Air Yards 2011

Imagine that football were just invented and we had to figure out how to credit passers with all the yards they generate, and I said let's include all the yards that the receiver runs after the catch. I'd be laughed out of the room. But that's the convention we've been handed.

It's not that QBs shouldn't be credited at all, because if a pass isn't completed there can be no yards after catch at all. And the argument that a very accurate QB that can hit a receiver in stride or lead him to open space to create YAC can't be ignored. Scheme matters too. But it's evident that some QBs feast on the YAC-gaining abilities of their receivers much more than others, distorting their overall stats.

I came up with Air Yards a few years ago as a method to compare passers with their receiver's YAC removed. Air Yards is simply the complement of YAC. It's the yardage a pass travels through the air forward of the line of scrimmage. AY is a unique and interesting way to view QB performance, but it's not perfect. A QB would be penalized for completing a short screen that's caught a yard to two behind the line of scrimmage. But those plays are few and far between, and I don't lose any sleep worrying about them. We should also keep in mind that a QB's performance is never just his own. But AY might be an inch closer to isolating the individual QB's contribution than if we look at total passing yards alone.

Here are the leaders in AY for the 2011 regular season. The table is sorted by default according to Air Yards per Attempt (AirYPA), but you can re-sort the table by clicking on the column headers.



RankNameTot YdsAir YdsAY%AirYPA
1A.Rodgers4643248053%4.9
2C.Palmer2753159058%4.8
3E.Manning4933275856%4.7
4M.Moore2497159664%4.6
5D.Brees5476296554%4.5
6M.Vick3303185956%4.4
7B.Roethlisberger4077219954%4.3
8P.Rivers4624246453%4.2
9M.Ryan4177239257%4.2
10M.Schaub2479121449%4.2
11T.Romo4184216852%4.2
12D.Orlovsky120180167%4.2
13T.Brady5235253148%4.1
14J.Cutler2319126855%4.0
15C.Newton4051208752%4.0
16M.Stafford5038263952%4.0
17T.Tebow1729106462%3.9
18K.Kolb195599351%3.9
19J.Campbell117064555%3.9
20A.Dalton3398197058%3.8
21J.Skelton1913104955%3.8
22R.Grossman3151173055%3.8
23A.Smith3144166953%3.8
24K.Orton175894354%3.7
25T.Jackson3091167354%3.7
26M.Cassel171395656%3.6
27S.Bradford2164126058%3.5
28D.McNabb102654853%3.5
29J.Flacco3610186152%3.4
30M.Sanchez3474184753%3.4
31T.Yates94944747%3.3
32J.Beck85843851%3.3
33M.Hasselbeck3571171848%3.3
34C.Ponder185396552%3.3
35J.Freeman3592180650%3.3
36C.McCoy2733149455%3.2
37R.Fitzpatrick3832183648%3.2
38C.Painter154173648%3.0
39B.Gabbert2214103347%2.5

A few quick observations:

Rodgers beats Brees handily, 4.9 to 4.5 in terms of AirYPA, which would be a vote in favor of Rodgers for MVP.

Eli ranks third in AirYPA, which is largely why the Giants find themselves returning to the Super Bowl for the 2nd time in 5 years.

There are a few surprises at the top of the list, such as Matt Moore and Carson Palmer. Dan Orlovsky beat out Tom Brady, Matt Stafford, and many other notable passers.

Brady tied for third in the league for the highest percentage of YAC.

Look how low Flacco is--29th. 48% of his passing yards were YAC, undoubtedly many which came from Ray Rice. If I had Rice to throw to. Flacco is expecting a contract extension soon, and his 29th-ranked AirYPA number of 3.4 might come in handy for Ozzie Newsome at the negotiating table.

Ryan Fitzpatrick ranks 37th(!) with 3.2 AirYPA. Last season he ranked 23rd with 3.8, which isn't much better. I'm not sure the Bills are thrilled with the 6 yr/$60 million contract they gave him before the season, and maybe if they had noted how much of his performance was thanks to YAC, the offer may not have been so generous.

And what's a football blog post without a Tim Tebow mention? Tebow fares relatively well under the Air Yards light. He ranks 17th in AirYPA, and is 3rd in % Air Yards. This is because he rare throws dump-offs--he's his own check-down.

