All stats entering Friday's games.

There’s more than one way to define the lowest point of the 2021 Yankees' season, but we’re going to say it came at around 5 p.m. ET on July 4. In the first game of a doubleheader against the Mets, Gerrit Cole lasted only 3 1/3 innings, then his club blew leads of 4-1 and 5-4 as Aroldis Chapman failed to retire any of the three batters he faced, and then Lucas Luetge allowed four hits to five batters. The Yanks lost, 10-5. They fell to .500, a season-high 10 games out of first place in the AL East.

They won the nightcap, thanks in large part to Nestor Cortes Jr. and Chad Green, and the next game, and the next game too. Since July 4, what had looked like baseball’s most disappointing team (with apologies to the Twins) has been baseball’s best team, posting a .718 winning percentage over the last six weeks. If the season ended today, the Yankees wouldn’t just be in the AL Wild Card Game, they’d host it in the Bronx. They may even yet take the AL East, considering the season ends with three at home against the Rays.

In a picture, the Yankee season looks like this:

But how? What, exactly, has changed for the Yankees since Independence Day? Or, for that matter, since we investigated their disappointing offense on June 7, at which time we wrote that “they are alone in terms of just how deeply they’ve underperformed,” given this year’s high expectations? Or since July 16, which was technically the nadir of their playoff odds?

What’s different?

It’s time to find out.

1. They are who we thought they were.

This is the boring answer that no one wants to hear, but it’s where you have to start. The current Yankees are not actually the best team in baseball. The early-season Yankees were also not really one of the weakest teams in baseball. Over the course of a full season, you might be what your record says you are. Over portions of it? Well, a lot can happen. Things even out.

The Yankees, right now, have a .574 winning percentage.

The Yankees, before the season, were projected to have a .588 winning percentage.

That’s … almost exactly right, and if they continue to pound the Twins this weekend, they’ll get there.

In fact, if you take what’s happened to date, and combine it with the rest-of-season projections, you’ll get a 93-69 mark ... just barely shy of the projected 95-67. Now, obviously, the shape of it all has been a mess, with wild ups and downs and little consistency -- the club’s monthly winning percentage for the first four months went .462, .607, .462, .609 before getting to August’s .789 -- but this is sort of why baseball seasons are six months long, ideally. You can point to them being lousy in June; you can also point to them being on a 98.6-win pace since their 6-11 start. This didn't just happen a few weeks ago, anyway.

Obviously, there’s a lot of fluctuation within there. In April, that .462 winning percentage came while the team was outscoring its opponents in a way that would make it seem like it was playing more like a .538 team. In July and August, that .690 winning percentage has come with a run differential that suggests a .599 winning percentage. The Yankees were better than they seemed before and not as good as they seem now. This is how long seasons work.

So the boring, unexciting, mostly correct answer is this: Talented roster, given enough time, plays well.

Still, we’re not going to chalk all of this up simply to water finding its level. Rosters change. So do players. What’s fueling the recent hot stretch?

2. The pitching kept this ship afloat.

As recently as July, the Yankees had the third-lowest slugging percentage in baseball for the month, and they scored the third-fewest runs that month. It’s not that the lineup has no role to play here -- we’ll get to that -- but it’s only been very recently that the bats have started to vaguely wake up, and despite that, the team was good in July, too.

Instead, look how good the pitching has been, save for a bad June stretch.

Overall, Yankees pitching has the sixth-best ERA (3.67) in baseball, and that group has burdened the effort of supporting an extremely inconsistent offense.

That said, the current Yankees pitching staff looks a lot different than the group we saw in April. While Cole and Jordan Montgomery have been rotation mainstays, we've seen massive improvement from Jameson Taillon, who had a 5.49 ERA through the end of June but has a sparkling 2.01 ERA since -- fueled in part by making in-season changes to his curveball. Rookie Luis Gil still hasn’t allowed a run in his first 15 2/3 innings. Cortes Jr., a journeyman now in his third stint as a Yankee, has a 2.55 ERA as a swingman.

