The current method of quantifying the performance of NFL quarterbacks is with the much-aligned Passer Rating.  It's calculated with typical quarterback stats, including completion percentage, yards per completion, touchdowns, etc.  However, it has two glaring downfalls:  (1) It ranges from 0 to an obscure 158.3, and (2) it consistently ranks bad quarterbacks quite highly. 

A new statistic called Defense-adjusted Value Over Average has been developed that seeks to avoid these pitfalls.  It's based on a percent scale, making it easy for normal people to understand what's good and bad.  If a quarterback's passer rating is 95, is that good?  Who knows?  With the DVOA, it's easy to see that a value of 100% is good while a value of 50% is not (although DVOA uses negative percentages, which to me seems to defeat the purpose of using percentages).  Also, a quarterback like Donovan McNabb, who traditionally has a high quarterback rating, is ranked lower with DVOA (sorry there's no proof for that; it was in a Popular Science magazine article). 

Personally, I'd like to develop my own rating system that takes into account whether the team won or lost.  Because when it comes down to it, the best quarterback on a losing team is still on a losing team. #sports