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

34 Responses to “Air Yards 2011”

  1. JoelW says:

    Obviously I think we intuit that this underrates Brady because he has such good YAC receivers and he finds them because that's the best offensive option, and that's there system. How do we adjust for that? It's very useful information to know how much the QB did in the air, I just want to think of a way to blend the information.

    I think it probably needs to be normalized for completion percentage. If I am completing 70% of my passes for 3 air yards and getting 3 YACs on average, isn't that better than completing 60% of my passes for 3.5 Air yards (same air yards per attempt) and getting 3 YACs on average? I think the better stat would be Air Yards/Attempt plus completion percentage x League Average YACs.

    It certainly makes me think more highly of Matt Moore.

  2. Keith Goldner says:

    Gotta love the top three AY% are Dan Orlovsky, Matt Moore, and Tebow.

    Thought there might be a strong correlation between the percentage of deep passes thrown and AY% (mostly because of Tebow) but it's only about .32. It is pulled down a ton by Orlovsky (and somewhat by Josh Freeman). If we take Orlovsky and his 8 games out of the picture, there is a larger, 0.56 correlation.

  3. Anonymous says:

    "Obviously I think we intuit that this underrates Brady because he has such good YAC receivers"

    Actually I intuit that perhaps most people's intuitive judgement of Brady is flawed. Looking at his career numbers, there appears to be a slight negative correlation between his AYPA and his team's performance.

  4. Anonymous says:

    "If I had Rice to throw to."

    I think you forgot to finish your thought there

  5. Dave Archibald says:

    "Looking at his career numbers, there appears to be a slight negative correlation between his AYPA and his team's performance."

    Are you using AYPA to stand for "Air Yards per Attempt?" It looks like it tracks logically for me; in 2006 when the Pats had just an OK offense Brady had 3.5 Air Yards/A (http://www.advancednflstats.com/2007/08/2006-qb-air-yards.html); in 2007 when the Pats had one of the great offenses in the history of the league, he had 4.9 Air Yards/A (http://www.advancednflstats.com/2008/07/brett-favre-is-overrated.html).

  6. Whispers says:

    "Rodgers beats Brees handily, 4.9 to 4.5 in terms of AirYPA, which would be a vote in favor of Rodgers for MVP."


    "There are a few surprises at the top of the list, such as Matt Moore and Carson Palmer. Dan Orlovsky beat out Tom Brady, Matt Stafford, and many other notable passers."

    Are you seeing the flaw in your logic?

  7. Whispers says:

    "Actually I intuit that perhaps most people's intuitive judgement of Brady is flawed. Looking at his career numbers, there appears to be a slight negative correlation between his AYPA and his team's performance."

    I've yet to see an argument that explains why we should care about AYPA at all.

  8. James says:

    I don't think it's a question of over- or under-rating players, I think it's a measure of the QB's system.

    AY% is definitely more dependent upon the skill of the receivers. Just look at Aaron Rodgers - he clearly bested Drew Brees in AirYPA, but because Rodgers' receivers ALSO had so much YAC their AY% are nearly the same. That underrates Rodgers' contribution, but that's not the primary stat here.

    To Joel's point about completion percentage, comp% is already a part of AirYPA the same way comp% is built into YPA:

    AirYPA = Air Yards / Attempt = [(Air Yards / Completion) * (Completion / Attempt)]

    More specifically, the answer to your example is the passing game with 70% comp% is better... but NOT because of air yards. 70% of 3 AirYards/comp is the same as 60% of 3.5 AirYards/comp because .7 * 3 = .6 * 3.5 = 2.1 AirYPA.

    The 70% team is better because the receivers catch the ball more often and therefore contribute their 3 yards of YAC on more plays. So all else equal the 70% team is better, but not because of the QB.

    HOWEVER, not all else is equal - shorter passes lead to more YAC. Therefore, you can assume that the receivers on the 60% team are better because they are getting the same amount of YAC further downfield where it's harder to gain YAC. Ultimately this seems like an unlikely scenario.

