What Happened to Paul Skenes?

Skenes won the 2025 Cy Young with a 1.97 ERA. He finished 2026 at 3.82 over 32 starts. I pulled every pitch from both seasons off Statcast to see what changed. It is the home runs. He gave up 20 this year after giving up 11 all of last season, and 13 of them came off a four-seamer that lost 1.4 mph. The strikeouts never went anywhere.

I thought this was bad luck. It wasn't.

Cumulative ERA and FIP by start. Opening Day was five runs in two-thirds of an inning against the Mets, which put his ERA at 67.50, so the chart starts at his second outing. Through May the two lines sat almost on top of each other. Then the ERA pulled away: seven runs in Philadelphia on July 1, five each against the Cubs and Reds later that month. By early August it was running a full run ahead of the FIP, which is the kind of gap that usually closes on its own.

It closed from the wrong direction. The FIP came up to meet the ERA instead, climbing from 2.80 in late May to 3.40 at the finish, and the gap shrank to 0.42. FIP counts home runs, and the home runs never stopped. Six of them came in September alone. Red dots mark them on the chart, and they land on the same starts where the ERA jumps.

The home runs have a fingerprint

The arsenal, 2025 vs 2026

Whiff rate is down on almost everything. The changeup and slider still miss bats, but hitters did a lot more damage on the changeup when they got to it, .233 against after .103. The sinker went from 15% whiffs to 12%. The splitter got hit at .323. The curveball was a real weapon in 2024 and he has basically stopped throwing it. The four-seamer is the strange one. Hitters swung and missed at it slightly more often than last year and batted just .220 against it, but the home runs off it went from 5 to 13. They were not hitting it more. They were hitting it farther.

Usage and velocity deltas vs 2025. Green = improved, red = worse. Whiff% = swinging strikes per swing. BAA hidden under 10 at-bats.

Velocity

Average four-seam velocity by start. He opened at 97.9 and never got back there. From mid-May on he was under 97 in most starts, and the 95.7 against Miami on August 11 was his low for the year, two and a half ticks off his 2025 average of 98.2. He finished the season at 96.8 on average, 1.4 below last year. It stepped down in May and stayed down for four months, which is a long time to call it a blip. Nothing that looks like an injury, but this is the number the rest of the page keeps pointing back to.

So what actually happened?

The home runs. 20 in 174.1 innings after 11 in 187.2 last year, mostly on four-seamers over the middle of the plate. Everything else on this page is either downstream of that or beside the point.

The blowups look smaller than they did in July. Opening Day and the July 1 game in Philadelphia are 12 of his 74 earned runs, and pulling both drops his ERA to 3.29, which is not the rescue it sounded like at the time. Nine of his 32 starts cost him four runs or more. That is a pattern, not a couple of bad nights.

What did not change is worth saying too. He struck out 10.3 per nine, the same as last year. He walked 2.5, up a little from 2.0. He was healthy enough to make 32 starts and throw 174.1 innings. This is not a pitcher coming apart.

The honest read is narrower and less dramatic than the one I started with. He lost 1.4 mph off the fastball, hitters stopped being late on it, and the extra home runs turned a 2.00 ERA into a 3.80 one. That is the whole story. I spent most of the summer calling it bad luck because the FIP was a run below the ERA, and then the FIP climbed to 3.40 and took the argument with it.