≡ Menu

Back when passes were completed for first downs.

The New York Jets have the second best completion percentage in the NFL through five weeks.  That’s a shocking thing to say for many reasons, including the key fact that 38-year-old Josh McCown has taken every snap at quarterback for the team this season. The Jets are completing 71.6% of their passes, which is truly remarkable for this franchise.

Today I want to compare the 2017 Jets to their predecessors from 45 years earlier. The 1972 Jets were an interesting team.  That year produced a low key entry for the best Joe Namath season: he went 7-6 (missing one game due to injury) but led the NFL in passing yards, touchdowns, yards per attempt, Net Yards per Attempt, and Adjusted Net Yards per Attempt.  Namath was the best QB in the NFL that year, and was named a first-team All-Pro by the Pro Football Writers, the NEA, and Pro Football Weekly.1 But Namath completed just 50% of his passes that year, and as a team, the Jets completed just 49.6% of their passes.

It’s easy to look at the 2017 Jets with their 71.6% completion rate — a whopping 22 points higher than the ’72 squad — and conclude that, era adjustments aside, the 2017 Jets passing offense is more efficient. To be clear, era adjustments are enormously important when comparing passers across eras. You almost never want to compare players from different eras without making those adjustments. But today is the rare day where that’s not where I want us to focus. Because as discussed yesterday, completion percentage ignores two key elements of a passing game.

  • Sack Rate.
    • Namath had a 3.3% sack rate that season, and the Jets as a team had a 4.7% sack rate (yes, the Jets backup quarterback was sacked 6 times and had just 22 pass attempts, while Namath took just 11 sacks in 13 starts. Namath remains underrated by the general public in part because of his legendary ability to avoid sacks (which, of course, tanked his completion percentage).
    • The iteration of that Jets team 45 years later is not so adept: the 2017 Jets have a 9.2% sack rate. McCown has taken 15 sacks this year, many in situations where a better quarterback would have thrown it away (or completed a pass).
  • First Downs per Completed Pass.
    • When Namath was throwing passes, he was moving the chains.  A whopping 68% of the completed passes thrown by the 1972 Jets went for first downs.
    • Not so much under McCown: just 47% of completed passes have gone for first downs.

The 1972 Jets had 364 dropbacks (pass attempts plus sacks) and threw for 117 first downs.  That translates to 32% of all pass plays turning into a first down. The 2017 Jets have had 163 dropbacks, and have thrown for 50 first downs. That translates to 31% of all pass plays turning into a first down.2

Make sure you re-read that again to see how remarkable that is: despite the 22% advantage in completion percentage, the 2017 Jets are actually picked up first downs at a lower overall rate than the 1972 Jets passing attack. This isn’t on a per completed pass basis, but a per pass play basis! That’s how much the NFL today is different than the NFL of 1972.

The 2017 Jets, as it turns out, are the team most focused on completing passes rather than throwing for first downs. I put together a quick table (prior to last night’s game between Minnesota and Chicago) of passing stats through five weeks. Here’s how to read the Jets line.

Through five games, the Jets have completed 71.6% of their passes. They have a Z-Score of 1.83, which means the Jets are completing passes at a rate that is 1.83 standard deviations higher than average. They also rank 2nd in completion percentage. But the Jets have taken 15 sacks and only have 50 passing first downs. The Jets “passing first downs per pass play” rate — the 1D% column — is 30.7%. That is 0.51 standard deviations below average, so they have a Z-Score of 1D% of -0.53. The Jets rank 21st in percentage of pass plays that have picked up a first down. Finally, to compare the two categories, I subtracted each team’s Z-Score in 1D% from its Z-Score in Cmp%: so for the Jets, you take 1.83 and subtract -0.51, which gives you a grade of 2.34, which shows how biased in the direction of completion percentage the Jets offense has been. That’s the strongest in the league.

