It's not really a statistical analysis until you can calculate the probability of an event (or likelihood of a hypothesis). So let's do that here. First of all, you need to either adjust all ratings before combining them and then calculate the correlation OR calculate the correlation each year without adjustment and then take the average. If you don't do that you'll just be picking up passer rating inflation. Best method is calculate correlation each year (with no adjustments) and then take the average. For Tannehill in Miami the average correlation between passing attempts and passer rating was -0.2358, and in Tennessee it's -0.0893. Those aren't directly comparable numbers however because the one from Miami was over 6 years while in Tennessee it's 1 year, so we have to adjust for sample size. I'll show how to do that in a moment. First, we need to know the distribution of such correlations. So let's look at the correlation between passing attempts and passer rating for every QB season where that QB threw 150+ passing attempts, from 1978 because the nature of the game changed from 1978. The mean correlation is -0.092 and the standard deviation is 0.332. Now.. you can't just look at +-2 standard deviations from the mean here because we took the mean over 6 years in Miami. That means the standard deviation has to be adjusted to 0.332/sqrt(6) = 0.1355. So the question is whether Tannehill's average correlation over 6 years is outside of the interval (-0.092-2*0.1355, -0.092+2*0.1355) = (-0.363, 0.179). And Tannehill's -0.2358 is within that interval. So when you actually do the statistical analysis (well.. now it's a hypothesis test) you find that even in Miami Tannehill's correlation between passing attempts and passer rating is NOT significant. In other words it can be explained by random variation alone. That really should settle this issue from a statistical point of view. Not only in Tennessee is Tannehill's correlation between passing attempts and passer rating not significant, it wasn't in Miami either.