≡ Menu

The Ebb and Flow of the NFL Passing Game Since 1932

One of the two greatest quarterbacks of the first half of the 20th century

One of the two greatest quarterbacks of the first half of the 20th century.

The game played by Peyton Manning and Joe Flacco last night bears little resemblance to the game Sammy Baugh played. Teams pass much more frequently and efficiently than ever before. And there are external effects, too: In 2011 and 2012, the average carry went for 4.3 yards, the first two years the average has ever been so high. But the details are often lost when discussing how the game has changed, and today’s post will help to refine exactly how, and in what way, the game has changed.

Nine men have thrown for 4,900 passing yards in a season. Seven of them did so in either 2011 or 2012. How did we get here? The NFL has turned into a pass-heavy league, but these changes didn’t happen overnight: a series of rules changes since the berth of the league have promoted the pass-happy environment we see today. The first ever playoff game came in 1932, pitting the Chicago Bears against the Portsmouth Spartans, predecessors of today’s Detroit Lions. The game was scoreless until the fourth quarter, when Bronko Nagurski threw a controversial pass to Red Grange for the game-winning touchdown. The pass was controversial because in 1932, a player needed to be at least five yards behind the line of scrimmage in order to be eligible to pass. The league eliminated the rule after the game1, allowing quarterbacks and other passers to be able to pass from anywhere behind the of scrimmage.

In 1934, a slimmer and more aerodynamic football was introduced to make life easier for quarterbacks. In the 1940s, most teams switched from the single-wing to the T-Formation, which placed the quarterback in the center of the offense and helped promote the passing game. Sammy Baugh in Washington and Sid Luckman in Chicago dominated the league, and the Redskins or Bears won the title every year from 1940 to 1943. During World War II, every franchise was playing with depleted rosters, so the league experimented with a rule change that would permit liberal substitutions. Finally, in 1950, the league finally decided to go with free substitution rules on a permanent basis. As pro football historian Sean Lahman explains:

For the NFL’s first three decades, versatility was the most important trait for a player. Your starting quarterback had to be quick enough to play safety, your running backs tough enough to play linebacker. The downside to this approach was that a player with one specific skill – say blazing speed – might not be enough of an all-around player to crack the starting lineup. With free substitution legalized, specialization became the norm…. Free substitution helped the passing game immensely because it allowed coaches to use quick players at offensive end who weren’t big enough to play defense. Their speed could be used as a weapon, and many teams moved to formations that featured three ends and just two backs.

The changes produced immediate results, and by 1954, the passing game had exploded. For the first time in league history, the league average completion percentage topped fifty-percent, and the average pass attempt gained 7.2 yards. By point of reference, teams averaged 7.1 yards per pass attempt in 2012.

Looking at just yards per attempt, one would think that NFL teams aren’t more effective at passing now than they were in the middle of the 1950s. But the next stage in the aerial evolution didn’t make the passing game more productive: it made it more consistent. From 1950 to 1978, the league average sack rate hovered around 8.3 percent. Since then, the sack rate has steadily declined, and dropped to just 6.1% since 2007. We’ve seen similar declines in interception rate and steady increases in completion percentage. Woody Hayes famously said that when you pass, three things can happen, and two of them are bad. In reality, four things can happen, and three of them are bad (sacks, interceptions, and incomplete passes). That’s why the most notable change over the last 60 years of football is that when teams pass, there have been fewer and fewer bad things happening.

And while free substitution was helping out offenses, defenses were becoming more innovative. Tom Landry dropped a fifth lineman into coverage to bring his 4-3 defense to the NFL, a direct response to the improvements in the passing game. In response, teams saw slight upticks in completion percentage and threw fewer interceptions — by throwing shorter passes (yards per completion began to steadily decline). Teams like the late ’50s 49ers began using West Coast Offense principles to make the pass a less risky alternative, but offenses eventually found themselves up against a glass ceiling.  By the middle of the 1970s, defenses had begun dominating again. As a result, NFL coaches recommitted to the running game. In 1977, teams averaged just 25 passes per game (versus 29.4 in 1954 and 34.7 last year) and completed only 51.3 percent of passes.2 That low completion rate existed despite the presence of more horizontal passing offenses3, leading to a puny 6.5 yards per pass attempt average. The Falcons, Rams, and Broncos all held opponents under 150 points that season, and the league had seen enough. Before the next season would be played, the NFL enacted the biggest rules changes since free substitution occurred. Pete Rozelle stated the desired intent unambiguously: “more long-gainer plays, more passes, more scoring.”