RkTeamGCmpAttCmp%Z-ScoreCmp% RkSk1D1D %Z-Score1D% RkComp Z
1New York Jets510614871.6%1.832155030.7%-0.51212.34
2Chicago Bears49314066.4%0.677.584228.4%-1.15291.82
3New York Giants513720267.8%0.985136429.8%-0.76261.74
4Baltimore Ravens59815364.1%0.161594729%-0.99281.15
5Miami Dolphins48213461.2%-0.4924103927.1%-1.52311.03
6Oakland Raiders59815264.5%0.2513125030.5%-0.56220.81
7Cincinnati Bengals510415965.4%0.4511155531.6%-0.25170.7
8Kansas City Chiefs512115876.6%2.951177241.1%2.4210.53
9Green Bay Packers512719066.8%0.766197134%0.4212.50.34
10Indianapolis Colts58614559.3%-0.9229.5184628.2%-1.21300.29
11Dallas Cowboys511217962.6%-0.181775831.2%-0.3719.50.19
12Detroit Lions511117364.2%0.1814186232.5%0160.18
13Minnesota Vikings48813366.2%0.63954734.1%0.45110.18
14Pittsburgh Steelers512019561.5%-0.422196230.4%-0.59230.17
15Buffalo Bills58513662.5%-0.218184831.2%-0.3719.50.17
16New Orleans Saints410515269.1%1.27345736.5%1.1360.14
17Carolina Panthers510115266.4%0.677.5155834.7%0.629.50.05
18Seattle Seahawks511117962%-0.3120136031.3%-0.34180.03
19San Francisco 49ers511519459.3%-0.9229.5156129.2%-0.93270.01
20Arizona Cardinals513622759.9%-0.7826.5197430.1%-0.6825-0.1
21Tennessee Titans59115259.9%-0.7826.5104930.2%-0.6524-0.13
22Cleveland Browns510719754.3%-2.0331145526.1%-1.832-0.23
23New England Patriots513319568.2%1.074167937.4%1.384-0.31
24Denver Broncos47912662.7%-0.1616134633.1%0.1714.5-0.33
25Atlanta Falcons48813565.2%0.41285236.4%1.17-0.7
26Washington Redskins48012166.1%0.61074837.5%1.413-0.81
27Philadelphia Eagles511017762.1%-0.2919136634.7%0.629.5-0.91
28Houston Texans59715861.4%-0.4522.5206234.8%0.658-1.1
29Los Angeles Chargers511619459.8%-0.82866834%0.4212.5-1.22
30Los Angeles Rams510216661.4%-0.4522.566437.2%1.325-1.77
31Jacksonville Jaguars57413754%-2.13254733.1%0.1714.5-2.27
32Tampa Bay Buccaneers49415560.6%-0.632576338.9%1.82-2.43
League Total3307521463.4%0384182232.5%00

The team most focused on picking up 1st downs rather than completing passes would be the Tampa Bay Bucs under Jameis Winston. The Bucs only rank 25th in completion percentage, which sounds pretty bad. But they rank 2nd in 1st down percentage, which sounds really good! The Chiefs are first in both categories — yeah, they’re having that sort of season — but Tampa Bay has been great at picking up first downs. In fact, 67% of Winston’s completions have gone for a first down, which is Namath-like. And Winston has a pretty good sack rate, too (4.3%). So when it comes to the Jets and the Bucs, completion percentage doesn’t tell a very useful story.

  1. Namath was a 2nd-team choice by the AP, which went with Earl Morrall, 9-0 QB of the undefeated Dolphins, as their first-team choice. But it’s not controversial to say that Namath was the best QB in the NFL that year, given that he led in ANY/A and won the majority vote for best QB, and also beat out Morrall in the organizations that made All-Conference (Sporting News and UPI) votes rather than All-Pro votes.  Of the five organizations that chose between Namath and Morrall, only one went with Morrall. []
  2. Note that this includes Namath’s backup, who was terrible, and a long completion by the Jets punter on a fake.  Namath in ’72 probably picked up a first down on 33-34% of his pass plays, while McCown is at 30% this year. []
  • sacramento gold miners

    I feel this issue is magnified in the postseason against better competition. The QB who is eager to unload the ball for a shorter completion, can damage his teams ability to sustain drives and win games.

  • Joseph Holley

    So, Chase, one thing I think you should clarify with the article–if a team is within -0.5 and 0.5 on the far right hand column, their passing game is doing fine. I don’t think that the Chiefs, my Saints, the Pats, etc. are doing anything “wrong” with their passing games at all, at least this year. On at least 1st and 2nd down, and even some 3rd downs, a completion that gains ~5 yards, but not a first down, is better than an incompletion. I hope that part 3 of this series focuses on “Success Rate” for those completions. (I know that you are familiar with DVOA–which creates a really good comparison for completion %, because the QB is penalized for that 2 yard dumpoff on 3rd and 10.)

    • Good point, Joseph.

      There are a number of completed passes that don’t go for first downs but are still good. I’m not sure how much more I am going to take this, though, as I don’t want to go too far down the rabbit hole with a complicated stat. My main issue is when I see something like 3 yard pass, 2 yard pass, sack, and that a QB had a 100% completion rate on a 3-and-out, well, that drives me crazy. It should really bad viewed as bad play, bad play, bad play.

      • Joseph Holley

        Don’t know if you have a 3rd part to this series, or if there is a way to easily measure and make a table with said data, but a Successful Completions total and percentage would be enlightening comparison with these numbers. DVOA measures this somewhat too, and a Success % would probably look very similar to QB DVOA.

