While much of the focus in Washington before the year was on its offense and quarterback drama, perhaps the defense in the nation’s capital should get more of our attention.

Washington has allowed 27 points in its two games this season, and only Carolina and the Jets have allowed fewer. New defensive coordinator Joe Barry’s unit has also allowed just 4.6 yards per play, ranking sixth in the league.

In terms of our advanced metrics, Washington ranks seventh in points prevented above expectation after a loss to Miami in Week 1 and a win over St. Louis on Sunday.

The early performance marks a vast improvement over last year, in which it ranked 29th in points allowed per game (27.4), 27th in yards allowed per play (5.9), and 30th in Adjusted Washington ranked 13th in yards allowed per carry (4.09), and sixth in both Adjusted Rushing NEP and Adjusted Rushing NEP per carry.

This season, Washington is 26th in yards allowed per carry (4.5), but still ranks in 11th Adjusted Rushing NEP and fifth in Adjusted Rushing NEP per carry.

Before going on to Washington’s pass defense, it would be useful to see what is driving the disparity between the raw and advanced statistics.

The first place to look would probably be turnovers, but Washington’s rush defense has not recovered a fumble (Ryan Kerrigan did force one against St. Louis, but it was recovered by Rams receiver Stedman Bailey).

The next area worth examining is third and fourth down performance. While yards per attempt treats all yardage the same, expected points models give greater weight to stops or successes on third and fourth down.

This is why there is a 21-spot difference between Washington’s rankings in yards per carry and Adjusted NEP per attempt.

Washington had faced five rushes on third or fourth down, and only allowed one first down, tying them with Kansas City for the best third/fourth down rush defense. Opposing offenses have needed to gain an average of only 3.6 yards on these plays against Washington, but they have only gained 1.8.

One of these stops occurred on 4th-and-1, another happened on 3rd-and-1, while another was made on 3rd-and-2 (the other was a one-yard run by Rams quarterback Nick Foles on a 3rd-and-6; the conversion was an eight-yard rush by Miami’s Lamar Miller on a 3rd-and-8).

An 80.0% success rate on third and fourth down is not sustainable in the long run (more on that later), but it does go a long way in explaining why the Washington rush defense rates highly in our advanced metrics.

It also is a component of Washington’s continued strength in short yardage situations, as its rush defense was sixth-best in conversion rate allowed against runs with three yards to go or fewer on all downs last season

Ryan Kerrigan has led the way this year, upping his performance to soften the blow of fellow outside linebacker Brian Orakpo's departure. In 2014, Orakpo was tied for sixth at the position in run stop percentage, while Kerrigan came in at 13th (among 46 qualifying players); this season, Kerrigan ranks fifth at the position (per Pro Football Focus).

Pass Defense Improvement

While Washington was strong against the run last year, it had a poor defense overall, due mostly to its inability to stop teams from moving the ball through the air.

Only Atlanta allowed more net yards per pass attempt than Washington, which allowed 7.2. Advanced stats paint a bleak picture as well, as Washington was dead last in both Adjusted Passing NEP and Adjusted NEP per pass.

Despite the strong rush defense, Washington’s poor pass defense dragged down the overall unit’s effectiveness (as mentioned, it ranked 30th in overall Adjusted NEP in 2014). This makes sense given the modern NFL, in which passing efficiency on both sides of the ball has a greater correlation with success than rushing efficiency.

This season, though, Washington looks vastly improved against the pass, ranking fourth in net yards per pass (4.7), fourth in Adjusted Pass Defense NEP, and third in Adjusted NEP per pass.

Changes in the defensive backfield seem to be driving the early improvement (as opposed to its pass rush; Washington ranked 19th in sack rate last season and is 13th this season).

Corner David Amerson is out of the starting lineup after raking 106th out of 116 qualifying corners in yards per coverage snap (1.7) last season, according to PFF. Chris Culliver has taken his place after signing with the team in the offseason, and the former 49er has allowed just two catches for three yards on three targets and 37 coverage snaps.

Brandon Meriweather (75th out of 90 safeties in yards per coverage snap last season) and Ryan Clark are also gone, giving way to safety @Dashon GolDashon Goldson.

Like its run defense, the Washington pass defense has been good on third and fourth down, ranking seventh in success rate on these downs. That figure might be less impressive given that when opponents pass in those situations, they need an average of 8.5 yards (the sixth-highest distance to go in the NFL).

Then again, it is another testament to Washington’s defense that it has been able to put opponents in disadvantageous situations in the first place.

Can It Continue?

It is hard to make forward looking predictions here with much certainty, given the two-game sample size we are dealing with.

There are, however, a few things we can look for.

For example, we should predict Washington will have better turnover luck, which could at least partially offset regression in other areas.

Washington has only forced one turnover this season, a fumble recovery after a pass reception against Miami. Opponents recovered the other two fumbles forced by Washington.

Turnovers are incredibly random events, and past turnovers (or lack thereof) are of little use when predicting future ones, and this is particularly true on defense.

As for Washington’s aforementioned performances on third and fourth down, there is also a low correlation between third down success rate on defense from one half of the season to the other.

Aaron Schatz found that third down performance tends to regress to a team’s overall level of play over time, which makes sense given that teams face a comparably small number of third down plays (this is especially true of defenses; newer research seems to suggest offensive performance on third down may be more consistent).

So while Washington’s rush defense has been good on third down, is its poor ranking in raw yards allowed per carry cause for concern?

Not necessarily, as while the 80.0% success rate on third down is more of a sample size fluke than indicator of true talent, the same could be said of Washington’s yards per rush average.

Yes, the team has allowed 4.5 yards per carry, but this is a stat that can be inflated by a few long runs, especially in a small sample.

Washington has faced 31 rushes this season, allowing 141 yards. If we take away just one run (let’s go with Lamar Miller’s 17-yard Week 1 carry), the numbers drop to 124 yards on 30 carries; this comes out to 4.1 yards per attempt, which would put Washington right around league average.

The more predictive stat would be success rate (the percentage of plays that yield positive expected points), a metric that correlates with itself at higher rate than yards per carry.

Only 12 runs have yielded positive expected points for Washington’s opponents, giving the Washington run defense a success rate of 61.2%; this would have ranked eighth in the NFL last season.

Based on this and the fact the Washington defense was stout against the run last year, we can be fairly confident the rush defense will continue to be successful.

As for the secondary and defense as a whole, it is harder to make concrete predictions, but the early returns have certainly been positive.