What key stats say about NFL season's second-half expectations

Woodson says Giants have been most unlucky team (1:37)

Darren Woodson explains that the Giants have been the most unlucky team in the NFL as injuries have plagued their season. (1:37)

Nine weeks can tell us a lot about the NFL season. It doesn't seem like a lot, but nine weeks were enough for us to take Jared Goff for real as a franchise quarterback. They were enough to convince us that the Jaguars had a defense capable of getting into the heads of Ben Roethlisberger and A.J. Green. They were even enough to turn the Giants into a cautionary tale against believing in teams that forgot about fielding offensive lines.

Quantitatively, the halfway point of the NFL season gives us an opportune time to both look back and project forward. Eight or nine games don't sound like much, but when you consider that we've seen more than 1,000 total snaps from most NFL teams, the sample is bigger than it seems on the surface. Stats aren't going to tell the whole story, but by using the right stats, we can gain some meaningful insight into what has happened this season. More important, we might even be able to muster up an idea of what's going to happen next.

Let's look at some key statistics for the 2017 season and what they suggest about the remainder of the campaign. You can find a primer explaining many of these stats and how they apply to the 2017 season here.

Pythagorean expectation

Historically, a team's point differential has been a better indicator of its future win-loss record than its present win-loss record. The simplest and most effective way to project whether a team is likely to improve or decline in advance of a season is by comparing its actual win-loss record to its Pythagorean expectation. In July, the method identified the Eagles and Jaguars as teams likely to improve, while the Cowboys, Giants and Raiders were likely to decline. Each of those five will either top last season's win or loss total with a half-season to go.

We also can use the Pythagorean expectation to project games over a half-season, although it's more subject to sheer variance over a smaller sample, in the same way that a baseball player is more likely to hit .400 over a week than he would over a season. Teams that have outperformed their Pythagorean expectation in the first half of the season, all else being equal, would be likely to decline during the second half. Here are the five biggest overachievers from the first half of the season:

The 2016 Dolphins also grossly outperformed their point differential, winning 10 games despite being outscored by 17 points. The 2017 team has looked truly awful for stretches and pulled out some fortuitous narrow victories. The Dolphins beat the Chargers in Week 2 when they successfully iced a 44-yard Younghoe Koo field goal try. They got to face Matt Cassel as opposed to Marcus Mariota and won on a fumble return against the Titans. Even the games without luck have been a struggle: a late interception of Matt Ryan kept their game against the Falcons from heading to overtime, and it took a 17-point fourth-quarter comeback to beat the Jets. Miami is likely to improve on offense, but it's not a good football team by any stretch of the imagination.

The Eagles are on this list, but they still deserve to be considered as one of the top teams in the league. Philly is on pace to produce 13.4 Pythagorean wins, which would be the sixth-highest total since 1989. If the Eagles are on pace to finish 14-2 and their point differential says they're merely a 13-3 team, I don't think anyone should be shouting them out as frauds.

And of course, here are the five teams that have underperformed by point differential:

Pythagorean expectation projected both the 49ers and Browns to improve after underperforming their point differential a year ago, but they've combined for zero wins and continued to come up short. The good news, I suppose, is that they don't need to do much to deliver on their promise -- the Browns would need to win two games and the 49ers three to top their marks from a year ago.

The AFC South presents a pair of interesting candidates. The Jaguars have been one of the most surprising teams in the league, but their dominant victories -- Jacksonville's average win this season is by more than 24 points -- suggests they're playing nearly as well as the Eagles. With the Browns, Colts, Cardinals and 49ers still to come on their schedule, it would hardly be a surprise to see the Jags top 10 wins.

The Texans, meanwhile, deserved to win at least one of their incredible road performances against the Patriots and Seahawks, but I wouldn't count on them living up to their numbers in the absence of Deshaun Watson, who made Houston a different team altogether. The best way to overcome or fail to live up to your Pythagorean expectation is by upgrading or downgrading at quarterback, the one position that defies the numbers.

Biggest takeaway: The Dolphins and Cardinals aren't as good as their .500 records, while the Browns and 49ers are probably going to sneak in a win or two during the second half.

Record in close games

Closely correlated with Pythagorean expectation is a team's record in games decided by seven points or fewer, which is markedly inconsistent from year to year. Five teams won eight or more one-score games last season, as the Lions, Texans, Dolphins, Giants and Raiders went a combined 40-13. Those five franchises, however, are a combined 7-12 in such games this season.