Green and Jonathan Loaisiga have been excellent (combined 2.76 ERA), but several of GM Brian Cashman’s under-the-radar bullpen acquisitions also paid off while veterans Chapman, Zack Britton, Darren O’Day and Justin Wilson were ineffective, injured or both.

Luetge, back in the Majors for the first time since 2015, has a 3.09 ERA in increasingly important situations. Wandy Peralta, acquired for Mike Tauchman in April, has allowed one earned run in his last 10 games. Clay Holmes, who came in a little-noticed pre-Deadline deal from Pittsburgh, allowed five batters to reach in seven games as he piled up grounders. Joely Rodríguez, who came with Joey Gallo from Texas, has allowed just two runs in his first eight games.

The pitching is good, and has been good, for nearly the entire season. It’s just a different kind of good now, with different pitchers doing different things.

But, you might correctly ask, the pitching was good early on, too, and the team wasn’t winning. What else is it? Well, we said we’d get to the offense.

3. There’s plenty of credit to spread around in the lineup.

Let’s split this offense up into three groups -- the ones here mostly all along, the pair of big trade acquisitions and the various lesser-known players who have come and gone along the way.

A. The core.

In the first week of June, we looked at the 10 Yankees hitters with more than 100 plate appearances to try to identify their struggles, if there were any. (There were.) Those 10: DJ LeMahieu, Aaron Judge, Gleyber Torres, Gio Urshela, Clint Frazier, Giancarlo Stanton, Gary Sánchez, Brett Gardner, Aaron Hicks and Rougned Odor. What have they done since?

Those 10, through June 6: .716 OPS
Those 10, since June 7: .782 OPS

Considerable improvement, that. Now, Hicks was already out for the year by that point, and Frazier hasn’t played since June. Of the remaining eight, six have played better since that June 7 article than they did before that. Odor, Gardner, Sánchez and LeMahieu have all raised their OPS by at least 100 points, though LeMahieu’s year has still been a slog; he hasn’t homered since June 26. Urshela and Stanton have each improved a little.

Torres didn’t, though he was showing signs of life before his thumb injury. While Judge has also been somewhat worse, that’s only in comparison to his spectacular start (dragged down by a COVID-interrupted June); he’s off to a .300/.393/.557 start in August and is generally having an MVP-caliber year. Throw in Luke Voit, who was injured earlier and now has seven hits in his last five games, and a lot of this is just that some of the guys you needed to be more productive ... have been.

Maybe it's easiest to put it this way, really: The top five spots in the lineup were below average in both April and May and have been above average in each of the last three months. The bottom four spots were above average only in June and have been average-ish or below in each of the other four months. Don't mistake this as LeMahieu or Torres being "fixed," because they're not -- and even in August the underlying metrics suggest only a middle-of-the-pack offense. But no matter how it happened, this always had to start from within.

B. The two big acquisitions.

The pair of trades for Gallo and Anthony Rizzo were hailed as badly needed moves to add lefty bats with plus gloves who would both fit well in Yankee Stadium as well as add some lineup diversity to a group that desperately needed it. Those moves made sense at the time and still do, though to this point, it's difficult to point to these trades as the ones that turned the season around.

That's mostly because the rebound was already happening -- New York was 14-9 in July, nearly all of which was spent without either player. It's also because Rizzo has still played in only 11 games as a Yankee, thanks to a COVID-19 diagnosis, and Gallo has hit just .160/.308/.387, though with good defense. They’ve had their moments, most notably Rizzo’s fantastic first week with the team, and you can expect more. It's easy to see one or both having some huge moments in the weeks to come. But overall, this isn’t a story of the new guys coming in and turning everything around, either.

C. The "little feisty guys."

Those were Voit’s words after a win on Thursday, when the bottom of the lineup scored five of the first six runs. This group has changed over the course of the season, really.

Just look at the July 18 lineup that posted a 9-1 win over Boston, the one that was necessary after a COVID-19 outbreak sidelined a quarter of the team. The outfield starters were Ryan LaMarre, Greg Allen and Trey Amburgey. Chris Gittens was the first baseman. But none of them are still on the team, and now Voit is talking about backup infielder Tyler Wade, backup catcher Kyle Higashioka and hometown hero Andrew Velazquez, who was signed as a non-roster invite last winter and recalled on Aug. 9 when Torres was injured.

