The Nationals' bullpen in the first half was, to put it kindly, not so great, and you already knew that. In the first half, Washington relievers had a 4.80 ERA, the second highest in the game. The lack of competition in the National League East helped paper over the continued blown leads, but the feeling that the relievers could single-handedly torpedo a playoff run was pervasive. Something had to be done.
So something was done. General manager Mike Rizzo acquired Sean Doolittle and Ryan Madson from Oakland on July 16, then added Brandon Kintzler from Minnesota on July 31. You figured things would get better, and they have; nearly by definition, they had to be. But what if it's not better just by comparison to what came before? What if they're actually ... good?
It's been slightly more than a month since the first additions arrived, and we can report that things are much, much brighter at Nationals Park. Washington's bullpen has a 3.75 ERA in the second half, 10th best in baseball, a massive step up. But more importantly, they've been very good in the biggest spots, and isn't that what being a championship-caliber bullpen is all about?

We can show that relatively simply, with a stat called Win Probability Added. It's actually not as complicated as it seems, because it's all about game context. A solo homer may count for one run no matter when it happens, but it sure matters more in the ninth inning of a tie game than it does in the fourth inning of a 10-run blowout. WPA quantifies that by calculating the percent change in a team's chances of winning from one event to the next.
As you probably can guess, WPA didn't look favorably upon Washington in the first half, not after things like Blake Treinen blowing a 4-2 lead to lose 5-4 on June 29, or Matt Grace helping turn a 5-2 lead into a 6-5 loss on May 7, or a 6-0 lead that became an 8-7 loss in Miami on June 19. In the first half, only five teams had a worse bullpen WPA than the Nats.
But over the past month, or essentially since Doolittle and Madson arrived, Nationals relievers haven't just been better. They've been the best, just a hair above the super-pen the Yankees have built in New York.

Now let's clarify here; this does not mean the Nationals "have the best bullpen" now. They don't, not compared to the depth of talent teams like the Yankees and Dodgers have put together. But they don't need to be the best -- they just needed to get to good enough to win, and they've done that.
It all starts with Doolittle, who has long been one of baseball's more underrated relievers, partially because he's spent his entire pre-Washington career in Oakland and partially because he's had endless troubles staying healthy. (He's hit the disabled list five times in the past four years, including as recently as this May, with a shoulder strain.) While the 2.57 ERA he's put up for the Nats looks good enough, a deeper look shows just how valuable he's been.
A pitcher really has three jobs: to get strikeouts, limit walks and avoid dangerous contact. It's easy to see that Doolittle does the first two well, and his 81.0 mph average exit velocity is tied with Andrew Miller for the lowest -- i.e., best -- mark of any regular reliever. But we can do better than that by tying all three skills into a single number, which we call Expected wOBA (xwOBA). It combines actual strikeout and walk totals with a quality of contact measurement, based on exit velocity and launch angle, to express total pitching skill.

So far, 179 relievers have faced 125 batters this year, and you'll enjoy seeing how how Doolittle ranks.
Expected wOBA leaders by relievers (minimum 125 batters)
.191 -- Kenley Jansen
.207 -- Doolittle
.211 -- Craig Kimbrel
.217 -- Pat Neshek
.226 -- Miller / Roberto Osuna
.229 -- Felipe Rivero
.231 -- Will Harris
.233 -- Tommy Kahnle
.234 -- Dellin Betances
That's a pretty easy list of baseball's most dominant relievers, and Doolittle is up there with the best. Even better, going back to our friend Win Probability Added, no pitcher has given more WPA to his team over the past month than Doolittle. He's offered what the Nationals so desperately needed.
As much as Doolittle has offered, this was not going to be a one-man fix, as was noted so often before the trades, and the other additions have helped, too. Madson threw only nine innings before going on the disabled list with a finger injury on Thursday, yet they were nine dominant innings, as he struck out 13 of the 33 hitters he faced, allowing one walk. Kintzler has allowed only a single run in his nine innings.
Even with all the new faces, the Nats still needed some of those who remained to step up; as we said, this was going to require a big lift. Aside from Matt Albers, who had been basically the only reliable member of the bullpen before the trades -- he's 17th on the xwOBA list, better than Greg Holland or Archie Bradley -- veteran lefty Oliver Perez has retained his strikeout skills, whiffing 11 of the past 25 he's faced.

Beyond all that, the Nats can still hope on a pair of names who were supposed to have been big pieces all along. Shawn Kelley, activated when Madson was hurt, hasn't pitched since June due to injury, and he wasn't very good when he did (7.00), but it was only last year that he was quietly one of baseball's best relievers. Koda Glover, on the DL since early June with back and shoulder issues, was off to a fine start (2.08 ERA, 16/2 K/BB in 20 games) before ruining his stat line trying to pitch through pain.  
Again, the point isn't to say this is baseball's best bullpen. It's just that it's a much different bullpen, and that's a good thing. When the season began, the group consisted of Treinen, Kelley, Joe Blanton, Sammy Solis, Perez, Glover and Enny Romero. What if in October, it's Doolittle, Madson, Kintzler, Kelley, Perez and Albers, with either Glover, Solis or Blanton in the final spot? That's better, right?
That's a different group. It's a better group, a good enough one to win with. Earlier in the year, we saw news being made by departed Nationals fans who made light-hearted references to the struggles of the bullpen in their obituaries. Over the past month, that narrative has turned considerably. In October, they'll get the chance to prove it further.