An action of change.

An engine of change.

Cornerbacks like the Steelers’ Mel Blount had been allowed to bump, chuck, or otherwise make contact with receivers at any point until a pass was thrown. Beginning in 1978, the rules permitted such contact only within five yards of the line of scrimmage. Offensive lineman were finally permitted to extend their arms, and push with open hands, allowing for better blocking and fewer holding penalties. These two changes coincided nicely with the arrival of Bill Walsh and Joe Montana in San Francisco in 1979, and teams have adopted principles of the West Coast offense ever since.

From 1970 to 1977, teams averaged 5.3 interceptions per 100 passes; that average steadily dropped over the next 15 years, eventually falling by two full interceptions to 3.3 in 1993. At the same time, yards per completion dropped and completion percentage rose, further reflecting the desire of teams to complete shorter passes instead of waiting for big plays. By 1992, yards per completion dropped below 12.0 for the first time in league history, and it has remained that way nearly every year since. Pass attempts steadily increased, too, and teams averaged over 30 passes per game every year in the ’80s.

But while teams started passing more frequently, completing more passes, throwing fewer interceptions and getting sacked less frequently, the yards per attempt average for the league has continued to hover around 7.0. Meanwhile, in the ten year period from 1956 to 1965, NFL4 teams averaged 7.4 yards per attempt. Even accounting for sacks, net yards per attempt from ’56 to ’65 was 6.1, while from 1979 to 2003, the league average net yards per pass ratio hovered around 5.9. It helped that Johnny Unitas accounted for 7.3% of all passes thrown during that decade, but that was a much passer-friendly era than many remember. The evolution of the league has been far from linear, and that period stands out as an efficient but very different era. From ’56 to ’65, teams completed 51% of their passes but averaged 14.4 yards per completion, which are Tim Tebow numbers. Over the last ten seasons, the average has been 60.3% and 11.5 yards per completion, while pass attempts per team game have risen from 26.3 to 33.0.

Polian does not approve.

Polian does not approve.

Three more changes in the past decade have helped send passing numbers through the roof. Following the 2003 playoffs, the league decided to more strictly enforce rules regarding illegal contact, pass interference and defensive holding. After Carson Palmer and Tom Brady suffered serious, high-profile knee injuries, the NFL enacted several rules to restrict the actions of defensive pass rushers. Rules have also been enacted to prevent concussions, including a prohibition on hitting defenseless receivers and launching and leading with a helment or shoulder into the opponent’s helment or neck area. As a result, the passing game is more wide open than ever. Combined with the increased prevalence of the shotgun formation and the incorporation of spread elements into traditional offenses, and teams have been passing more frequently and more effectively than ever before. Three seasons in NFL history have seen at least 34 passes per team game: 1995, 2011, and 2012. And quarterbacks are staying healthier than ever, too.

In the ’50s and ’60s, teams viewed the pass as a big play option. You expected to be intercepted and sacked, but considered that simply a cost of doing business when teams averaged over 14 yards per completion. But the various rules changes have now turned the pass into a normal way to run your offense. Even when the passing game peaked in the ’50s and ’60s, the league average pass to run ratio barely topped fifty-percent. That ratio has been steadily rising following the ’78 rules changes (the ratio dipped as low as 42% in 1977), and teams passed on 57.7% of all pass plays in 2012.

The ratio of team passing to rushing first downs has following a similar pattern. After dropping below 50% during various points of the dead ball era of the ’70s, more and more teams are gaining first downs via the pass. 2010 was the highest mark in football history, when teams gained first downs 65.6% of the time through the air; then it jumped to 65.9% in 2011, and finally 66.4% last season. The latest revolution seems to involve tempo, as we say with Manning last night, and as we’ll likely see with Chip Kelly and the Eagles. Within the next few years, I think we’ll see that first down ratio creep even closer to 70%.

