[CTcard]

[quote author=teejers1 link=topic=18454.msg220249#msg220249 date=1515104384]
I will just end with this:  it's no coincidence that Stanford's best teams under Shaw had prolific offenses:  2011 (excellent) and 2015 (very good) on a YPG metric.  [And again, if the Os under Shaw have been good or effective these last couple years, then why do we run so few plays with such a "good" offense?].

Our 2012 and 2013 Rose Bowl teams called to disagree with you.
[/quote]

How we should think about and measure our offense is a good question. I personally believe in the premise that there is more than one way to skin a cat, and thus believe we should consider measures that don't depend upon type of offensive system, tempo, etc.

Yards per game is not a good metric for comparing teams over the course of a season if the teams operate at drastically different tempos. It's not irrelevant, just not good. It's roughly equivalent to using total yards on the season for teams that play a different number of games.
Yards per game is a good metric for comparing opponents within one game - as both teams will typically have the same number of possessions (just about).

Yards per play at least takes tempo into account, but is highly style dependent.
It becomes a better measure over a large data base, but is quite poor over a single game.
[I give this years Stanford-SDSU game as an example (only an example - one game doesn't prove anything for general trends). Stanford was outgained for the game 254-353. Stanford outgained SDSU per play by 5.9 to 4.5. I'd say the former was more indicative of the flow of play, with the issue being that Stanford couldn't get first downs so didn't have many plays to average out the two big Love runs.]

Per possession measures seem the obvious best choice, but for whatever reason they are rarely reported for football. They are quite standard for basketball and some other sports.
The best site I know of for reporting per possession stats is bcftoys.com, which several people here use from time to time.
Points per drive is the most easily understood of their stats. There are also versions of yards per drive but they get a little complicated. Putting it all together gives an offensive efficiency value that they then adjust for opponent. That is probably the best way to describe the quality of an offense in one number, with the very serious caveat that one number cannot answer a complex question such as "how good is this offense".


Stanford's best offenses under Shaw were 2011 and 2015. They were also the best at yards per game, because yards per game is relevant. However, Stanford hadn't slowed down quite so much by 2011, so the per game numbers slide quite a bit for 2015 despite great per possession numbers.


So, to look at how these things compare, and Stanford's offense over the Shaw years, here are the overall Offensive Efficiency rating (OFEI), national rank, Points Per Possession, national rank, and yards per game, national rank.

Year      OFEI      Rank        PPP    Rank      ypg      Rank
2017      3.01*      17          2.78    17        389        79 
2016    -0.21        79          2.22      62        367      103
2015      1.45          2          3.58        2        436        40
2014      0.20        50          2.30      50        389        77
2013      0.60        27          2.67      30        406        69
2012      0.15        54          2.08      71        374        86
2011      1.16          6          3.56      2        489        8

*  the numerical formula for OFEI obviously changed in 2017, so the raw number isn't comparable to other years.

I would stipulate that the first two categories are far better measures of the quality of Stanford's offense, and everybody's offense, than the third category. Still, yards per game is related enough that it goes up for good offenses and down for bad.

[CTcard]

[quote author=teejers1 link=topic=18454.msg220249#msg220249 date=1515104384]
I will just end with this:  it's no coincidence that Stanford's best teams under Shaw had prolific offenses:  2011 (excellent) and 2015 (very good) on a YPG metric.  [And again, if the Os under Shaw have been good or effective these last couple years, then why do we run so few plays with such a "good" offense?].

Our 2012 and 2013 Rose Bowl teams called to disagree with you.
[/quote]

How we should think about and measure our offense is a good question. I personally believe in the premise that there is more than one way to skin a cat, and thus believe we should consider measures that don't depend upon type of offensive system, tempo, etc.

Yards per game is not a good metric for comparing teams over the course of a season if the teams operate at drastically different tempos. It's not irrelevant, just not good. It's roughly equivalent to using total yards on the season for teams that play a different number of games.
Yards per game is a good metric for comparing opponents within one game - as both teams will typically have the same number of possessions (just about).

Yards per play at least takes tempo into account, but is highly style dependent.
It becomes a better measure over a large data base, but is quite poor over a single game.
[I give this years Stanford-SDSU game as an example (only an example - one game doesn't prove anything for general trends). Stanford was outgained for the game 254-353. Stanford outgained SDSU per play by 5.9 to 4.5. I'd say the former was more indicative of the flow of play, with the issue being that Stanford couldn't get first downs so didn't have many plays to average out the two big Love runs.]

Per possession measures seem the obvious best choice, but for whatever reason they are rarely reported for football. They are quite standard for basketball and some other sports.
The best site I know of for reporting per possession stats is bcftoys.com, which several people here use from time to time.
Points per drive is the most easily understood of their stats. There are also versions of yards per drive but they get a little complicated. Putting it all together gives an offensive efficiency value that they then adjust for opponent. That is probably the best way to describe the quality of an offense in one number, with the very serious caveat that one number cannot answer a complex question such as "how good is this offense".


Stanford's best offenses under Shaw were 2011 and 2015. They were also the best at yards per game, because yards per game is relevant. However, Stanford hadn't slowed down quite so much by 2011, so the per game numbers slide quite a bit for 2015 despite great per possession numbers.


So, to look at how these things compare, and Stanford's offense over the Shaw years, here are the overall Offensive Efficiency rating (OFEI), national rank, Points Per Possession, national rank, and yards per game, national rank.

Year      OFEI      Rank        PPP    Rank      ypg      Rank
2017      3.01*      17          2.78    17        389        79 
2016    -0.21        79          2.22      62        367      103
2015      1.45          2          3.58        2        436        40
2014      0.20        50          2.30      50        389        77
2013      0.60        27          2.67      30        406        69
2012      0.15        54          2.08      71        374        86
2011      1.16          6          3.56      2        489        8

*  the numerical formula for OFEI obviously changed in 2017, so the raw number isn't comparable to other years.

I would stipulate that the first two categories are far better measures of the quality of Stanford's offense, and everybody's offense, than the third category. Still, yards per game is related enough that it goes up for good offenses and down for bad.