Quarterback is the most important position in all of sports, and statistics for tracking quarterback performance have evolved nearly as much as the players who play the position in the NFL. As such, new acronyms like ADOT (average depth of target) and QBR (quarterback rating) keep popping up frequently, leading to consistent confusion and a distrust in analytics. Here, the 33^{rd} Team breaks down what each of these statistics mean and where they came from.

*(Note: Although quarterbacks are frequently rushers, we will only focus on passing-based statistics.)*

The most basic of quarterback stats, these are the earliest form of analytics that simply involved counting whenever something occurred on the field. Although much of the field has moved on to more-advanced categories that we will get into, fans will continue to obsess over “4,000 passing yards” or “50 passing touchdowns.” Many of these statistics are flawed in some way or another, from ignoring opportunity, scheme or the skill level of the passer’s teammates, but they will remain due to the ease of fan understanding.

Smarter coaches began to realize the flaws with the basic counting statistics and created the earliest forms of rate statistics. Each of these statistics accounts for the opportunity that a quarterback had – whether that’s dividing by the number of pass attempts (as seen for each of completion, touchdown, interception, and sack percentage) or some other counting statistic like completions or games played. You’ll find that, outside of completion percentage, many of these rates have a mixed reception among football coaches and GMs, but they do start to solve some of the issues mentioned for the counting statistics.

In the early days of the NFL, there was a lot of debate on how to crown the best passer. Passing yardage and completion percentage were each used for several years in the 1930s, but in 1941 the earliest ranking was created to compare quarterbacks. This system had several major drawbacks, most notably an inability to rank quarterbacks across multiple seasons, so commissioner Pete Rozelle created a committee in 1971 to develop a solution. Their answer was the formula for **passer rating**, adopted by the NFL in 1973. Although the formula itself is complex, it accounts for attempts, completions, passing yards, touchdown passes, and interceptions with the highest possible rating a 158.3 (a “perfect” rating). This rating has steadily improved from year-to-year, with an average in 2020 of 93.66. The NCAA created a similar statistic with the same inputs, but a few differences in the calculation – their maximum score is a 1,261.6.

Passer rating has several drawbacks, such as not measuring the rushing impact of a quarterback or accounting for the difficulty of a specific play, so in 2011 ESPN announced **Total Quarterback Rating (QBR)**. Also known simply as quarterback rating, QBR functions on a scale from 0-100, with the average passer at a 50 and a Pro Bowl passer at around a 75. To keep it simple, the statistic includes the same counting statistics as passer rating, but adds in rushes, sacks and penalties on the QB as well as a series of adjustments for down-and-distance (a 4-yard pass is more valuable on third-and-3 than fourth-and-8), yards after the catch by the receiver, whether the QB was pressured, the overall ability of the opposing defense, and even discounts statistics gained in garbage time. Although the formula itself is not publicly available, the final score functions as a percentile – meaning that a score of 60.5 indicates that a quarterback performed better than 60.5% of all other quarterback outings since 2006. To read more on QBR, click here.

**Approximate Value (AV )** is another popular statistic that aims to put a single number on the seasonal value of a player at any position from any year. Created by Doug Drinen in the late ‘aughts, this formula (for offensive players) takes a team’s offensive points per drive, scales it against the league average for points per drive, and splits up the result to each member of the offense based on their contributions. This formula is notable for ignoring (nearly!) all passing statistics, but rather taking a percentage of the offense’s overall output and then crediting each quarterback on the team based on his percentage of the team’s passing yards and how pass-focused the team was. This is rounded to a final, whole number that can compare the season-long output between any two players since 1960. The all-time record is a tie between Marshall Faulk’s 1999 campaign and LaDainian Tomlinson’s 2006 showing, each with 26 AV and a combined 4,752 total yards and 43 touchdowns, while Lamar Jackson’s 2019 campaign (25 AV) barely beat out 2007 Tom Brady and 2011 Aaron Rodgers (23 AV each) for the all-time passing mark. To read more on the founding and exact calculation of AV, click here.