  1. There must be an inverse relationship between passing prowess and rule volatility, since it took the NFL twelve years to get rid of the Tuck Rule. []
  2. The NFL also saw a modern record 37.4 runs per team game that season. []
  3. Yards per completion dropped from 14.2 in ’54 to 12.7 in ’75, ’76 and ’77 []
  4. By NFL, I mean NFL, and not the combined NFL/AFL statistics. []
  • Guru

    I’m not sure which feat is more impressive: Manning’s game, or the fact that you researched and typed all of this in one evening.

    • Chase Stuart

      Ha, thanks Guru.

  • Richie

    I thought I had seen on some website where somebody calculated game scores for passers. I couldn’t find anything. But I whipped up a quick formula to rate a QB’s passing performance in a game. I am using Comp %, Yards, Y/A, TD, Int%. A perfect score would be 100, and would be done by setting (or tying) the record for passing yards, touchdowns and Yds/Att while completing 100% of the passes and throwing no Interceptions.

    The best performance since 1960 looks to be YA Tittle in 1962. He had a score of 85.0 (27-39, 505, 7-0).
    Last night, Manning had the 2nd-best game ever with an 80.5.

    The top-10:

    Player Year Score Comp Att Pct Yards Yds/att TD Int Int%
    YA Tittle 1962 85.0 27 39 69.2% 505 12.9 7 0 0.0%
    Peyton Manning 2013 80.5 27 42 64.3% 462 11.0 7 0 0.0%
    Johnny Unitas 1967 80.3 17 20 85.0% 370 18.5 4 0 0.0%
    Joe Namath 1972 79.5 15 28 53.6% 496 17.7 6 1 3.6%
    Craig Morton 1981 78.5 17 18 94.4% 308 17.1 4 0 0.0%
    Steve Grogan 1979 77.6 13 18 72.2% 315 17.5 5 0 0.0%
    George Blanda 1961 77.2 20 32 62.5% 418 13.1 7 1 3.1%
    Joe Kapp 1969 77.1 28 43 65.1% 449 10.4 7 1 2.3%
    Matt Flynn 2012 76.9 31 44 70.5% 480 10.9 6 1 2.3%
    Craig Morton 1970 76.2 13 17 76.5% 349 20.5 5 1 5.9%

    • Richie

      (Obviously, I am just looking at a pure passing statistic ranking – not taking into account sacks, running, opponent strength, etc.)

      • Red

        Richie, I have always wanted to create a Game Score metric for QB’s, similar to what baseball has for pitches. Any chance you could share the details of your methodology for brainstorming purposes?

        • Richie

          No problem. I just whipped up something simple. I am using 5 statistical categories, and weighting them equally with a maximum score of 20 points.
          Passing Yards
          Completion Percentage
          Passing Touchdowns
          Interception Percentage

          I simply take the passing yards, and divide it by the all-time record (554 yards in a game) to give a percentage. Then multiply by 20. Repeat for each category.

          The only oddball is interception percentage, since it’s a “low” statistic. I went with a maximum interception percentage of 18% (anything higher, and a QB would lose points). So I took 18% and subtracted the QB’s percentage and multiplied by 111. (18% x 111 = ~20).

          But it’s funny, as I pulled up my spreadsheet to check my numbers, I found a file that I had begun working on back in January. It looks like I was working with a different methodology then, based on the findings in “The Hidden Game of Football”. I don’t have the sack data handy, but it looks like that formula was:

          For that one, I was using the old PFR data download file (which apparently is no longer available), that only went back to 1995.
          In these ratings, it looks like Tittle and Manning are still 1-2. But overall my ratings above are giving a lot more credit to guys with high completion percentages.

          I guess the question is: do you want a game score to be something that just reflects excellent statistical performances, or something (like Hidden Game of Football) that does a better job to express the total value to the offense?

    • Scott L

      You forgot Drew Brees vs New England in 2009

      18-23 371 yards, 5 TD 0 Int

      • Richie

        That’s worth 79.4 – just behind Namath.

        • Scott L

          Yeah, I would list it if that’s the case.

          • Richie

            I only skimmed the list of top passing yard and TD games, and somehow missed Brees. I didn’t skip him intentionally.

