Sport · The Record
Roberts said Díaz's fastball was good; Statcast shows it lost 24 points of whiff rate
The manager's diagnosis named the slider. On the same player page, the four-seam is the pitch that lost more whiff rate, gave up the higher batting average, and lost real velocity.
Scoreline first. Edwin Díaz's 2026 line on his Baseball Savant career page reads a 10.80 ERA across 12 games and 10.0 innings, with 17 hits, 8 walks, 15 strikeouts and a 2.50 WHIP. That is the record.
The explanation attached to it, in the reports cited here, is that the slider is the problem and the fastball is on the way back. The pitch-level table on the same site does not support the second half of that.
A note on titles before anything else: no snapshot Daily Pol holds states Dave Roberts' job title. He is described as the Dodgers manager in the cited reports, and he speaks in the club's voice - of Díaz, before the reliever's return, he said "Once he comes back, he's going to be our closer. So there is no like, somebody behind him that could potentially cost him a game." That is the basis for the description used below, and it is attribution, not a document.
What was said
Roberts, quoted by Yahoo Sports: "I thought the fastball was good. I thought the fastball life was good, but the slider just isn't having much effect." In the same passage he said Díaz left a pitch close to Waldschmidt, who hit a homer, and he returned to the fastball a third time: "But the fastball seems to be getting better, command of it, the life of it."
A second Yahoo Sports report carries overlapping remarks - both reports contain the "feel for that pitch since he got back" sentence, the "look at the options" passage and a version of the "clear-cut alternative" line, so they are best read as separate reports of the same availability rather than two sessions. In that second report Roberts said: "I just think he just hasn't had the feel for that pitch since he got back".
That sentence is the first line of the excerpt Daily Pol holds, and it does not itself name a pitch. Reading "that pitch" as the slider is an inference drawn from the slider reference earlier in the first report's passage, where Roberts names the slider before turning to the homer.
Díaz's own read matched his manager's. "I think the fastball was great today. Just I missed the slider right in the middle," he said, per Yahoo Sports, adding: "I'm a big-league pitcher, so I gotta make pitches." Two people, one diagnosis. That is what makes the data worth checking rather than one man's slip.
What the record shows
From Díaz's Statcast pitch-arsenal table at Baseball Savant (pitches, usage %, average velocity, whiff %, batting average against, run value per 100):
| Season | Pitch | Pitches | Usage | Velo | Whiff | BA | RV/100 | |---|---|---|---|---|---|---|---| | 2026 | 4-Seam | 126 | 55.5 | 96.1 | 15.4 | .375 | -2.7 | | 2026 | Slider | 101 | 44.5 | 88.9 | 23.6 | .348 | -3.1 | | 2025 | 4-Seam | 561 | 52.6 | 97.2 | 39.4 | .133 | 1.6 | | 2025 | Slider | 505 | 47.3 | 89.1 | 44.0 | .179 | 1.7 | | 2024 | 4-Seam | 475 | 52.1 | 97.5 | 36.6 | .190 | 0.6 | | 2024 | Slider | 437 | 47.9 | 89.6 | 39.4 | .186 | 0.6 | | 2022 | 4-Seam | 389 | 41.9 | 99.1 | 41.2 | .266 | 0.2 | | 2022 | Slider | 539 | 58.1 | 90.8 | 55.0 | .114 | 3.5 |
Against last season, the four-seam whiff rate is down 24.0 points, from 39.4 to 15.4. The slider is down 20.4, from 44.0 to 23.6. The larger fall belongs to the pitch Roberts said was good.
Batting average against runs the same way: .375 on the four-seam, .348 on the slider. Hitters have done more this season with the pitch Roberts said was good than with the one he said isn't having much effect.
One column does rank the slider lower. Run value per 100 pitches is -2.7 on the four-seam and -3.1 on the slider. It is also the noisiest column in a 227-pitch sample, and it is the column Daily Pol is least confident naming (see Limits).
The velocity check
The snapshot carries three prior seasons, so run the comparison against all three. The four-seam sits at 96.1 mph: down 1.1 from 2025 (97.2), down 1.4 from 2024 (97.5), down 3.0 from the 99.1 of 2022. The slider sits at 88.9: down 0.2 from 2025 (89.1), down 0.7 from 2024 (89.6), down 1.9 from 2022 (90.8).
Against both recent baselines the fastball's loss is the larger one. Against 2022 both pitches are slower and the fastball is still down more.
Analysis
This is Daily Pol's reading of the figures above, not a claim by any source.
A feel-and-command story is the right shape for a pitch that kept its stuff and stopped landing. It is the wrong shape for a pitch that lost 1.1 to 1.4 mph and 24 points of swing-and-miss. On the record as published, the diagnosis has been fitted to the pitch that held its velocity, while the pitch that lost velocity is the one being called good and improving.
None of that makes the public backing dishonest. A manager defending a struggling reliever in front of reporters is ordinary clubhouse management, and Roberts has been consistent about it: the closer job was promised before the return - "He's got to ensure us that he's in a good spot to close out major league games" - and defended afterwards on scarcity, with "I just don't think there's a clear-cut alternative."
That scarcity defense is untested here. Nothing in Daily Pol's evidence base documents how the other Dodgers relievers are throwing, so the alternatives claim is neither confirmed nor rebutted by this piece. Only the pitch diagnosis was checkable, and it is the half that does not hold.
A checkable prediction
If the "getting better" reading is right, the record will say so. Daily Pol predicts that as of 2026-09-30, the 2026 four-seam row on Díaz's Baseball Savant Statcast page will still show an average velocity below 97.0 mph and a whiff rate below 30.0 percent. If either figure clears those marks, this piece was wrong and the fastball recovered.
Limits
The 2026 sample is 227 tracked pitches over 10.0 innings. Rate stats move hard at that size and run value per 100 is the least stable of them; nothing here should be read as a settled season.
The Savant table in Daily Pol's snapshot carries no column headers. The columns are identified by position and format: the .xxx column as batting average against, the 15.4 / 23.6 / 39.4 / 44.0 / 55.0 column as whiff rate, and the final column as run value per 100 pitches. Those identifications are Daily Pol's reading, and the headline's 24-point figure rests on the whiff one.
The same snapshot gives 2025 pitch counts two ways: 561 four-seams and 505 sliders in the arsenal rows, 560 and 504 in the velocity rows. The table above prints the arsenal figures; the one-pitch gap does not move any rate cited.
No 2026 save total is cited: the figure is not resolved within the evidence base. Reported medical detail about an elbow procedure and an injured-list stint is not in the evidence base either, and is not asserted here.
Disclosure
This piece makes negative professional claims about two named living people. Daily Pol's ledger records Dave Roberts, Edwin Díaz and the Los Angeles Dodgers as subjects of a required comment request; a human editor documents that request before publication. No response had been recorded when this draft was written.