    That's why I look at air yards as an analysis of the system - the higher the team AirYPA, the more the team relies on the QB throwing downfield. The lower the AirYPA, the more the team relies on it's receivers to make plays. I'm not going to fault Brady for using his high accuracy on short passes to maximize his receivers' YAC ability.

    On the other hand, you can look at Colt McCoy, who is throwing some of the shortest passes in the league and yet has an extremely high AY%. Clearly his receivers are terrible and not contributing any YAC; contrast with Hasselbeck and Fitzpatrick who throw similarly short passes but benefit from Nate Washington and Fred Jackson.

  9. Anonymous says:

    Ok, so can pundits stop talking about how Rodgers is getting carried by his receivers, while Brady is playing with scrubs now? You can't have a successful dink and dunk offense without receivers being good after the catch. Just ask Cassel, Bradford and McCoy.

  10. Tay says:

    One way of looking at Moore's figures is that he's getting hardly any yac for his completed air yards,either though his receivers not breaking free or him not leading them into open field.

    A community post looked at the relationship between yac and air yards here

    http://community.advancednflstats.com/2011/01/yac-may-be-quarterbacking-skill.html

    and if you plug Moore's numbers in,the average QB throwing his amount of air yards should be getting about 4.5 yards of yac.He's getting half a yard less.

    Manning,by contrast who throws about the same air yards as Moore is getting almost a yard and a half more yac than you'd expect from an average QB/WRs combination.

    Brady similarly out runs the average expected yac for his air yards by around a yard and a half per attempt.

    Should make for an interesting game on Sunday.

  11. MattieShoes says:

    AYPA is often adjusted yards per attempt, which folds touchdowns and interceptions into the normal Y/A numbers.

  12. Joshua Northey says:

    Which is why he refereed to it as AirYPA...

  13. Alex says:

    ESPN doesn't have passes broken out by QB-to-receiver, so this is team total and mixes in some of Chase Daniel and Matt Flynn. But...

    NO attempted 197 passes to Sproles, Thomas, Collins, and Ingram in the regular season for 1231 yards. According to ESPN, 1265 of those were yards after the catch. Green Bay attempted 91 passes to Grant, Starks, Kuhn, and Saine for 630 yards, with 702 being YAC. In the rest of their passes, GB had 4531 yards against 1646 YAC on 461 attempts and NO had 4274 yards with 1181 YAC on 465 attempts.

    If AirYPA is just (total-YAC)/attempts, on non-RB throws GB had 6.26 AirYPA and NO had 6.65 AirYPA. Perhaps Brees isn't so far behind Rodgers?

  14. Ty Will says:

    Brian:

    Inspired by your Air Yards statistic, I came up with a statistic I call "QB Yard Credits per Snap", which is essentially Air Yards + Rushing Yards - Sack Yards - 30(Turnovers) / Snaps Played. It is meant to identify the number of yards that one can credit directly to the Quarterback, taking into account any yards he may have cost them by taking a sack or turning the ball over.

    Yesterday, using Yahoo Sports YAC statistics, I calculated the QBYC for the post 1998 careers of Peyton Manning, Tom Brady, Aaron Rodgers, and Brett Favre.

    The NFL average this season for all QBs was 1.32 QBYC per snap. In 4864 career snaps, Aaron Rodgers a QBYC average of 2.08 yards. Peyton Manning has a QBYC of 1.98 yards in 12752 snaps, Tom Brady has a QBYC average of 1.53 yards in 8891 snaps, and Brett Favre has a QBYC average of 1.31 yards in 12524 post-1998 snaps.

    So, by this measure, Aaron Rodgers has been the most valuable of the 4, and this holds for the postseason also. Rodgers QBYC actually improves in postseason to 2.18 yards in 375 snaps, Manning is again second with a QBYC of 1.82 in 1096 snaps, Brady is again third with a shockingly low QBYC of 1.29 in 1055 snaps, and Favre again brings up the rear with a 1.22 in 513 post-1998 postseason snaps.

    I want to calculate Brees and Eli Manning in the near future.

  15. TheLegend says:

    Shenanigans. The only way I see to learn anything about a QB from this chart is sorting by AY% from lowest to highest.