At the coldest possible level, that group -- the seven names listed in the preceding paragraph -- has not contributed much with the bat. On the season, for the Yankees, they’ve posted a line of .212/.299/.363, a .661 OPS, and it’s too easy to conflate “a more entertaining alternative to a slow, lumbering lineup” with “better baseball players.”

But while there’s really no quantifying “energy” or “chemistry” or whatever qualities you think this group has brought at various times, it’s also too simplistic to look at the batting line and think that value hasn’t been provided. That’s maybe especially true in the field; does anyone, for example, think that the defensively limited Torres makes the game-saving play that Velazquez did to end Wednesday's game?

One thing we can quantify is baserunning, where the Yankees were absolutely miserable for much of the year. As we wrote in June, “no team has had more runners thrown out on the bases (28, most in baseball) and no team has taken a lower rate of extra bases (30%, lowest in baseball).”

While the Yankees have made a massive turnaround in stolen bases -- they were dead last in the first half and have the most in the second half -- what’s more interesting is when you look at all the other things a runner can to do add or subtract value, like taking that extra base or getting thrown out (or not). How’s this for a trend?

In April, the Yankees were the worst baserunning team.
In May, 28th.
In June, 29th.
In July … 15th.
So far in August? 5th.

It’s a massive turnaround. We can guarantee you it wasn’t Stanton or Sánchez suddenly gaining speed, either, and it wasn’t. It’s Wade. It’s been Allen. It’s been Gardner.

None of this has fundamentally changed the Yankees' offense. The second-half Yankees don’t strike out less (24.8%) than the first-half group did (24.3%). They don’t have a better batting average (.240, compared to a first-half .236). It’s not like they’re bunting for hits meaningfully more (just eight all season, though six since the break). Manager Aaron Boone hasn’t bowed to talk radio pressure suggesting he call more sacrifices; the Yankees have just four in the second half after only six in the first half. It’s not what this team does.

What the Yankees are doing is hitting it on the ground a lot less (45% before the break, and 40% after), which is a good thing if you like power and avoiding double plays. The Yankees went from having the most double plays grounded into in the first half to "only" the eighth most in the second half.

But it’s this, too. It’s not running yourselves out of opportunities, and being able to create some when the slug isn’t there. That's all important, because ...

4. It’s about all of this paying off in tight games.

…. of this. If you’ve watched a game, especially recently, you know the Yankees tend to play excruciatingly tight baseball games. They’ve played the most games decided by two runs or fewer (69), and they have the best winning percentage in those games (.571). (They’re five games under .500 when the margin is three runs or more, which points back to how punchless the offense has often been.)

“We’ve played more than our share of really close ones,” said Boone recently. “The good thing is, I think our guys are really comfortable in those situations, whatever the outcome may be.

No team’s pitchers have had fewer low-leverage plate appearances this year, which shows just how often they’re in tight spots. (On the hitting side, only three teams have had fewer low-leverage trips to the plate.) In spots considered "medium" or "high" leverage, only two teams have had better pitching performances than the Yankees, and generally being on a list "just behind the Dodgers and Giants" and "just ahead of the Rays and White Sox" is a place you want to be.

As you’d expect, the Yankees' performance in these games has done a lot to affect their win/loss record. They were just 5-7 in tight games in April (.416), but they’ve won 22 of their last 27 (.815) since July 1. That, it should go without saying, is not close to sustainable for an entire season. Then again, it doesn’t have to be. Entering Friday's game against the Twins, there were only 40 games left. The Yankees are on their way to a number of wins in the 90s, just as we expected. As grim as it may have seemed in June, when they were swept by the Red Sox twice, or in early July, when the catastrophe against the Mets happened, there's a reason we prefer six-month seasons.

The Yankees might be who we thought they were -- and they're more likely than not to play October baseball.