Freddy Peralta’s decline is real. But the stuff is not gone.

From April 1 through May 12, Freddy Peralta posted a 2.64 ERA. From May 17 through July 20, that number jumped to 6.56.

This is not only bad luck. FIP rose from 3.46 to 5.18. Overall run value per 100 pitches — how much his stuff helped or hurt him versus an average result — flipped from +1.11 to -1.39. CSW, the share of pitches that end as a called strike or swing and miss, fell from 27.3% to 23.7%. Opposing wOBA climbed from .282 to .393.

Those metrics moved negatively together. But it also does not look like a pitcher whose raw stuff disappeared. His four-seam velocity increased, his walk rate did not rise, and overall xwOBA — expected offensive results based on contact quality and plate-appearance outcomes — barely changed. That tension drives the rest of the story.

The four-seam is still the centerpiece of Peralta’s arsenal. It is also the clearest window into the decline.

Average velocity rose from about 93.7 mph to 94.5 mph. The fastball got harder. It also became less deceptive and less forgiving.

Hitters chased it less (27.4% → 20.0%) and missed it less (21.5% → 17.1%). Hard-hit rate jumped from 19.1% to 30.6%. Four-seam run value per 100 flipped from +1.90 to -1.55.

The evidence suggests a real effectiveness drop. It does not prove that hitters fully solved him.

The easy story would be that Peralta started throwing more four-seamers down the middle. He did not.

Heart-zone share on the four-seam fell slightly, from 23.4% to 21.6%. Shadow-zone share stayed nearly identical (39.0% → 39.1%). The general location map remained similar.

What changed was the damage against those locations. Heart-zone hard-hit rate rose from 25.8% to 34.1%, and heart-zone run value per 100 flipped from +6.25 to -0.86. Four of six decline-period four-seam home runs also came in the heart.

This was broader than one broken pitch. It looks more like an execution and approach problem across Peralta’s most important situations.

When he threw a four-seam after another four-seam, run value per 100 flipped from +1.26 to -2.18. First-pitch four-seam results worsened (+1.16 → -2.73), and two-strike four-seam results did the same (+2.37 → -5.08).

The changeup also lost effectiveness, with run value per 100 flipping from about +1.45 to -2.97.

Inside the decline window, the second time through the order was the worst phase: .471 wOBA and -2.71 RV/100, compared with .338 and -0.52 the first time through.

The decline is real. It does not look like a complete physical collapse.

Velocity is intact — and up. Fastball shape did not fall apart: horizontal movement stayed near -6.9 to -7.1 inches, and vertical movement near 16.9 to 16.4. The overall location distribution did not dramatically worsen. Walk rate stayed roughly stable (10.1% → 9.6%). Overall xwOBA moved only from .305 to .307. No reliable physical or mechanical adjustment was detected.

BABIP also rose from .264 to .341. That does not erase the process decline — FIP and pitch run value worsened too — but it does mean some of the extreme ERA jump may still involve sequencing, contact placement, defense, or variance.

Those indicators suggest Peralta may not need to rebuild his arsenal. He may need to improve execution, regain deception, and avoid predictable or poorly located fastballs in important counts. That is a reason for hope, not a guarantee of a rebound.

The heatmaps below compare similar four-seam location maps with much harder contact and worse run value in the decline window.

These are the clearest measurable signs that the decline is reversing — or not.

Over the next three to five starts, watch this short list:

If the misses return and hard contact falls while velocity remains intact, that would be an encouraging sign that the decline is correctable. If the same contact and sequencing problems persist, the concern becomes more serious.

Peralta’s decline is real, but it does not look like a simple loss of stuff. His fastball is still hard, and his general location profile has not collapsed. The problem is that hitters are missing less and doing more damage against similar locations, while his changeup and important-count execution have also slipped. There is reason to believe the decline could be corrected, but the next few starts need to show better chase, more whiffs, and less hard contact.

Windows compare a strong opening (April 1–May 12) with a decline period (May 17–July 20) using Statcast pitch data and Baseball Pulse’s published finding pipeline. Attack zones are a documented Savant-style approximation from plate location and strike-zone height, not an official Statcast column. The decline is associated with contact quality, sequencing, count leverage, and pitch interaction — not a confirmed cause.