Before I begin, I want to give a shoutout to my man, OCC for putting in a lot of time gathering an amazing amount of Pro Football Focus data, that has been used in this post....
We always hear the arguments, does the pass rush make the secondary better, or does the secondary make the pass rush better, and there are advocates on both sides of the coin. I decided based on some Graduate School work that I was going to determine how significant pass rush is when looking at the number of wins a team could obtain.
I used a statistical method called regression, which is used to measure how dependent one variable is upon another. This creates 2 values that are important, the first being R-Squared which tells you for example how much of a teams win total is explained by the amount of pressure they put on opposing QB's. The other is the P-Value of the independent variable, in our case Pass rush, this tells you essentially whether or not that variable is statistically significant to determining the result.
The method, after the jump.
In order to consolidate all the measures of pass rush (Sacks, Hurries, Hits, etc) into one number that could be easily analyzed, I determined each team's Pressure % by adding the team's Sacks, Hurries and Hits, and dividing that by the total number of drop backs they faced (Attempts+Sacks). This number tells us what percentage of drop backs against a certain team faced pressure.
Based on this data, the regression determined that the PR% only explained 6.5% of a teams win total, and based on the P-Value, we can say that we are 95% confident that PR% is not statistically significant in determining a team's win total.
I was honestly shocked, based on the fact that it is commonly thought that putting pressure on the QB is extremely important to being successful in the NFL. I thought about it some more, and wondered what might explain more of this, and thought about the opponents QB rating. I won't post the table, but the regression showed, that opponents passer rating explained 31% of a teams win total, and that we can be 95% that it is statistically significant.
So Passer Rating against, is 5 times more effective in explaining wins, so I thought about the fact that interceptions are important to Passer Rating, I wondered if pressure might influence takeaways and therefore passer rating, The regression for these data sets, showed that pressure explained nothing about total takeaways.
I then set out to determine what in the heck does influence wins the most. So I pulled all kinds of data about turnover differential and QB ratings and Takeaways and 3rd down efficiency, and I determined that we can explain 78% of a teams win total by examining their opponents QB's Passer rating, combined with their QB's passer rating. Looking at those 2 numbers alone gives you 29% and 60% determination, which combine to the 78% mark. Ideally you'd like to achieve a higher determination than this, but attempting to combine more than 3 or 4 stats creates issues because so many stats in football are interdependent (Passer Rating and Ints thrown for example).
Interesting to note that take aways and your QB rating can explain 74% of a teams wins, but when you combine all 3 variable, takeaways becomes statistically insignificant, (probably because interceptions are a factor in Passer Rating).
So what in the world does this mean for the Cowboys?
To me, it means they made the perfect decision about where to upgrade their defense, and I believe it explains why they made the decisions they did. Incompletions and interceptions are big influences into the calculation of Passer Rating, and by going out and getting guys who can play man coverage, and make plays on the ball, will help increase takeaways, and lower their Passer Rating, and if they can get a top 5 Passer rating out of Romo again in 2012, we as fans can feel confident the win total will increase, which should result in a playoff birth.