[gocard14]

[quote author=washingtonismoney link=topic=16412.msg186087#msg186087 date=1484198953]
The other funny thing is, let's stipulate some Shaw playcalls are Bad. How many of them are they? Given the harping over the Utah playcalls, which was, jeez, three years ago now, it seems Bad playcalls are pretty rare. Once or twice a game, we'll say. Is that really that big a deal?

That being said, I'm pretty skeptical of the notion we can identify good playcallers reliably anyway. I think Lane Kiffin has metamorphosed three or five times between Bad Playcaller and Good Playcaller, like the world's most indecisive butterfly/caterpillar.

As I pointed out, if Buddy Teevens had Christian McCaffrey running behind Murphy, Garnett, and company, his shotgun draws would look like genius plays.

Someday, someone is going to turn towards AI/neural networks to use the all-22 views to determine the optimal playcalls in every situation.  But until then, I agree that whether a play is good or bad play seems to be more susceptible to an individual's own biases.  We used to hate the fade, but now we seem to rarely complain when it is thrown.  Likewise, the Wildcat has been the object of much derision, except if I recall, someone went through and showed that we gained more yards out of the wildcat than on regular running plays despite the complete lack of deception (and, of course, every once in a while, you get a big completion out of it).

BC
[/quote]

A play calling AI would be really interesting to see, as so much of play calling success depends not just on being in the correct play for a given defense, but a measure of unpredictability. An AI that would always counter the defensive alignment with the correct play could be extremely predictable, and as such would require a degree of randomness incorporated into the programming. Because of this, it would sometimes call the "wrong" play, as an unexpected draw or screen on 3rd and 6 sometimes looks like a great call and sometimes blows up in your face.

[gocard14]

[quote author=washingtonismoney link=topic=16412.msg186087#msg186087 date=1484198953]
The other funny thing is, let's stipulate some Shaw playcalls are Bad. How many of them are they? Given the harping over the Utah playcalls, which was, jeez, three years ago now, it seems Bad playcalls are pretty rare. Once or twice a game, we'll say. Is that really that big a deal?

That being said, I'm pretty skeptical of the notion we can identify good playcallers reliably anyway. I think Lane Kiffin has metamorphosed three or five times between Bad Playcaller and Good Playcaller, like the world's most indecisive butterfly/caterpillar.

As I pointed out, if Buddy Teevens had Christian McCaffrey running behind Murphy, Garnett, and company, his shotgun draws would look like genius plays.

Someday, someone is going to turn towards AI/neural networks to use the all-22 views to determine the optimal playcalls in every situation.  But until then, I agree that whether a play is good or bad play seems to be more susceptible to an individual's own biases.  We used to hate the fade, but now we seem to rarely complain when it is thrown.  Likewise, the Wildcat has been the object of much derision, except if I recall, someone went through and showed that we gained more yards out of the wildcat than on regular running plays despite the complete lack of deception (and, of course, every once in a while, you get a big completion out of it).

BC
[/quote]

A play calling AI would be really interesting to see, as so much of play calling success depends not just on being in the correct play for a given defense, but a measure of unpredictability. An AI that would always counter the defensive alignment with the correct play could be extremely predictable, and as such would require a degree of randomness incorporated into the programming. Because of this, it would sometimes call the "wrong" play, as an unexpected draw or screen on 3rd and 6 sometimes looks like a great call and sometimes blows up in your face.