Completed Air Yards Over Expectation - Fantasy Football Risers, Fallers

If you play fantasy football, you know your great deal of statistics and metrics. At the end of the day, we need some way to measure production. While the basic numbers (targets, receptions, yards,...) are often enough to give us a solid idea of what we're dealing with, the truth is that the deeper we dig, the more information we get, and the more data points we have to make sound decisions that help our teams.

At this point in time, the NFL analytics movement is going forward at a good pace and through the last few years, we have seen concepts make it to the surface and become mainstream when assessing players. Enter Air Yards, and "stats over expectation". For example, the very own NFL is publishing data related to how quarterbacks are completing passes over or under expectation over NextGenStats, that is, if they are connecting with receivers in plays they shouldn't (based on historical data) or the opposite. The same goes for rushers and the yards they gain, or receivers and the yards after catch they are able to rack up.

One metric that doesn't appear in NGS, though, is Completed Air Yards Over Expectation. Today, I'll be writing about it and how the field of NFL players have feared in that metric through two weeks of games and leading up to this Sunday's slate of games. Let's get to it.

• Quarterback fantasy football rankings
• Running back fantasy football rankings
• Wide receiver fantasy football rankings
• Tight end fantasy football rankings
• Kicker fantasy football rankings
• FLEX fantasy football rankings
• Defense (D/ST) fantasy football rankings
• Superflex fantasy football rankings
• IDP fantasy football rankings
• Dynasty fantasy football rankings

Completed Air Yards Over Expectation (CAYOE) Model

Although the NFL has made public multiple "over-expectation" metrics already, one that is not available but can be easily calculated is the one named in this section's headline: Completed Air Yards Over Expectation (CAYOE). I'm not using NGS data nor model to calculate those values, though, so you might find some differences if you make calculations yourself using those numbers.

As we know the completion probability of every pass thrown by quarterbacks (from zero percent to 100 percent), the air yards traveled by each of those passes, and whether they finished as being completed throws or incomplete ones, we can calculate the CAYOE of each play/receiver and come up with a total and an average per target easily.

This is the formula I'm using:

CAYOE = (IF (Complete Pass) THEN (AirYds) ELSE 0) - (IF (AirYds)>0 then COMP%*AirYds ELSE 0)

Basically, I look for air yards gained in a play (AirYds if the pass is completed, 0 if it is not), and subtract CMP%*AirYds from it. That means that for example, a pass of 10 AirYds with a 100% chance of being completed would yield the following formula:

CAYOE = 10 - (1.0*10) = 0

Meaning the receiver was 100% expected to make the catch, and did so, so he gained no CAYOE at all. If the pass only had a 50% chance of being completed and the receiver hauled it in, though, that'd mean:

CAYOE = 10 - (0.5*10) = 10 - 5 = 5

The receiver would be expected to make that catch 50% of the times he's being targeted with it, and as he completed it, he gained 5 CAYOE. Had he missed the reception, then it'd have been "0 - 5 = -5 CAYOE".

CAYOE Leaders/Trailers Through Week 2 (Individual Games)

First of all, here is the leaderboard (top-10 leaders and bottom-10 trailers) through two weeks of games, with the performances taken individually--meaning each game is considered separate from each other instead of all combined (we'll tackle that in later in the next section). I have set the minimum targets at 5 for each player/game and sorted the leaderboard by CAYOE/Target.

Some quick takeaways:

• Julian Edelman had an impossible Week 2 performance. He was targeted more than 10 times (it has happened only 28 times this season through two weeks of games) and racked up the most yards of every player in Weeks 1 and 2 with his 179 (16.2 per target). 86% of those yards came through the air, with just 25 after the catch.
• Edelman's 83 CAYOE almost doubled the second-highest mark of the first two weeks (Darius Slayton's 43), although on a per-target average the difference wasn't that large between Edelman (7.5) and Marquise Brown (6.5).
• Edelman's biggest bump in CAYOE came from this insane downfield completion (28% completion probability):