  • Deacon Drake

    More fuel to the “Kirk Cousins is woefully underrated by Redskins’ management”.

    This is really good stuff. The league needs more risk takers like Favre, Mike Martz, Chip Kelly. I think one of the reasons Norv Turner left Minnesota last year is that Sam Bradford could not do what he wanted, which is nuts, because even Alex Smith got the ball down the field under him.

    I get why inexperienced QBs checkdown… It is actually a valuable part of the learning experience, particularly in preseason: Go through the reads, both pre and post snap, watch the defense respond to the action, and deliver the ball. In the preseason, when you aren’t trying to win or subject yourself to injury, hit your three reads and dump it off, hope to move the chains. But once the bullets start flying, you have to be willing to chuck and duck.

    Trubisky took some heat from the booth for a “bad read” that killed the game, but the truth was that the play call and pattern were great, and he read that Miller had his guy beat for a significant gain. He just made a shitty throw… big difference. Much better than the slew of failed completions from Matt Cassel.

  • I never would’ve guessed Jacksonville would be where they are. In my mind, when Blake Bortles isn’t throwing checkdowns, he’s getting sacked or getting his passes tipped. Though he’s only been sacked 5 times this year. Jacksonville is also 31st in passing attempts, so he must be making the best of his limited opportunities.

    Also, if this continues, Jared Goff’s turnaround has to be one of the most remarkable we’ve ever seen. The Rams look “good” with this metric, Goff is tied with TOM BRADY for the most completions of 20+ yards (24), and he’s currently +4.95 in ANY/A over last year. That is remarkable. They’ve been a blast to watch, even with their struggles last week. Gotta do something about those unis though. Looks like their real ones got lost at the airport so they bought a bunch of stuff from a thrift store.

    • Actually looking at the Jags more closely, they’re more a case of a balanced attack: they suck both at throwing for first downs and completing passes. I guess Bortles is who I thought he was. I shouldn’t have let him off the hook.

      • Yeah, I think that’s right lol. Although FWIW, Bortles is picking up 1st downs at a slightly above average rate. So that’s …. something.

        • Josh Sanford

          Did you guys see the ESPN article today that reports Fournette was clocked at 22 MPH on his long TD? He is decidedly NOT the Blake Bortles of running backs.

    • Richie

      ” Gotta do something about those unis though. ”

      The blue/yellow they wore on Sunday against Seattle were spectacular. I wish that is what they were going to switch to in 2019.

      This hideous uniform limbo for the 2016-2018 Rams is one of the stupidest, foot-shooting ideas in NFL history.

      • LightsOut85

        Agreed 100%. They are representative of great older NFL uniforms (when they were better on average, IMO). Colors that “pop”, lots of contrast (& no need for outlines on the numbers). That’s why I’ve loved a number of the Color-Rush uniforms; bold colors & contrast.

        • The Rams actually had really cool color rush uniforms too. They were reminiscent of the 50s Uniforms from the Norm Van Brocklin/Tom Fears days. The ones the wore Sunday are awesome (they should’ve never switched from these in 2000 IMO). They have good options. Pairing practice pants, with the St Louis Rams tops, and the 60s Roman Gabriel white helmet isn’t one of them.

      • Tom

        I don’t understand it…I don’t really pay too much attention to the uniform thing because I feel like everytime I watch a game, it’s some uniform that I don’t recognize (“old-timey”, TNF, away, home, etc.). Are they sticking with the old school white/blue helmet, but then blue/gold uniform?

        Yeah, the blue/yellow in Sunday was perfect…just freaking do that and be done with it. One of the reasons, I love the Steelers, is that – apart from that freakish bumblebee thing they make them wear – their uniform is always the same.

        • Richie

          Yeah. Most of the historically well-run franchises (in all sports) don’t mess with their uniforms much.

          On a side note: I cover high school football games for the newspaper. There is an insane trend with some teams to have the numbers be the same (or very similar) color to the jerseys, with a contrasting outline color. You know how hard it is to figure out a player’s number when he’s 30+ yards away and both his jersey and numbers are WHITE?!?!?

          • Tom

            First of all, that’s awesome that you cover high school football. I’ve watched a few CIF games, don’t know what divisions, but the teams are Bosco, Mater Dei, St. Francis, etc., lots of fun.
            I have no idea why the hell a team would make the numbers the same color as the jersey…I would think that would look pretty lame.