    I'd argue that the difference between NFL quarterbacks has very little to do with how far or how often they can huck it down the field, and much more to do with their ability to hit a receiver in stride, throw a receiver open, etc.

    My experience as a player and a fan leads me to believe that a QBs ability to throw precision passes has as more or more to do with the amount of YAC gained as any other factor. I'd be curious to see how Steve Young, Kurt Warner or other great QBs from elite 90s style west coast offenses would stack up here.

  16. TheLegend says:

    I can't help but notice that the top 9 on this list play in very aggressive vertical offenses... I'm not really sure what this is supposed to tell us about a QB others than how far he can huck it and how often he's asked to.

  17. Tarr says:

    Interesting to see Orlovsky and Painter on opposite ends of the scale, with the same system and the same personnel.

    I've had the arguments on the community end of the site and I don't have the time to re-hash here, but I do put a lot of stock in this stat.

  18. Anonymous says:

    @thelegend

    "My experience as a player and a fan leads me to believe that a QBs ability to throw precision passes has as more or more to do with the amount of YAC gained as any other factor."

    And this is why "your experience as a fan" is not a valid argument. First time visiting the site?

  19. Anonymous says:

    Interesting stat for discussion, but it is much like yards per completion. 40 years ago the yards per completion were phenomenal because of the style of football played then. This stat is more of an indicator of style than QB play. Brady throws quite a few screens. Brees throws a more equal combination. Not many throw it down field while the pocket is collapsing. A break down of completion % (5, 15, 25, and 40 yards)is far more appropriate to compare QB play.

  20. BP says:

    AirYards% is probably mostly a scheme-dependent stat, but it could be very useful if it is consistent for a QB playing in the same scheme from year-to-year. We could then expect a QB with a stretch of unusually low AirYards% to regress to the mean in future games. This was probably the case for Curtis Painter, whose early-season passing stats were inflated by huge YAC before returning to normal.

  21. Joseph says:

    I think one thing this stat allows us to do is sort out part-time QB's and younger QB's. In other words, the average seems to be between 52% & 56%; so if a QB is outside of this range, either he throws a lot of bombs (if he's above) or his receivers stink. If he's below, his stats are probably inflated by speedy receiver X breaking a slant for an 80yd TD.
    Another good idea was posted by "Alex" comparing Rodgers with Brees. If you separate out the RB's, you can see how much the QB throws the ball down the field, and who is better at it. Rare is the throw to the RB who is further than ~5yds downfield.
    Another interesting idea, if you have the data to do it with--cap all YAC at 10 yds, then sort by passing yds. In other words, 10 yds after he caught the ball, anything else gained is probably 100% result of a receiver's speed and/or breaking tackles. But those first few YAC have something to do with the QB's ability to hit him in stride or on time, where he can lunge forward for a couple of yards.
    In other words, it's not the QB's fault that his receivers can break tackles (Gronk), have sprinter speed (Cruz, Julio Jones), or have good moves (Sproles). Nor is it his fault that he threw a ~10 yd pass on his 20 that was taken to the house, instead of completing "normal" passes down to the opponent's 20 when his receiver then did the same thing as before--there just wasn't any more field to accumulate YAC.
    I think this is a good stat, but it doesn't tell us much.

  22. Dave Archibald says:

    The thinking here makes sense, and I'm sure there's some value to be gleaned from this stat. But part of Air Yards is elective, too. If you have good YAC receivers, it might be a smarter thing to do to dump the ball off to Ray Rice than to try to squeeze it through a narrow window 15 yards downfield. The preference to do the former does not necessarily convey an inability to do the latter.

    Is this stat available for the 2009 and 2010 seasons also? I see it for '06, '07, '08, and '11, but not those two years.

  23. TheLegend says:

    If you're gonna take a shot at somebody, at least do it in the form of a complete thought (and preferably not anonymously). But to answer your question, no, this is not my first time on the site.

    I'm just saying that I don't get what we're supposed to learn from this... which seems to be the same thing a bunch of other people are saying...

    If my opinion is so invalid, I'd be perfectly happy to have someone explain why and what it is that I'm missing about this set of stats.