The last few statistics mentioned here are best described as improvements on earlier rate statistics. After the popularity of yards per attempt (YPA), several spinoffs were created. **Net Yards per Attempt (NY/A)** is the simplest, accounting for yards lost when sacked in addition to the basic calculation. **Adjusted Yards Per Attempt (AY/A)** and **Adjusted Net Yards Per Attempt (ANY/A)** each include a bonus for touchdown passes and a detriment for interceptions.

**New-Age Statistics**

With these complicated formulas out of the way, where is the field going now? With player tracking data recently becoming available as well as the increase of fully-staffed statistics outfits like Pro Football Focus (PFF), Football Outsiders, Sports Info Solutions (SIS), Pro-Football-Reference (PFR), Elias Sports Bureau, Sharp Football Stats, and the NFL’s NextGen Stats (NGS), the sky is the limit. Notably, these companies each use different methodologies to compute their statistics, from direct tracking data (NGS) to employing human trackers (PFF and SIS), so the actual values can vary slightly even when the statistic is the same. Regardless, here are some of the statistics that have stood out to us for one reason or another:

**Indexed Statistics** may not be a specific statistic in itself, but the idea is in taking a rate or formulaic statistic and comparing it to all other quarterbacks with an exact average of 100. For example, completion percentage index would take Tom Brady’s 7.6 yards per attempt in 2020 and compare that to the rest of the league, awarding him a 107 and indicating that his yards per attempt was 7% better than league average. This group of statistics is helpful because you can quickly compare a passer to their peers instead of wondering how good a 7.6 YPA or a 7.12 NY/A is.

Several newer statistics also attempt to separate a quarterback’s passing from the yards-after-the-catch ability of his receivers. PFF’s **Average Depth of Target (ADOT) **and NGS’ **Average Intended Air Yards** both calculate how deep down-the-field a quarterback is attempting his throws, rather than only considering completions. Each outfit goes further with (respectively) **Air Yards Percentage** and **Average Completed Air Yards**, now looking at completions and ignoring any yards the receiver picked up after the catch. Sharp Football Stats computes the same with its **Yards In Air**.

What about identifying the quarterbacks who are completing 5-yard passes on third-and-long? PFF and Sharp Football Stats each calculate a **Short of Sticks Percentage**, measuring what percentage of a passer’s throws don’t go beyond the first-down marker, while NGS’ **Air Yards to the Sticks** counts the difference between the spot where a passer attempts to place his throw and the first-down-marker. Football Outsiders uses **Air Less Expected on third downs (ALEX)** to roughly the same effect. To close the book on PFF and NGS, both calculate **Time to Throw**, the average time from the snap to the release on all non-sack passing downs.

NGS’ tracking data also allows them to compute several highly-technical statistics that could be seen as subjective if it weren’t calculated. **Aggressiveness**, which is the percentage of passing attempts that a QB attempts to a receiver with less than a yard of space from the nearest defender, is one, while **Completion Probability** computes the probability that any given pass will be completed based on the receiver’s separation from the nearest defender, his spot on the field, the amount of pressure on a QB, and more. This can also be used to calculate **Completion Percentage over Expectation** for QBs, a stat led by Deshaun Watson in 2020 with Josh Allen, Aaron Rodgers and Kirk Cousins close behind.

In this crowded field, Football Outsiders has carved out their niche with largely opponent-adjusted statistics. With **Defense-adjusted Yards Above Replacement (DYAR)**, they compute how many passing yards a quarterback truly added to his team, adjusted for the defense. Similarly, **Defense-adjusted Value Over Average (DVOA)** calculates how much better the passer is than an average quarterback but is represented by a percentage and now looks at more than just passing yards. This is later translated into **Effective Yards**, a yards per attempt figure that lets you compare how much better a quarterback played compared to their basic counting stats.

The field of football statistics has grown from something you can count on your television to formulas that you need advanced tracking data and a math degree to measure, but there is no sign of stopping and each new addition brings the possibility of finding a competitive advantage over the opposing team. As Vince Lombardi once said, “The measure of who we are is what we do with what we have.” We now have more than ever. It’s up to the prepared mind to figure out what to do with it.

## Comments