For years, media talking heads have only discussed and judged Defensive backs based off interceptions. Obviously, interceptions don't tell us much about a DB outside of playmaking ability. It's like debating wide receivers based off just touchdowns, in that it can be correlative, or show a nose for something, but not the whole picture. For example we know that touchdown king James Jones is good, but not better than the touchdown deprived Calvin Johnson. Likewise, we know that despite having more picks, NYG S Stevie Brown isn't nearly as good as Jairus Byrd. With that being said, I see two stats as telling the entire story for DBs- Yards per Coverage Snap, and target percentage.
Let's start with YPCS, a PFF stat that is pretty intuitive. You take total Yards allowed for a DB, and then divide it by total Coverage Snaps (i.e. every passing play that the DB is on the field). The lower the YPCS, the better. Elite YPCS are generally around .9. The league leaders in the past two years have been Brent Grimes in '11 at .58, and Champ Bailey in '12 at .82. Guys like Revis, Cromartie, Sherman, and Tillman are some of the next guys in line in case you are doubting the stat. YPCS really takes into account everything about coverage, and allows a corner to demonstrate his coverage worth in multiple ways, by rewarding Corners for not being targeted, or for preventing completions when targeted. Take the Cowboy corners, for example. Both are quality corners. Brandon Carr allowed a YPCS of 1.17, which is decent. Throughout the season he was targeted more than Claiborne, but at the same time had more pass breakups and interceptions. Conversely, the less targeted Claiborne had a YPCS of 1.21, which rewarded him for not being targeted that much, but also noting his lower INT and PB stats. If anything, YPCS is best as a comparative stat when judging DBs who play the same position. Comparing YPCS across positions might not be that useful. A FS would likely have wild fluctuations in YPCS due to deep coverage responsibilities (in today's nfl, probably a very high YPCS), while 3rd corners tend to have steady, low YPCS (Mike Jenkins had the 15th best YPCS last year).
Next, we have target percentage, which is just the percentage of throws that go a certain DB's way (or defender) when he's on the field. Target percentage is entirely comparative within a secondary. Based off the rational assumption that QBs target the weaker links in a secondary (the guy who allows his man to be open, or cannot maintain his zone), a lower target percentage is better. Unfortunately, I could not really find much on target percentages for the Dallas DBs (I would imagine Claiborne's was lower than Carr's), but the Oakland Raiders of a few years back is a good example. Nmandi Asomwagh (sp?) was good enough to be significantly better than anybody else in their secondary. As a result, QBs ignored Nmandi's side and just went after the other DBs in the Oakland Secondary. While he has proven that he is not a world beater at CB (Eagles Dream team secondary), he was comparatively better than the rest of the Oakland Secondary.
There you have it. When you compare DBs at the same position, use YPCS. When you compare guys within a secondary, use target percentage. The stats are near perfect, and have two drawbacks. The first drawback is that neither statistic really takes into account how much pass rush affects coverage quality/success. The second is that it does not take into account INTs, or Pass Defense penalties. But hey, you can't have everything, and these two stats are pretty damn good. What do you guys think? Tell me in the comments below!