern, compared to about 71 plays per game for the average team. So C'al ran about 40% more plays than average. Statistically, it's as though C'al played an extra 1.6 quarters of football compared to the average team. Just based on the sheer number of plays, you would expect C'al to have more of everything -- more possessions, more yards, more points, more punts, more turnovers. But C'al also is giving its opponents more possessions, thus more opportunities. The "per game" numbers for both C'al and its opponents should be considerably higher this year. But that does not mean C'al's offense is more effective.

C'al gained 5.5 yards per play. That's the same as Stanford's yards per play last season. C'al gained 7.1 yards per pass attempt, which is average. C'al gained 2.7 yards per rush attempt, which is well below average. That low figure is due partly to the four sacks C'al allowed. But even after taking out the sacks, C'al still ran for a mediocre 3.8 yards per attempt.

Goff threw for 445 yards, but his pass efficiency rating was 120.6, which is just so-so. As a point of reference, Hogan's pass efficiency rating last year was 147.9. Goff threw as many interceptions in one game as Hogan threw last season.

C'al scored 30 points, which is not all that impressive given the number of plays they ran (99). C'al scored .303 points per play, which would have ranked 95th in the nation last year. As a comparison, Stanford averaged about 28 points per game last year while running just 68 plays per game. Stanford scored .403 points per play last year, which was just slightly above average (52nd).

I expect them to get more comfortable with their offense. It will be interesting to see how they do over time.

[terry]

Well, except yards per play is the meta-statistic. Total yardage is meaningless without knowing how many attempts it took to get there. If you pump out 80 plays or whatever kal had, you're just going to get more yardage, scores, and first downs -- everything. It's just simple math.

An extreme example. Let's say I threw 100 times in a game, for a total of 300 yards. That, by traditional reckoning, is a pretty successful game -- but of course at 3 yards per attempt it's entirely possible that my hypothetical passer and offense never converted a first down.

The correct way to think about offense is on a per-play or per-drive basis. Because possessions are typically equal between teams (excepting onside kicks), the team that more efficiently uses possessions (and plays) will be the winning team. The particular raw totals that you achieve in getting there are just aesthetics and shouldn't be confused with actual effectiveness.

I agree with this. We are accustomed to evaluating offense by yards per game. But in the no-huddle spread offense era, yardage per game is becoming a poor metric. A team can increase its yardage simply by running plays quickly to increase the number of plays per game, rather than by being more effective.

D.ykes' offense is designed to run as many plays as possible. C'al ran 99 plays against Northwestern, compared to about 71 plays per game for the average team. So C'al ran about 40% more plays than average. Statistically, it's as though C'al played an extra 1.6 quarters of football compared to the average team. Just based on the sheer number of plays, you would expect C'al to have more of everything -- more possessions, more yards, more points, more punts, more turnovers. But C'al also is giving its opponents more possessions, thus more opportunities. The "per game" numbers for both C'al and its opponents should be considerably higher this year. But that does not mean C'al's offense is more effective.

C'al gained 5.5 yards per play. That's the same as Stanford's yards per play last season. C'al gained 7.1 yards per pass attempt, which is average. C'al gained 2.7 yards per rush attempt, which is well below average. That low figure is due partly to the four sacks C'al allowed. But even after taking out the sacks, C'al still ran for a mediocre 3.8 yards per attempt.

Goff threw for 445 yards, but his pass efficiency rating was 120.6, which is just so-so. As a point of reference, Hogan's pass efficiency rating last year was 147.9. Goff threw as many interceptions in one game as Hogan threw last season.

C'al scored 30 points, which is not all that impressive given the number of plays they ran (99). C'al scored .303 points per play, which would have ranked 95th in the nation last year. As a comparison, Stanford averaged about 28 points per game last year while running just 68 plays per game. Stanford scored .403 points per play last year, which was just slightly above average (52nd).

I expect them to get more comfortable with their offense. It will be interesting to see how they do over time.