• Don't get it wrong, though: Edelman's 83 CAYOE had little to do with his season average as he posted a negative minus-8 (-1.1 CAYOE/Tgt) in Week 1. The 83-mark was definitely a weekend-winning stat, but it was also a large outlier.
• Minnesota had a great debut this season with Adam Thielen as its new WR1 after Diggs departure, and he played to the expected level--actually, over it. He was targeted 9 times, reached 112 yards virtually without needing YAC at all (4), and averaged 4.3 CAYOE/Tgt to make the top-10 so far this season. But look at what happened in Week 2: Thielen reached just under a quarter of his Week 1 yard-tally when he should have at least completed receptions for double his yards, and posted one of the bottom-10 CAYOE/Tgt of the first two weeks.
• Both Odell Beckham Jr. and A.J. Green were the only players to post CAYOE marks at-or-under minus-45. No matter what, they were targeted 10+ times each and wildly underperformed the expectations with less than 30 completed air yards while expected to reach at least 59 in each of those two games.
• Shame on Christian Kirk. The volume was low at just five targets, but he and his minus-3 completed air yards didn't help Kyler Murray at all while he should have racked up 26 for (at least) an average of 5.0 yards per target without taking potential YAC into account. His worst play in terms of CAYOE was a ball that he was so close to catching but ultimately couldn't:

CAYOE Leaders/Trailers Through Week 2 (Season Totals)

With the individual games covered, here is the season-long leaderboard (top-10 leaders and bottom-10 trailers) through two weeks of games. I have set the minimum targets at 10 for each player and sorted the leaderboard by CAYOE/Target over the year.

Some quick takeaways:

• As I told you above, Edelman's Week 1 was far worse (he posted a negative CAYOE of minus-8) than his Week 2, but even with that he still leads the field through two weeks of play for an average of 4.2 CAYOE/Tgt. Brown also retains his no. 2 spot, in this case with steadier performances of 39 and 6 CAYOE respectively.
• While Stefon Diggs has been targeted the fifth-most times this season (22), he ranks fifth overall in CAYOE/Tgt after hauling in passes for a combined 67 air yards over expectation. That raw mark is the second-largest only behind Edelman, but it's much healthier as Diggs had 27 CAYOE in W1 and 40 in W2 compared to the much-more-volatile Edelman numbers.
• This has been Diggs' best play so far, good for all of 24 CAYOE in W2:

• Atlanta's Calvin Ridley is the only player to have accrued more than 200 completed air yards through Week 2 games. He's 13 yards over second-best Stefon Diggs, although he's averaged 0.5 CAYOE/Tgt less than the latter and both are tied with 239 receiving yards.
• If we go by expected-CAY, then the best receivers of the season should actually be D.J. Moore (153 eCAY) and A.J. Green (148). While Moore has produced at least to those expectations (160 CAY, 0.3 CAYOE/Tgt), Green has been absolutely atrocious with just 67 CAY for a minus-3.7 CAYOE/Tgt.
• No wonder why Green is the worst receiver so far, no matter the angle you look at him from:

• Although Corey Davis missed on the top-10, he's got the 15th-best CAYOE mark of the year so far and he's converted the most targets in first downs (77%). Only Cooper Kupp (73%) and Calvin Ridley (70%) are at-or-above 70% on the season while targeted at least 10 times.
• One of the most talked-about players during the first couple of weeks of the season has been TE Logan Thomas, from Washington. He's been good for TE18 through Week 2, but he's been far from efficient completing just 36 air yards of an expected 73 (minus-2.2 CAYOE/Tgt) and still tied with Terry McLaurin (another underperformed) in targets (17).
• This pass-incomplete combination between QB Dwayne Haskins Jr. and TE Logan Thomas has been the most costly CAYOE-wise from the tight end (minus-9.7 CAYOE, 70% completion probability):

Get ready for another Sunday of action, packed-full of interesting matchups. With three matches in the bag when all is said and done after today's (and tomorrow's) slate of games, we should start to see trends solidifying and outliers getting wiped out of the map. Here's to hope Julian Edelman keeps it up and A.J. Green rebounds to his years-prior self and helps Joe Burrow at least a bit more than he has so far this season.

More Fantasy Football Analysis