≡ Menu

Want to see how passing has changed in the NFL over the last 63 years? A picture is worth at least 1,000 words in this case. The graph below shows the number of interceptions per dropback (red), sacks per dropback (purple), non-INT incomplete passes per dropback (yellow) and completions per dropback (green). Of course, a dropback is simply a pass attempt or a sack. The information is stacked on top of each other for ease of viewing.


In 1950, only 43% of dropbacks resulted in a complete pass; that number was 57% last year. As you can see, sacks and interceptions have been trending down for decades.

Update: People often wonder what’s the point of twitter. Well, a few minutes after I tweeted this post, Trey suggested that it might look better as a stacked area graph. I didn’t know what that was, but after trying it in Excel, I’m with Trey. Here was the original graph (which still has some value if you really want to trace a specific year):


  • Tim Truemper

    I read recently in some post (do not remember) about the “dead ball” era of Pro Football being the 1960’s. But from 1970 to 1975 passing was really going down compared to the previus decade. Knowing you have a lot to do, playoff data on its own would be interesting.

    • Chase Stuart

      The dead ball era was definitely not in the ’60s… ’73 to ’77 is what I would say off the top of my head. I think part of it (arguably) had to do with just bad luck with good quarterbacks retiring/getting old (Unitas/Namath/Starr/Lamonica/Dawson) and some of the new stars not being great then (Fouts/Anderson).

    • Richie

      I think there was a conventional wisdom that prior to the 1980’s, the NFL was all about running the ball. But upon further investigation (not sure if it was FO or PFR that pointed it out to me) it was shown that the 70’s were an extreme running decade, but the 50s and 60s had quite a bit of passing.

  • There was an article that came out a few years ago in the Journal of Quantitative Analysis in Sports (http://www.degruyter.com/view/j/jqas.2009.5.2/jqas.2009.5.2.1166/jqas.2009.5.2.1166.xml?format=INT) that looked at the trajectory of passer rating variables over time (1960-2007), and identified statistically significant “structural breaks” in the leaguewide averages for qualifying QBs. “Structural breaks” is just a fancy way of saying when one “era” shifts to another. Here are the eras identified for each variable:

    Passer rating — ’60-’79, ’80-’94, ’95-’07
    YPA — ’60-’07
    Comp% — ’60-’79, ’80-’99, ’00-’07
    INT% — ’60-’83, ’84-’92, ’93-’07
    TD% — ’60-’69, ’70-’07

  • Tim Truemper

    Upon second thought, and to Chase’s point, there did seem to be a transition problem with some great QB’s retiring and the slowness of the development of new ones. Still, the change to the new blocking rules and the 10 yard holding penalty change from 15 seemed to be quite a change in increase in passing. For example, the tremendous increase in passing yards and TD’s by Staubach and Bradshaw from 1977 onward.

    The eras link was very interesting too.

  • scott t

    Could you also show us the corresponding adjusted yardage data? My gut tells me it’s all about Bill Walsh and the safer, West Coast-style short passes in the flat.