Weaver's Turnaround Is Real. The 1.88 ERA Is Running Ahead of It.

Twenty-five consecutive appearances without an earned run. Per MLB.com, that’s the second-longest single-season streak in Mets franchise history, behind only Mark Guthrie in 2002, and it’s why, with the Aug. 3 deadline days away, Weaver has become the name every contender keeps circling. Tim Britton of The Athletic put it plainly: if the Red Sox, Astros and Padres decide against dealing their closers, Weaver might be as desired as any reliever on the market.

The coverage around him is almost entirely transactional. Who’s calling. What the Mets want back. Whether the Yankees, of all teams, come back for a reunion. What almost nobody has asked is the question that should come first: what actually changed?

Because something did. This is not a fluke season dressed up by sequencing. But the answer isn’t the one the shape of the story implies, there’s no velocity rebound here, no mechanical fix, no new weapon. Weaver is throwing the same stuff with the same command from the same arm slot. What changed is how often he throws each pitch. And once you see that, you can also see the gap between the pitcher teams think they’re bidding on and the one the evidence supports: a genuinely excellent reliever whose 1.88 ERA is running meaningfully ahead of even his genuinely excellent skill.

The number the market is buying, and the number underneath it

Start with what the reporting says: a 6.00 ERA in April, then a reported 0.29 ERA since May 1. That’s the number driving the bidding.

The full-season picture is a two-layer story. The top layer is spotless, the 1.88 ERA and the clean lines around it. The layer underneath is still excellent, and noticeably less so: his FIP is 2.88, nearly a full run above the ERA. And every luck-sensitive indicator beneath the surface, production against relative to the quality of contact, ball-in-play outcomes, strand rate, leans the same fortunate direction at once, across 166 plate appearances, 106 balls in play and 41 outings. Any one of those, alone, is a shrug. All of them together is a pattern.

[CHART: ArticleRegressionPanel | Every luck-sensitive indicator leans fortunate at once, the expected-stats version of Weaver is elite, but not sub-1.00 elite.>

Read the panel as a cluster, not a checklist. ERA a run under FIP. Results running ahead of expected contact quality. Ball-in-play outcomes unusually fortunate by any standard. A strand rate elevated well beyond his own prior season. None of this says the season is fake, the expected-stats version of Weaver is a legitimately elite reliever. It says the results have outrun even that version, and outcomes this extreme are unlikely to persist. The honest regression target is the 2.88 FIP pitcher. Which, to be clear, is a pitcher worth trading for. It just isn’t the pitcher the 0.29 implies.

One indicator on that panel never leaned fortunate: barrel rate against, at 8.4%, slightly worse than league average.

What actually changed: the mix, not the stuff

The usual suspects all come back clean. Four-seam velocity is 95.3 mph, up 0.2 from last year. Walk rate, first-pitch strike rate and zone rate all moved less than a point year-over-year, command did not meaningfully change, so newfound strike-throwing can’t be the story. And no pitch cleared any meaningful threshold for a mechanical change. Same arm, same command, same velocity.

One caution on that velocity point, because the reporting history invites a misreading. Pinstripe Alley documented “a fastball velocity that had dropped 0.7 MPH” during his difficult 2025. Nothing in the 2026 data contradicts that, the +0.2 figure compares this year to last, not last year to the one before. Both can be true. What the 2026 number does establish is narrower and more useful: whatever happened to his velocity in 2025, this turnaround is not a velocity recovery. (You may also see his aggregate pitch velocity listed at 92.0, down from 92.6, that’s an artifact of throwing more slow secondaries, not decline.)

So if the stuff didn’t change and the command didn’t change, what did? The distribution.

[CHART: ArticleArsenalEvolution | Weaver cut his four-seam nearly 10 points and redistributed to the changeup and cutter, a mix change, not a stuff change.>

Weaver cut his four-seam usage from 59.2% to 49.7% and handed the freed-up volume to his two most-used secondaries: the changeup, up from 29.8% to 34.9%, and the cutter, up from 9.4% to 15.2%. The changeup is the centerpiece, a 38.7% whiff rate across 220 pitches, a .224 xwOBA against, and it’s now his designated finisher: he throws it 45.1% of the time in two-strike counts against 27.4% on first pitches. His fastball has become less predictable too, the gap between how often he leads with it and how often he finishes with it shrank from 22.9 points to 17.0.

The detail that makes this a selection story rather than a stuff story: per-pitch outcomes didn’t improve. The four-seam’s whiff rate actually fell from 24.5% to 22.1%, the changeup’s from 43.9% to 38.7%. Overall CSW ticked up to 30.6% from 29.2% anyway, because he’s simply throwing more of what already worked. He didn’t get better pitches. He got a better portfolio.

One honest limitation. All of these splits are full-season versus prior-season, so the rebalance can’t be dated to the streak’s May inflection, for all the evidence can say, the new mix may have been in place during the rough April too. That’s the difference between an explanation and a timestamped cause. What the full-season comparison does establish is the more important thing: the mix shift is the leading identifiable explanation for the year-over-year improvement, the one change in his profile large enough to carry it.

How much of the .208 is earned

Before writing the BABIP off as pure fortune, the batted-ball profile deserves its own look, because it transformed, and in a direction that would legitimately suppress hits.

