≡ Menu

NFL Passing, 1950 Through Week 13, 2014

In case you haven’t noticed, 2014 is on pace to become the greatest passing season in NFL history. Which may not be surprising, since just a few months ago, the three best passing seasons in NFL history were the 2012, 2011, and 2013 seasons. Falling into fifth place will be the… 2010 NFL season. So passing numbers are on the rise, but you already knew that.

Through week 13 of the 2014 season, the NFL average Adjusted Net Yards per Attempt — defined as gross passing yards, plus 20 yards for every touchdown pass, minus 45 yards for every interception, and minus sack yards, all divided by the total number of pass attempts plus sacks — was at 6.26.  Most passing statistically typically take a trip south in December (and prior to SNF, the week 14 average was 5.85), but 6.26 would be a significant outlier even in our high-flying times. The graph below shows the NFL average ANY/A for each season since 1950.  Of course, we are doing a bit of apples-to-oranges comparisons by using full season numbers for all years and through-13-weeks numbers for 2014, but so be it:


Obviously ANY/A spiked after the 1978 rules changes, and has been steadily increasing over the last 35 years. But 2014 — depending on what happens in December — has the chance to mark another large increase in passing efficiency.

Since ANY/A is a combination of many statistics, it’s natural to wonder what exactly is driving the increase in ANY/A. Let’s first look at generic yards per attempt: i.e., gross passing yards divided by pass attempts (excluding sacks). While the peak Y/A was in the early ’60s (this if NFL-only data, so the AFL numbers are not included), the 2014 season stands out relative to recent years:

yards per attempt

Why has yards per attempt increased? Well, yards per attempt is the product of two other statistics: completion percentage and yards per completion. Here’s a look at completion percentage over time:

Completion Percentage

The 2014 season is set to smash the completion percentage record of 61.2%, set last year. Through thirteen weeks, teams are completing an incredible 63.0% of passes. But while a higher completion percentage often leads to a lower yards per completion average, that hasn’t really been the case this year. Take a look:

yards per compl

So teams are completing passes at an insanely high rate, but yards per catch averages aren’t dropping by a commensurate rate; as a result, yards per attempt is high. But what about the other parts of ANY/A?

Relative to 2013, the sack rate has declined from 6.7% to 6.1%; that’s a sizable difference, and one reason why ANY/A has increased this year. But in general, the 2014 sack rate is not a big outlier relative to other recent seasons:

sack rate

On the other hand, quarterback interception rate looks to set yet another new low. This is the metric that has seen the steadiest pro-offense change over the last few decades, and 2014 continues that trend:

INT rate

But what might be the most interesting is the change in touchdown rate. Take a look:

td rate2

This is pretty surprising, and not something you have heard much about. Consider: in every year in the sixties, the NFL average was at least 4.6 touchdown passes per 100 attempts. That level was never reached from 1970 to 2013, but through 13 weeks in 2014, the rate was 4.7 touchdowns per 100 throws. That’s kind of incredible, isn’t it?

A quick thought as to a possible explanation would be that teams are throwing more often near the goal line, and plays that used to be a rushing touchdown are now passing touchdowns. I haven’t crunched the numbers on the average length of touchdown passes this year, but I will note that rushing touchdowns per game are slightly down, so that jives with that hypotheses.


Completion percentage is through the roof in 2014, as the short-passing game has dominated the NFL. But yards per completion hasn’t dropped much at all, which has led to an increase in yards per attempt. Throw in small improvements in both sack rate and interception rate, plus a large increase in touchdown rate, and it all adds up to a big improvement in ANY/A. We’ll have to check back in the off-season to see where the final 2014 season numbers.

  • Ty

    All good points made. By the way, I remember an article on FO talking about Play Action passes and how QBs (Even crappy ones) become more efficient on such types of passes. I can’t remember if it was the author that mentioned this, or it was a comment, but the remark was that Play Action passing would be like what the corner 3 is to the NBA. Arguably the most efficient possession (aside from a dunk inside the paint), and it took a team like the Spurs to really figure that out… then other teams began to catch on. Compared to the NBA, though, the NFL seems to move at a glacier pace when it comes to analytics (the NBA isn’t that fast, either, but it has greatly improved compared to 10 years ago).

    Something that people will mention with PA passing is that you need a good running game to have a good PA, but if you look at the 2013 Ravens (which had a historically bad running game), they were actually pretty good at PA. If there is a coach with enough balls, or desperate, or a coach with a poor QB situation, they could try to use this to their advantage, at least until teams catch on and it no longer becomes an inefficiency in the market.

    All this that I mentioned relates back to the original point of the article. If more teams utilized more PA passing, it will cause the average ANY/A to keep increasing.

    • Rodrigo

      Isn’t that the basis of the read option? It has worked pretty well and has better YPA than regular plays, both if you decide to run or to pass; And like in the NBA most talking heads think it is a “fad” or a “flavor of the month” while the use has only increased, and producing good results even with more exposure* just like the three pointers weren’t accepted at the time.

  • Red

    During the first half of the season, officials were calling defensive penalties very tight, which is probably what caused the spike in ANY/A. However, the last month has seen fewer penalties and ANY/A has dropped back down to 2011-2013 levels. Unless they choose to “re-emphasize” defensive penalties again (please don’t!), I would expect ANY/A to continue dropping through week 17 and stabilize around 5.9 next season.

    Chase, any chance you could list week-by-week ANY/A for the 2014 season?

    • James

      You need to control by comparing to other years, because ANY/A is always higher earlier in the year and drops down, mostly for weather as Chase said.

  • ammek

    Is YAC a factor? Is there a difference in the air yards / yac split compared with previous years?

  • James

    “I haven’t crunched the numbers on the average length of touchdown passes this year, but I will note that rushing touchdowns per game are slightly down, so that jives with that hypotheses.”

    I’d like to see this. I’ve noticed that while passing touchdowns have been up over the past few years total touchdowns haven’t increased very much, as they’ve mostly cannabalized rushing touchdowns.

  • Richie

    The consistent improvement over 60+ years on the charts for completion percentage and interception percentage is amazing.

    I wonder if there would be any value in a team implementing a 1954-style passing game in 2014? (Which I assume to be that most pass attempts were 10+ yards downfield.) Would defenses set up to defend against short passes have trouble defending it?

  • Clint

    Reading this late, but I think it’d be cool to figure out the biggest gaps between td% and int%. Like that great season Steve DeBerg had in 1990 http://www.pro-football-reference.com/leaders/pass_int_perc_single_season.htm. In general, I don’t think there’s enough focuse on these percentages. People judge solely based on the ratio but that doesn’t tell the whole story.

    • What do you mean by biggest gaps?

      • James

        Think he means difference: TD% – INT%

      • Clint

        I’m sure there’s a much better way to phrase it, but let’s say Aaron Rodgers has a TD % of 8.7 and an INT% of 1.3. That’s a 7.4% difference… or “gap” as I had put it. And there’d have to be a minimum amount of attempts required to be on the list.
        Just a thought