            • Richie

              I actually found a picture somebody took at the game. I can even see myself on the opposite sideline.

              https://vi.hudl.com/ot1/ysp/74815bb2-a8a7-4e7a-bd31-ebf605a0bce3/ds3dcmjftp97qhlu-3000_Thumbnail.jpg

              Both teams had same-color numbers.

              https://static.hudl.com/users/prod/8223457_b37313c906614bb2830e40e7ef7641a6.jpg

              • Richie

                Oops. Sorry to hijack this thread.

                • Josh Sanford

                  my woeful Arkansas Razorbacks got a win this season against some sad-sack team that we paid to come play us, and that team wore the white-on-white jerseys that you describe. It’s a violation of the NCAA uniform rules. And — no cheating –what do you think the penalty is? NO CHEATING!!

                  • Richie

                    I am very excited that there is an NCAA rule against that. Now if only my local high schools would follow.

                    What is the penalty for an illegal uniform? It probably SHOULD be a forfeit, just to make sure nobody ever does it. But if that was the penalty, I probably would have heard about this. So, I will guess that there is a fine. How about a $5,000 fine?

                    • Josh Sanford

                      The all-white team got charged with an expired time-out at the beginning of each quarter of the game.

                    • Richie

                      I wonder what would have happened if they used both of their remaining time outs in the first or third quarter.

    • Richie

      “Also, if this continues, Jared Goff’s turnaround has to be one of the most remarkable we’ve ever seen.”

      Goff had an ANY/A+ of 55 last year. That was the third-worst of all time amongst rookie QB’s with at least 150 pass attempts, ahead of Ryan Lindley (50) and Alex Smith (49). (DeShone Kizer is making a run this year, with a 57.)

      Smith’s first season with an ANY/A+ of 100 or better was in 2011, his 7th season. (Did Alex Smith miss the entire 2008 season? How do I have no recollection of that?) Smith had a 92 his 2nd season.

      Terry Bradshaw had a 65. He got to 100 by his 3rd season.

      Donovan McNabb had a 69. He was at 95 in year 2 and 103 in year 3.

      Aikman had a 72, and got to 100 in year 3.

      So, if Goff can keep his ANY/A+ above 100, it would be the first rookie QB to have a bad rookie season, and be this good his second season. http://pfref.com/tiny/HLw8t

      He made some mistakes on Sunday for sure. But he was playing against Seattle, who still has a good defense. But Goff still looked decent between the mistakes. The pass that Kupp dropped in the end zone wasn’t an easy catch, but was catchable. That’s how close Goff came to having a game-winning drive.

      • Tom

        Right…watched the whole game; that pass to Kupp was a bit ahead of him, but still catchable…I’d say both Goff and Kupp can be “blamed” for that. In any event, it certainly wasn’t a bad pass at all, and he was pretty darn close to getting a GWD. And his turnaround, if it continues as Eric noted, will be remarkable.

  • Of course, the fact that Peyton Manning is the co-GOAT at avoiding sacks and probably one of the best QBs of all time in era-adjusted first downs per completion is the reason why he’s underrated even by his ridiculous conventional statistics. If “era-adjusted first downs per dropback” was a real thing, he’d be the likely leader (outside of pre-modern guys like Luckman and Baugh where the era-adjustment is doing too much lifting). He’s definitely the best QB of the last 20 years or so in first downs per dropback.

    • Richie

      ” one of the best QBs of all time in era-adjusted first downs per completion”

      Where are you finding this ranking?

      • mrh

        I think by starting your quotation after the word “probably” you have changed the meaning of the original.

        • Richie

          Not really. I’m curious about where the era-adjusted first downs per completion stat is coming from.

          I’m not disputing it. I’ve just don’t remember seeing it anywhere.

          • Adam

            I wrote an article about first down kings during the offseason:

            http://www.thegridfe.com/2017/04/12/passing-first-downs-career/

          • I hadn’t seen Adam’s piece this offseason, though that serves as good confirmation. I’ve discussed it with Adam and Bryan before, though; we’re all pretty big on the value of first downs as a statistic. Given what I know about how much Peyton dominated the era for which we have first-down data, I find it hard to believe very many quarterbacks in history could have plausibly dominated their era to a similar degree.

  • LightsOut85

    I love comparisons like this, that highlight how much today’s passing game revolves around short passes (that inflate the perception of success/quality).

    What I was surprised to learn however, was that so far McCown is 4th in comp% on deep (16+) passes. However, he passes deep 2nd least often (of 32), only about 10% of the time. He’s still 3rd (comp%) when looking at all QB’s short passes, but then again, when you count all attempts within 15 yards there’s no way to distinguish between 11-15 heavy passers, and 5&under passers with play-by-play alone. I’d have to imagine there’s big comp% differences there.

  • Pingback: Passing completions and first downs | Get The Picture()