East Coast Offense: Don't Be Too Attached to Your Metrics

This article is part of our East Coast Offense series.

Don't Get Too Attached to Particular Metrics

The yards-per-attempt (YPA) stat for quarterbacks is a pretty good one. It allows us to evaluate passers on their efficiency rather than their total output which is often a consequence of volume. For example, in 2011, Matthew Stafford passed for 5,038 yards (eighth all time) and 41 touchdowns (T-9th), but his YPA was only 7.6. While that's above average (7.2 in 2011), it's nothing special - certainly not the mark of a quarterback putting up all-time totals. Stafford was merely an above average QB who led the league with 663 passing attempts.

As such, you could only expect him to continue being a top fantasy QB if he continued to lead the league in attempts. (In fact, he set the all-time record for attempts in 2012 with 727, but his efficiency was so poor (6.8 YPA) his numbers declined precipitously.) Since then, Stafford's been about league average, and his attempts have gone down every year - and his fantasy value along with them. So you can see how YPA - a rough measurement of a quarterback's efficiency - is a useful metric. For that reason, I tend to apply it widely when evaluating quarterbacks.

One quarterback, of whom I wanted no part this year was Derek Carr. Carr's YPA last year was 5.5, an abysmal number even for a rookie on a bad team. Consider that Carr's Raiders were 32nd in YPA, and the 31st-ranked team, the Blake Bortles/Chad Henne Jaguars, were at 6.2. Put differently, there was a greater gap between the No. 31 Jaguars and No. 32 Raiders than between the No. 6 Colts (7.7) and No. 19 Texans (7.1.) Carr was an outlier in efficiency, and he wasn't the only player or rookie in a bad environment.

Here are the YPAs of rookie QBs (minimum 250 pass attempts) over the last seven years:

The quarterbacks with less than 6.4 YPA include Stafford, Bradford, Clausen, Gabbert, Bortles and Carr. It's not good company, and Bradford and Stafford were only afforded so many chances to get better because of their elite pedigrees and status as top overall picks.

So you can see why I was not on Carr or his top receiver Amari Cooper this year. But through seven games this season Carr's YPA is a robust 7.6, and he's thrown only three picks (something he was admittedly good at avoiding last year too.) Where did I go wrong?

There's a limit to this type of analysis, where we categorize players by a certain metric or scouting description, look at how similarly categorized players performed and assume the player in question will follow suit. What's true generally is often false in a particular case. There is no law of physics that says QBs with bad rookie YPAs won't improve, perhaps significantly, even if it hasn't happened often in the recent past.

Moreover, many of these correlations are already priced in. While this particular one between rookie performance and career success isn't talked about much, many people, for example, target bigger wide receivers (whose size had been correlated with increased fantasy production), so there was value to be had with Jarvis Landry, Julian Edelman, Emmanuel Sanders, Antonio Brown and T.Y. Hilton the last few years.

The bottom line, it's useful to categorize players and evaluate them as a group, and use the group's performance with its larger sample to predict the fates of its particular members. But don't get too attached to the categorization and always keep an eye out for distinguishing characteristics that might portend a divergent path.

In Carr's case, it might have been good judgment with the ball, a franchise that had nowhere to go but up and a receiving corps that improved massively from 2014 to this season. Of course, it's only been six games, and it might turn out Carr was horrible after all.

End of Half Kneel-downs

At the end of the first half Sunday night, the Panthers got the ball with two seconds left at their own 29 yard line and took a knee. Why do teams do that? Do they think the defense has a better chance to score on one play than they do? At what yard line does the defense (who doesn't even control the ball) have the advantage? I'd have to think it would be about the 15. If that were true, shouldn't teams always run a play anytime they're not deep in their own end?

I realize this is but a small subset of the pathological risk-aversion in today's NFL, and the chances of either team scoring are incredibly slim. But if you have any kind of edge, you should run a play. Moreover, I bet a savvy team could line up like it was going to kneel down and catch a defense off guard. Of course, you could only do that once.

Week 7 Observations

• Every first-round rookie receiver besides Amari Cooper has been derailed by injuries this year and is virtually worthless. DeVante Parker (healthy now, but preseason injury prevented him from moving up the depth chart), Kevin White, Nelson Agholor, Breshad Perriman and Phillip Dorsett have combined for 321 receiving yards so far, and all are droppable in most formats.

•Eli Manning had a terrible fantasy day, but it wasn't entirely his fault. Rueben Randle dropped a would-be big play, and defensive and kick return TDs killed the game flow.