Here are the teams that have played at least three one-touchdown games this season and been notably successful or unsuccessful in those contests:

  • 3-0: Cardinals, Washington

  • 4-1: Panthers, Dolphins

  • 3-1: Eagles, Patriots, Steelers, Titans

  • 0-3: Giants

  • 0-4: Browns

  • 0-5: 49ers

  • 1-3: Buccaneers, Lions, Texans

By sheer randomness, one or two of those teams that have been hot in close games will keep it up over the rest of the season. Most will not. (I'd bet on the Patriots, who have a habit of defying convention with Tom Brady at quarterback.) Meanwhile, the 49ers, Browns and Giants might seem hopelessly resigned to their fate, but keep in mind that the Bills, Browns, Cardinals and Eagles were a combined 0-14-1 in one-score games this time last season and went a far more respectable 6-8 the rest of the way.

Biggest takeaway: The Panthers and Washington are both in the NFC playoff picture, but they'll need to improve or else they'll lose ground when they struggle to maintain their performance in close games. The Giants (and the two winless teams) are still going to be bad, but probably not as hapless.

Strength of schedule

While there are far more effective ways of projecting strength of schedule than the old habit of using the previous year's winning percentage, we don't get to use the best possible measures of schedule strength until the season has actually begun. Who could have figured the Rams would be a roadblock that teams would want to avoid?

Football Outsiders tracks schedule strength via its DVOA metric and splits it into schedule strength both before and after the week in question. Here are the teams with the largest positive gaps between their first- and second-half strength of schedule, suggesting their schedule is likely to decline over the remainder of the season. These are their ranks in average quality of opponent from Weeks 1 through 9 and then from Weeks 10 through 17:

The Chiefs play just one team with a winning record over the rest of their season, a home game against the Bills on Nov. 26. The same is true for the Lions, who have to face the Vikings on Thanksgiving Day but otherwise enjoy an extremely friendly slate. We're also probably underestimating Washington, which already has played the Eagles twice to go along with the Chiefs, Cowboys, Rams and Seahawks. Washington faces the Vikings and Saints over the next two weeks, but outside of a trip to Dallas, its schedule is comfortable over the final six weeks of the season. Could it be a repeat of Washington's late-season run to the playoffs in 2015?

Here are the teams whose schedule stiffens up over the final eight weeks of the season:

The Jets have been a great story, but after playing the Bucs and hitting their bye, they'll have to face the Chiefs, Saints and Patriots over the final six weeks of their season. The Falcons don't exactly need bad news right about now, but they still have to travel to Seattle and New Orleans and have home games against the Cowboys, Panthers, Saints and Vikings to come. And the Cowboys might wish that the Ezekiel Elliott suspension actually took place earlier this season, given that they could be missing him for a run of games that includes matchups with the Seahawks and one of two games against the Eagles.

Biggest takeaway: Don't be surprised if the Chiefs and Jaguars go on winning streaks during the second half, or if the Jets and Colts fall to earth and start solidifying their positions in or around the top of the 2018 draft.

Red zone performance

Everyone knows that teams that perform well in the red zone win more frequently. What isn't quite as obvious, though, is that red zone performance isn't remotely consistent from year to year.

Take last season, when the Titans averaged 5.66 points per red zone trip on offense, which was best in the league and the third-highest rate of the previous 15 years. The same Tennessee offense this season is averaging 4.83 points per red zone possession, which is 13th in the NFL.

Or consider the surprising Jets, who were last in red zone point differential (offensive points per trip minus defensive points per trip) in 2014, first in 2015, and last again in 2016. This season, a limited Jets team is 15th -- essentially league average -- in red zone point differential.

Let's start with the ball and the best and worst red zone offenses through nine weeks. Keep in mind that these numbers count a touchdown as worth 6.95 points to account for the occasional missed extra point and two-pointer:

It wouldn't be surprising to see teams with great offensive lines like the Cowboys, Eagles and Raiders ranking toward the top of the red zone charts, right? The Cowboys were second in points scored per red zone trip last season, but the Raiders were 11th and an Eagles team missing Lane Johnson for most of the year was 22nd. I wouldn't count on the Dolphins or Packers ranking atop this leaderboard by the time the season ends.

