I'm giving @calcuttarain credit for this for making RIDICULOUS claims about evidence and english and bla bla bla ;)
P.S. this is the 3rd try so I'm inserting this text so it sticks around [the really should fix the forum so you can actually edit posts]
And seeing as @Nobody thinks he so cool with his "All 22" and "film review". Psssh. Who nees film when you can just look at stats. Stats I say!
The argument is that a bad D leads to more offensive turnovers.
My argument is that this is nonsense. Let's look at the evidence!
In this first graph we see (high to low) pass attempts vs. giveaways. We actually see only 8% correlation which is, essentially no correlation. That is, in 2018 more passing does not lead to more turnovers. The one exception (a common theme) is Tampa Bay on the far left.
But that's just straight passing attempts, what about if offenses are more aggressive? Well, I used Yards/Attempt as a proxy for that. It's not a perfect proxy, but I think it does a good job of indicating which offenses are completing chunk plays.
What I expected is giveaways to INCREASE with a higher passing Y/A. I was completely shocked that the OPPOSITE is true. There is a slight correlation (-27%) between a high Y/A and more giveaways. That is, a more aggressive offense correlates slightly with fewer turnovers. Look at that one MASSIVE outlier: Tampa Bay again.
I have one more graph to show but I'll have to wait as I don't want to break the forum.
I'll say I am actually surprised by how little correlation there is between giveaways and both passing attempts and yards/attempt. There are some caveats, these are bulk stats that can miss some context because other factors are baked into the giveaways and passing stats. For example, having a high passing yards/attempt means you have a successful offense which will include fewer turnovers.
But the flip side is how much of an outlier Tampa is on both these stats. That's bad.