[CHART: ArticleYoYBars | The contact-quality gains are real but incomplete, exit velocity and ground balls improved sharply while barrel rate against never moved.>

Exit velocity against dropped from 90.3 mph to 87.5. Ground-ball rate jumped from 27.4% to 45.0%, with line drives at 17.4%. The four-seam, his most-damaged pitch a year ago, has gone quiet: 87.7 mph average exit velocity against it versus 90.8 last year, and one home run allowed on 49 balls in play after eight on 110 the year before. Softer contact on the ground doesn’t fall for hits the way hard contact in the air does, so part of the low BABIP may be earned, not borrowed.

You could watch this thesis get stress-tested in real time last weekend, when the Mets took on the Dodgers and their loaded lineup.

[VIDEO: wUYoK4cfWxc | The Mets’ most recent series against the Dodgers’ loaded lineup is exactly the kind of test where Weaver’s contact-management gains, 87.5 mph exit velocity against, .208 BABIP, get stress-tested in real time.>

Or take Wednesday: a scoreless eighth in the Mets’ 3-2 win over Atlanta in Game 1 of a doubleheader, his 15th hold, sealed by inducing an inning-ending double play from Matt Olson. That’s the contact-management profile in miniature, a dangerous left-handed bat, on the ground, two outs.

Now the counterweight, and it’s the single most important skeptical fact in his profile. Barrel rate against is 8.4%, statistically unchanged from last year’s 8.3%, and worse than the 7.7% league average, sitting in the 38th percentile. When hitters do square him up, they still barrel him at an above-average clip. The soft-contact story is real, but it’s a story about average exit velocity and ground balls, not about eliminating the most damaging contact. And the samples urge restraint everywhere here: 49 balls in play on the four-seam, 30 fly balls behind the home-run suppression, and the exit-velocity gain itself is the kind of mid-size, single-season shift that can shrink. The fair reading is that some unquantifiable share of the .208 is skill, and the barrel rate is the reason not to round that share up.

The righty-killer premium

The trade market has attached a specific label to Weaver. Jimmy O’Brien of Talkin’ Yanks made the case out loud: “He is the reliever the [Yankees> need... someone who can get [right-handed batters> out. Go get him.”

The observed 2026 split backs the framing emphatically, almost too emphatically.

[CHART: ArticlePlatoonSplits | A .303 OPS gap in favour of facing righties matches the trade-market framing, at sample sizes where such gaps can be substantially noise.>

Right-handed batters have hit .099 against him with a .322 OPS, a 28.2% strikeout rate and a 0.56 WHIP across 78 plate appearances. Lefties have managed a .625 OPS in 88. A .303 OPS gap is enormous, and 78 plate appearances is far too small a sample to establish a true-talent platoon skill. Two further reasons for caution. First, that .099 average against righties is deflated by the same unusually fortunate ball-in-play environment driving the season line. Second, the arsenal logic doesn’t obviously support the label: his best pitch is a changeup, the classic weapon against opposite-handed hitters, yet lefties are the side having relative success. The split, in other words, arrives without a mechanism the arsenal would predict, if anything, the arsenal points the other way. A team paying a righty-killer premium is paying for 78 plate appearances of results, not an established skill.

The process blinked first

Weaver, for his part, sounds like a man at peace with the noise. “It was certainly a lot harder in the past,” he told MLB.com of the deadline swirl. “My wife and I understand this game, and we’re mentally prepared, to an extent. But we’re also not by any means one foot out the door.”

The results have given him no reason to sweat: a 0.00 ERA over his last three appearances, and on July 29 he struck out Ronald Acuña Jr., the stuff is still retiring elite hitters. But underneath the clean lines, the most recent window carries an early flicker worth naming, with the sample caveat stated up front: this is 16 plate appearances and 14 balls in play, a window that regularly swings this much on nothing at all.

[CHART: ArticleTwoWindowTrend | Strikeouts, whiffs, expected damage and hard contact all cooled at once over a tiny recent window, the results haven't noticed yet.>

Over the last 14 days: strikeout rate 6.2% against 29.1% before, whiffs 17.9% against 31.4%, xwOBA against up to .334 from .258, hard-hit rate 50.0% from 31.6%. The 10-game rolling xwOBA tells the same story in slower motion, climbing from a season-low .141 on July 12 to .208 by July 29. Four independent process metrics moving the wrong direction simultaneously while the results stay spotless is the classic signature of results temporarily outrunning process. It is also, at this sample size, entirely consistent with noise, and zero walks in the window cuts the other way. File it as a watch item, not an alarm. But whichever team acquires him this week will be watching the same four lines.

So here is the answer the deadline coverage skipped. Weaver’s turnaround is real, and it has a real explanation: he stopped leaning on his most hittable pitch and reorganized his arsenal around his best one, and hitters started beating the ball into the ground. Some share of the hit suppression that followed may be earned, the barrel rate is the reason to stop short of promising more than that. But the 1.88, and the reported 0.29, sit on top of that skill, inflated by a season in which every bounce-sensitive number broke his way at once. The team that trades for Weaver on Aug. 3 is buying the 2.88 FIP pitcher: a legitimately elite setup arm built around a weaponized changeup. Pay for that pitcher, and the deal works. Pay for the streak, and you’re paying for the results rather than the mechanism that produced them.