  24. Tay says:

    Just a quick comparison with football or soccer if you like.Every player gets a chance to pass the ball and they have fairly similar associated stats such as completion rates.But the players who attract a premium in the transfer market are the ones who can consistently release their colleagues into space.It's a two way process,the attacker makes a run,the passer spots him from all the other passing options and completes the pass so it reaches it's intended target.

    I'd be surprised if something similar wasn't going on with yac,quarterbacks and receivers.

  25. Anonymous says:

    TheLegend,

    If you actually read the articles here, you'd know that the year-to-year correlation of receivers' YAC numbers are much higher than QBs', which are tiny. It clearly shows that QBs don't control YAC.

  26. James says:

    I think I have to agree with Anon here that as much as we would expect YAC to be partially or completely dependent on the QB, it seems to be almost entirely a result of the receiver.

    It may be true that a more accurate QB leads to more YAC, but the signal is drowned out because that only has an affect on a small number of plays. After all, how often is a receiver so open that he just needs a clean catch and not a broken tackle or to outrun a defender before making a big gain?

  27. Unknown says:

    Expanding on "The Legend"'s thoughts, I don't see why this statistic is useful. Air Yards appears to simply measure 2 things: 1) the propensity of a QB to throw deep as opposed to short/intermediate; and 2) the accuracy of a QB's deep balls.

    What air yards seems to ignore completely is a QB (or scheme)'s propensity to throw short/intermediate passes, and his accuracy on those short/intermediate passes. After all, a lot of YAC is about the QB's ability to hit the receiver in stride, or to throw the ball to a space where the WR can make YAC (something Brady and Eli seem to be exceptionally good at).

    Arguably, this stat completely ignores that fact, which is why it is at such odds with all other conventional and advanced statistical measures of QB performance.

  28. Brian Burke says:

    Dave, that is one of the most transparently poor comments I've read in a long time. Congratulations.

  29. Semifaded says:

    http://bayaniflores.blogspot.com/2011/11/overrated-tom-brady-and-giselle.html

  30. Anonymous says:

    I think this just indicates that Brady doesn't go up field as much and may be more of a systems QB than ESPN may like us to think. It may also suggest that Welker's ability to gain yards after a catch, and Gronk's ability to run someone over and get 10 more yards are undervalued.

    Brady is extremely accurate on short stuff, and the system has been designed to suit his strengths (as offenses should be) but he may not be as great a pure passer as he gets credit for.

  31. Ben says:

    Air yards do seem like a more accurate way to measure QBs' skill & true contribution than YPA.

    However, I'm a little surprised by where Tom Brady ranks.

    It is possible that he has the actual ability to throw more air yards per pass, but knowing that his receivers, while not exceptionally speedy, are great at churning out YAC (especially Welker), he intentionally holds back, throwing more high-percentage dart passes and letting his receivers get the yards. Strategic reasons.

    Going further, QBs' air yards could be affected by the type of wide receivers they throw to (e.g. deep threats vs. possession-type receivers.. which could be why Eli has such high air yards- all of his top 3 WRs belong to the former category).

    I'd say a good way to find out more is comparing QBs' air yards when they or their top receivers change teams.

    Brian, what do you think?

  32. Anonymous says:

    QBs can contribute to YAC in two ways: by making good decisions about which receiver to throw to, and by throwing accurate passes that can be caught in stride. So completely disregarding YAC in their stats may be as unfair as including it.

  33. Anonymous says:

    Ty for the stats and it really let's you see who is how good, well gives an idea I know there are a lot of varibles but I love the stats keep them coming.

    How bout those cowboys

  34. Happ says:

    I would like to see a stat that shows a QBs-AY in a direct ratio to his attempts.
    i.e.
    minus 5 to 0 AY = 7% of total attempts
    ...........0 to 10 AY = 43% of total attempts
    .........11 to 20 AY =28% of total attempts
    .........21 to 30 AY =16% of total attempts
    .........31 to 40 AY = 4% of total attempts
    .........41 to 50 AY = 2% of total attempts
    .........51 and over AY=0% of total attempts
    then we could see who is airing out the ball instead of dumping it off

Leave a Reply

Note: Only a member of this blog may post a comment.