  • Guru

    Thanks for the stats, Richie.

    It truly amazes me how Manning still amanges to be underrated despite his transcendent career. Today’s football chatter about this performance goes like this: “Yeah, awesome performance by the best regular season QB ever, but, you know, one Super Bowl ring…” It’s a fact that one player is not solely responsible for a team win. We see how fallacious the winining/championship logic is when we apply it to other team sport superstars like LeBron James, and James plays in a sport that is easier to impact as individual (five teammates on the field of play as opposed to 11, and those five also play defense). Heck, it seems that we still haven’t learned this lesson in baseball: Felix Hernandez wins a Cy Young a couple seasons ago despite his pedestrian “win” record, but we still want to make Max Scherzer a Cy Young front runner this season because of his 19-2 record. (Never mind the fact that he would be a candidate anyway when you actually measure his performance by metrics that, you know, actually pertain to the pitcher.)

    When will the stupid stop?

    • Red

      Seriously. Listening to sports radio the morning after Manning’s performance, the hosts basically said “That was nice and all, but none of it matters until he does it in the playoffs.” It seems like people actually penalize Manning for being so great in the regular season, like that implies his teams’ failures in the playoffs must be the result of choking or some sort of character flaw. QB win/loss record is perhaps the dumbest “stat” in all of sports, and it’s even more ridiculous over the tiny sample of single-elimination NFL playoff games. If a couple random bounces went the other way, Manning could easily have 3 SB wins and Brady could just as easily have only one.

      I saw a poll the other day (can’t remember the website) that asked fans to choose the most important stat for measuring QB play. The most common answer was Win %. All I could do was shake my head. I wonder if sports fans in other countries are as ignorant as most Americans?

      • Richie

        “That was nice and all, but none of it matters until he does it in the playoffs.” It seems like people actually penalize Manning for being so great in the regular season, like that implies his teams’ failures in the playoffs must be the result of choking or some sort of character flaw.

        I’m afraid Matt Ryan is heading down that same path. I think Ryan deserves a little credit for the fact that 2 of his playoff losses came against the eventual Super Bowl winners.

      • Joe

        “That was nice and all, but none of it matters until he does it in the playoffs.”

        The thing is he has done it in the playoffs (more or less):

        And I’m with all of you regarding the QB wins nonsense, but I’m afraid it will never end.

        • Red

          That’s the kicker – he’s actually played BETTER in the playoffs than a lot of QB’s who are known as “winners”. Here are the postseason AY/A for some current high-profile QB’s:

          Rodgers 8.38
          Brees 8.27
          Flacco 7.22
          Peyton 7.06
          Eli 7.01
          Big Ben 6.81
          Brady 6.54

          Interesting, huh? The three QB’s with multiple rings have all played worse than Peyton in the postseason. Brady’s numbers are especially glaring, since he’s widely considered to be the best playoff QB since Montana. It’s almost as if defense and special teams affect a QB’s playoff record…

          • Scott L

            You missed Kurt Warner. In terms of QBs who has elevated their game come playoff times, I would rate them as Warner, Brees, Montana and then Rodgers. Warner’s career playoff QB rating was nearly 10 points better than his regular season career. Brees is nearly 7 points higher, and Montana was about 3 points higher. Rodgers is a couple of points lower, although at 103.1 who can really complain?

    • Richie

      When will the stupid stop?

      I really thought that Super Bowl win would end it. As a Marino fan, I will always have to hear him be slighted because the lack of a championship. I guess when a guy is successful, he is going to take flack for the one thing that he hasn’t done very well (win in the post-season).

      • Scott L

        The bottom line is Peyton Manning is one and done in the playoffs 8 times. That’s a ridiculous number. And when his team has lost in the playoffs, they have averaged only 13.9 points per game with the offense (i.e., did not count the 14 points scored by Holliday on the kick returns from the Ravens game). I mean 13.9 points per game? That’s just not getting the job done. No other way to look at it, skirt around it, ignore it, sweep it under the rug, whatever you want to do. Facts is facts. Then the one and only year he wins the Super Bowl, his post season was actually worse than Trent Dilfer with the Ravens in terms of QB Rating. It is what it is. Tom Brady not far behind Manning since 2005, going 7-7 and failing in 2 Super Bowls to generate enough points (can’t put it on their defense).