On the flip side, the Browns are unbelievably bad in the red zone. More on them in a minute. The Cardinals aren't much better; since 2001, no team has averaged fewer than four points per red zone trip, which leaves the Cardinals in danger of becoming part of the historical record with Drew Stanton taking over at quarterback.

Here are the teams with the best and worst red zone defenses so far:

The Dolphins and Packers have given away most of the benefits from their offense performing wonders in the red zone; they should each be better on defense and worse on offense the rest of the way. The opposite is true for the Bengals. The Panthers have held opposing offenses to a league-low 17 red zone trips, which is even more impressive when you realize that they've played nine games already, but offenses permitted entry have scored touchdowns on 11 of those occasions.

Let's talk about the Browns, who have the league's worst red zone offense and the worst red zone defense. Hue Jackson's team has nine touchdowns and five turnovers in the red zone so far; for comparison, the rest of the league has 404 touchdowns against 36 turnovers. In those same situations, the Cleveland defense has allowed 17 touchdowns without forcing a single takeaway.

The Browns are being outscored by 2.25 points in the red zone. To put that in context, the next-worst team in the red zone since 2001 is the 2009 Rams, who were outscored by 1.87 points per trip, and they're the only team that topped a margin of 1.6 points. This isn't an annual thing, either; the Browns were 29th in red zone defense last season but a respectable 19th on offense. The 2016 team was outscored by 0.65 points per trip, which isn't memorably bad. The Browns should regress toward the mean and be much better inside the 20 on either side of the football, but they could still be the worst red zone team of this millennium.

Biggest takeaway: The Browns are significantly worse than any team in recent memory in the red zone, and it's almost impossible for them to be that bad over an entire season. The Eagles might struggle to keep up their impeccable red zone performance on offense, while the Panthers probably will improve upon the few times they actually let teams inside the 20.

Fumble luck

While forcing fumbles is a skill (and coughing up fumbles a weakness), no team has exhibited a consistent ability to recover a significantly high percentage of the fumbles in their games on an annual basis. Teams that recover a disproportionately high or low percentage of the footballs that hit the ground aren't going to be able to keep that up for any significant length of time, no matter how much you hear about how often they do fumble drills in practice or place an emphasis on running toward the ball.

Here are the teams with the highest and lowest percentage of fumble recoveries so far this season:

The Titans stand out as the enormous outlier, which is in part because they haven't seen many fumbles. Tennessee has fumbled a league-low four times on offense, and it has lost all four of them. Mike Mularkey's team is tied with the Cardinals with 15 fumble opportunities in their games. As Kevin Pelton noted on Twitter earlier this week in reference to Steph Curry, smaller sample sizes are more likely to create outliers.

The Steelers, who have the third-worst fumble recovery rate in the league this season, led the league by recovering 62.2 percent of fumbles last season despite possessing virtually the same roster. I'd be more concerned about the volume of fumbles if I were a fan of a particular team, which is where Washington stands out. Jay Gruden's offense has a league-high 20 fumbles, with no other team topping 17. Washington has lost 10 of those fumbles, which is also the most in football.

On defense, the Bills top the charts by forcing 18 fumbles, but they've recovered only six. For comparison, the Bears have stripped opposing ball carriers eight times and somehow managed to recover seven of those. Five of them have come on plays that started inside the Chicago 26-yard line, saving the Bears from what would likely have been field goal tries at a minimum.

Biggest takeaway: The Steelers were lucky with fumbles last season, but they're likely to start gobbling up more of those loose balls the rest of the way. As if they need the help. The Titans also will likely improve and create a few extra possessions for their offense. The opposite is true for the Seahawks and Cowboys.

Third-down performance

Teams that are particularly great on third and fourth down over any stretch of time often struggle to keep it up, especially if they're significantly better on third down than they are on first and second down. Take last season, when the only team in the league that was converting more than 50 percent of its third downs on offense through nine weeks was New Orleans. Over the remainder of the season, the Saints dropped down to a 43.7 percent success rate on third down; still good, but only sixth best in the league.

There's no team converting 50 percent of its third downs on offense this season, but a couple are close, thanks to this season's two breakout quarterbacks. Goff's Rams are up at 49.1 percent, while a recent slump has dropped Carson Wentz's Eagles under the 50-50 mark, down to 46.8 percent.