• Now that Orleans Darkwa (Let's call him "The Dark Horse") is in the mix, the Giants have a four-headed monster at running back. To illustrate how bad it is, all four saw at least four carries and none saw more than eight.

• Those Domino's commercials where they show you all the ways you can order it make no sense. It's still low-quality pizza no matter what device you use. I'd rather order real pizza by fax machine than Domino's by telepathy.

•Darren McFadden looked like the guy from five years ago. I have no idea how long that'll last, but I'd bust my FAAB budget on him. Christine Michael showed a little burst, but it looks like McFadden's job until further notice. Stick a fork in Joseph Randle. I don't know who Lucky Whitehead is.

• The Giants don't use Odell Beckham enough, and this explanation from offensive coordinator Ben McAdoo last week was not especially satisfying.

• The Giants defense is bad.

•Keenan Allen was robbed of a TD late in the game when the refs called a phantom offensive PI penalty. The Chargers eventually scored anyway, but it was Danny Woodhead who got a few extra catches and a TD instead. As it turns out, I lost in the DFS Crossover League because of it. (I had Allen, and Brad Evans had Woodhead.)

• What's with the Chargers? They're knocking on the door for overtime at Lambeau Field last week, and they get annihilated at home by the Raiders? At least Melvin Gordon learned his fumbling lesson.

• The Eagles plan to wear down the Panthers with their pace seemed to be working until Bradford made some bad throws, and his receivers kept dropping passes. The idea of the Eagles is great, but the reality is so uneven and flawed. I keep waiting for them to get it together, but through seven weeks it hasn't happened. Even in their blowout win against the Giants Bradford was terrible.

• Two years ago who would have thought three of the leading rushers for Week 7 of 2016 would be McFadden, Chris Johnson and Jonathan Stewart?

• It looks like the bye fixed whatever ailed Mike Evans. He could easily be a top-five overall receiver the rest of the way.

•Jordan Reed had 13 targets and two touchdowns. On the rare occasion he's healthy, Reed's firmly in the Travis Kelce-Tyler Eifert tier, i.e., about as good as anyone except Rob Gronkowski.

• For all the DFS talk I heard this week, no one mentioned stacking Ryan Tannehill and Jarvis Landry. There's usually one or two a week like that – some random quarterback who's not cheap enough to be a great value and not good enough or well-matched up enough to be a chalky play. Next week it could be Joe Flacco, Ryan Fitzpatrick or Matthew Stafford. It almost worth using a random-number generator to pick your QB and match him up with his best target.

• Tannehill amazingly completed 18 of 19 passing attempts. Lamar Miller had a huge day and is averaging 6.0 YPC this year after managing 5.1 last year. That's Jamaal Charles territory.

• It's too bad Arian Foster tore his Achilles' tendon, in garbage time no less. He's one of the great backs of his era, but he'll be 30 heading into next season coming off a severe injury. At least he had the decency to go down after compiling a huge day for you. So many players selfishly get hurt in the first quarter.

•Nate Washington was the Texans' primary garbage time star as DeAndre Hopkins didn't do much with his 12 targets.

•Todd Gurley predictably had a huge game against the Browns weak run defense, but what's especially encouraging is he caught four balls for 35 yards. He was a top-five back already, but if the receiving becomes a regular part of his game, he could be next year's No. 1 overall pick.

• Apparently, I'm the last one to realize how bad the Colts are this year. I thought they'd roll over the Saints and instead found themselves down 27-0 at home. They got it together in the fourth quarter, but it was too little too late. Chuck Pagano also punted away one chance on 4th-and-6 from the Saints 46, down 13 with nine minutes left in the game. They executed arguably the greatest onside kick of all time – it seemed to bounce 20 feet in the air – but couldn't corral it in the scrum. And their defense couldn't stop the Saints from sealing the game with a first down.

• What's going on with Matt Ryan and Julio Jones? They're almost as bad as Eli Manning and Beckham. Ever since Jones hurt his hamstring (apparently he's healthy now) teams have slowed the Falcons passing game down. While Jones went 9-92-1, he needed 17 targets to get there, a meager 5.4 YPT.

• The Patriots were life and death with the Jets at home all game, and suddenly they were up 10, closing in on the cover before Folk's miraculous kick. Brandon LaFell's return was a bust, as he couldn't hold onto the ball. Danny Amendola picked up the slack but Julian Edelman (who dropped an easy TD) won't draw Darrelle Revis every week.

•Tom Brady went for 355 yards, two TDs and no picks against arguably the leagues best defense. He wasn't especially efficient (6.6 YPA), but he's virtually matchup proof.