        • Richie

          Yeah, Tom Brady was a clutch player in the playoffs until he wasn’t.

  • Guru

    “I saw a poll the other day (can’t remember the website) that asked fans to choose the most important stat for measuring QB play. The most common answer was Win %.”

    Perhaps the most maddening thing about that answer is that it the winning criteria isn’t even applied consistently. If one should concede that winning is transcendent and trumps everything else – stats, production, you name it – then it only begs the following questions:

    Who’s the greatest RB ever? Surely it’s Franco Harris hands down, since with his four Super Bowl wins he’s clearly superior to losers and underacheivers like Jim Brown and Barry Sanders? Who wastes their time ranking RBs by rushing yards and the like when the object is to win games?

    How about WR? It’s no contest that Lynn Swann and not Jerry Rice is the greatest ever, correct? Why does the media fawn so much over a player who simply won less over his career? Wouldn’t Rice gladly trade all of his useless records and useless stats for the extra Super Ring that Swann wears on his hand?

    We could go on like this for each position. Heck, let’s not keep this discussion to football: for all of the importance that the maniacal Michael Jordan places on winning, why do his fans consider him a better player than John Havlicek, who played similar minutes over his career? MJ’s stats? Hasn’t it already been established that the truly great players only play to win, and not to whore stats? Don’t the true greats also possess the “intangibles” to win that their loser, stat-whoring counterparts don’t have? So who cares about a “mixture of rings and stats” when Havlicek simply WON more than MJ?

    Or Trent Dilfer more than Dan Marino? Or Johnny Damon more than Ted Williams?

    • Scott L

      What people would think of Eli Manning would be vastly different had he not “led” his team to Super Bowl wins. The Giants defense woke up in those playoffs (I don’t think they were great defensively in the regular season in either of those seasons). Eli played “well enough” to help his team win. He certainly didn’t carry them. Heck he only led them to 17 points in the 1st Super Bowl and it required the most flukish play in Super Bowl history in order to do it. If the reliable Wes Welker makes a semi-acrobatic catch late in the game in the 2nd Super Bowl matchup, Eli never gets the ball back to be the “hero” once again. It goes to show you every QB needs a little luck to win it all. My Saints needed 5 turnovers by the Vikings (most were not really induced, I’ll even admit) in order to escape a nailbiter.

  • Guru

    Defense and special teams certainly affect a team’s record, and this is true even if a QB is entirely responsible for his team’s production on offense …which he is not.

    Another fact is that the choker Peyton Manning ranks 3rd all-time in playoff Adjusted Value (not weighted for each playoff round, but I refrain from doing that because the games aren’t inherently more important in successive rounds. The Super Bowl’s importance is only based on the leverage that is [i]being carried over[/i] by games won in previous rounds; at the end of things, it’s still one game out of three or four games needed to win a championship.) Not only is the “Manning isn’t a playoff QB” narrative a flawed way to assess his ranking among other QB greats, it’s also false.

  • Guru

    Richie, I definitely understand your point of view on Marino. Following from my previous comment about the pitfalls of just looking at a QB’s performance in the playoffs, I’d take Marino over almost every QB in history (including Brady, Bradshaw, and other “playoff QBs”) even if he never played a single playoff game. His dominance was why the Dolphins were even contenders in the first place; Brady couldn’t achieve that success with the inferior roster that Marino helped get to contention throughout his career.

  • Tim Truemper

    Too bad Woody Hayes didn’t consider all the wrong things that can happen with a running play (as a matter of contrast to his thinking about passing).

    Nice history lesson. I think the changes in rules serve as the benchmark for the different eras of pro football and thus, within group (i.e. era performance comparisons). Another point about early football players requirements above and beyond the versatility element that was so important prior to 1950. Endurance was important too as rosters were less than 40 players (up to the mid 60’s they were a little more than 33 active players). That meant you had to play special teams as well if you were a starter. The smaller rosters also emphasized versatility as well, thus you had Jerry Kramer and Paul Hornung kicking for the Packers in the early 60’s.