Wentz actually has been better as a passer on third down than Goff, posting an unreal 125.1 passer rating. Goff is a relative plebeian in second place at 113.6. As I wrote about when discussing Wentz's breakout last month, even the best quarterbacks of their respective generations have struggled to keep up those sort of phenomenal passer ratings on third down for any length of time. Wentz had a passer rating of 137.8 heading into Week 6; since then, he has posted a merely great rating of 101.0.

Goff and Wentz should still be excellent in the second half, but their dominance -- and by proxy, their teams' dominance -- on third down is likely to fall back to the pack. The Dolphins, on the other hand, are likely to improve after converting just 29.7 percent of their third downs during the first half of the campaign. Would you believe the 49ers and Browns are 30th and 31st in this category? You would? OK.

One surprising bit of good news for the Browns: They're actually 18th in the league in terms of stopping teams on third downs. The worst third-down defenses belong to the dismal pass defense of the Buccaneers, the surprisingly mediocre Packers defense, and the 49ers, who can't escape the doldrums.

On the flip side, the Broncos have the best third-down defense, which might not be sustainable given how they've slipped on first and second downs. Last season's leading defense by DVOA is sixth through nine weeks. The Vikings are the only other team in the league keeping teams at a conversion rate below 30 percent.

Biggest takeaway: Goff and Wentz are still going to be among the league's best quarterbacks, but they're probably not going to stay on the field for as many long drives.

Sack percentage

Let's finish up with this measure of defensive performance. Everyone knows about sacks, but the NFL also records quarterback knockdowns or hits, which includes any play in which a defender makes the passer hit the turf. History tells us that edge rushers will turn about 45 percent of their knockdowns into sacks and that players who have high knockdown totals and lower-than-expected sack numbers are likely to generate more sacks the following year. The opposite is also true.

We can track this on a team basis, too. So far this season, 42.6 percent of knockdowns around the league have resulted in sacks. Naturally, we can track this same measure of sack percentage for teams.

Teams that turn a particularly high percentage of their knockdowns into sacks might struggle to keep up their sack rate. The poster boys this season are the Chargers, who have 26 sacks on 43 knockdowns, good for a 60.5 percent sack percentage. The Jags (58.3 percent) are right there, although they're also tied for fourth in the league with 60 knockdowns, so they would get plenty of sacks by any measure. More concerning are the Broncos, who have a league-low 34 knockdowns after racking up 113 last season. Denver has generated 19 sacks on those 34 hits for a 55.9 percent sack percentage.

Alternately, there are teams that should get more pressure the rest of the way. Start with the Buccaneers, who desperately need the help. They are tied with the Broncos (and Giants) at 34 knockdowns, but they've generated only eight sacks. Tampa has just two sacks over the past three games. The 49ers quietly have one of the league's best defensive lines, as they're second in the league with 65 knockdowns, but those hits have generated only 17 sacks for a 26.2 percent sack percentage. The Jets are in third at 31.6 percent.

As for players, the biggest overachievers by sack rate include the league's breakout pass-rusher. DeMarcus Lawrence has 10.5 sacks on 17 hits, and the leaguewide 42.6 percent sack rate suggests he would typically have generated 7.2 sacks. The ensuing difference -- 3.3 sacks -- is the second highest in the NFL so far. Joey Bosa's 8.5 sacks on a mere 10 knockdowns has him 4.2 sacks ahead of expectations, while Mario Addison's 6.5 sacks on eight knockdowns make him the only other pass-rusher three sacks ahead of schedule. These guys will either have to improve their knockdown rate or expect to rack up fewer sacks in the second half.

Likewise, there are three players who are three sacks behind what their hit total would expect. Bucs interior disruptor Gerald McCoy has just two sacks, but his 13 knockdowns would project to 5.5 sacks, placing him 3.5 sacks below his projected rate. There's a huge gap between McCoy and the two young pass-rushers ahead of him. Niners end DeForest Buckner is quietly having a breakout season, as he has racked up 15 knockdowns, usually good for 6.4 sacks. Buckner has only 1.5 sacks, though, 4.9 sacks below our numbers. And even he can't top the bad luck of Leonard Williams, who has a half-sack on 13 knockdowns. At an even five sacks below projection, Williams could emerge as a superstar in the second half.

Biggest takeaway: Lawrence probably isn't going to threaten the single-season sack record, even if he already has recorded one of the biggest contract years in recent memory with a half-season to go. Watch out for the underrated 49ers pass rush. And please pray for the Buccaneers to improve on defense over the rest of the season.