[CTcard]

I also don't think Shaw's concept of "efficient" is sound. What efficiency really is is the ratio of useful work performed by the offense to the total energy expended by the offense. A reductive but useful way to think about it is yards per play. Or, if one wanted to adhere to the Shawvita concept of avoiding negative plays, the yards per play of Stanford's typical/median play. It is NOT the number of positive plays you have, which is how we talk about it. An offense that has eight positive plays out of ten and averages 5.20 yards per play is exactly as efficient as an offense that has five positive plays out of ten and also averages 5.20 yards per play. They both get 50 yards in that ten play sequence, and are both very inefficient. For too long we've been stuck in a three yards and a cloud of dust era of football in which getting a few yards each play is seen as effective because you're then making third down manageable. That's antiquated. Good offenses avoid third down if they can.

That's not really the definition of efficiency, or an efficient offense. 

The most meaningful definition of offensive efficiency is points per drive. That is basically what FEI is, modulo some complicated manipulations that are supposed to account for strength of opponents and giving reductions for outcomes that really hurt (like turnovers). If you believe in it, you could say that FEI itself is the best measure of efficiency. 

Some advanced stats focus on yards per play. That's typically easier to dig out from many standard stats, but it's not really the same thing. Four yards each and every play means you don't punt. 40 yards on one play followed by 9 plays of zero typically means at least two or three punts every four drives. More broadly, if you are scoring frequently, less variation is typically better (i.e. it should increase your points per drive) whereas if you score rarely, more variation should help. 

There are definitely fan/stylistic issues on slow tempo versus fast tempo. I would tend to agree that I would rather watch more football rather than less and it offends my sense of competition that one can make a youtube of all the plays in a football game that lasts about 12 minutes. 
However, the winning versus losing considerations are different. Here are some reasons one might perform better in a slowed down version of a football game (i.e. fewer possessions).
1. Thin roster, particularly a thin defense. Fewer possessions and more time between plays might help a thin team get by. 
2. If you manage to get an early lead, lowering the number of possessions should give you better odds. (but vice versa)
3. If you have personnel that allows for overwhelming the defense in particular situations, changing the folks on the field for particular plays can provide a significant advantage - and changing personnel takes time. [This has been a particular point of contention on Stanford fan boards, as it might also signal what play is coming.]
4. If your opponent is simply not used to playing in this manner, it might throw them off; though it would seem easier to adjust to slow play than fast. 

Simply looking at won-loss records in previous Stanford seasons aren't really a good measure of whether playing slow is a good idea or not, as one cannot tell what would have happened with a different approach. 

An odd feature of Stanford in recent years is that we have tilted heavily towards the pass game while still sticking to a slow tempo. Since the clock stops after incompletions, this tends to run less total clock than the more traditional mode of playing slow with a dominant ground game, but it is still slow. We've also gone to a generally more dink/dunk passing offense since Hogan left. That may be as much about ability as intent, as an exception was 2018 when Costello was at his best.

[CTcard]

I also don't think Shaw's concept of "efficient" is sound. What efficiency really is is the ratio of useful work performed by the offense to the total energy expended by the offense. A reductive but useful way to think about it is yards per play. Or, if one wanted to adhere to the Shawvita concept of avoiding negative plays, the yards per play of Stanford's typical/median play. It is NOT the number of positive plays you have, which is how we talk about it. An offense that has eight positive plays out of ten and averages 5.20 yards per play is exactly as efficient as an offense that has five positive plays out of ten and also averages 5.20 yards per play. They both get 50 yards in that ten play sequence, and are both very inefficient. For too long we've been stuck in a three yards and a cloud of dust era of football in which getting a few yards each play is seen as effective because you're then making third down manageable. That's antiquated. Good offenses avoid third down if they can.

That's not really the definition of efficiency, or an efficient offense. 

The most meaningful definition of offensive efficiency is points per drive. That is basically what FEI is, modulo some complicated manipulations that are supposed to account for strength of opponents and giving reductions for outcomes that really hurt (like turnovers). If you believe in it, you could say that FEI itself is the best measure of efficiency. 

Some advanced stats focus on yards per play. That's typically easier to dig out from many standard stats, but it's not really the same thing. Four yards each and every play means you don't punt. 40 yards on one play followed by 9 plays of zero typically means at least two or three punts every four drives. More broadly, if you are scoring frequently, less variation is typically better (i.e. it should increase your points per drive) whereas if you score rarely, more variation should help. 

There are definitely fan/stylistic issues on slow tempo versus fast tempo. I would tend to agree that I would rather watch more football rather than less and it offends my sense of competition that one can make a youtube of all the plays in a football game that lasts about 12 minutes. 
However, the winning versus losing considerations are different. Here are some reasons one might perform better in a slowed down version of a football game (i.e. fewer possessions).
1. Thin roster, particularly a thin defense. Fewer possessions and more time between plays might help a thin team get by. 
2. If you manage to get an early lead, lowering the number of possessions should give you better odds. (but vice versa)
3. If you have personnel that allows for overwhelming the defense in particular situations, changing the folks on the field for particular plays can provide a significant advantage - and changing personnel takes time. [This has been a particular point of contention on Stanford fan boards, as it might also signal what play is coming.]
4. If your opponent is simply not used to playing in this manner, it might throw them off; though it would seem easier to adjust to slow play than fast. 

Simply looking at won-loss records in previous Stanford seasons aren't really a good measure of whether playing slow is a good idea or not, as one cannot tell what would have happened with a different approach. 

An odd feature of Stanford in recent years is that we have tilted heavily towards the pass game while still sticking to a slow tempo. Since the clock stops after incompletions, this tends to run less total clock than the more traditional mode of playing slow with a dominant ground game, but it is still slow. We've also gone to a generally more dink/dunk passing offense since Hogan left. That may be as much about ability as intent, as an exception was 2018 when Costello was at his best.