Two Starters Have Combined for Seven Wins. Seattle Needs Twenty-One More.
The number that should be keeping the Mariners' front office awake isn't the magic number. It's the one buried in the back of the rotation: a combined 7-20 record between Luis Castillo and Bryce Miller over 46 starts this season. Seattle needs 21 wins from 34 remaining games to clinch. Two of their five rotation slots have been net liabilities for most of the year, and Emerson Hancock — the one arm who had quietly steadied things at 8-9 with a 3.68 ERA — is now on the injured list. The math does not get easier from here.
What makes Castillo's collapse so jarring is the weight of expectation he carries. He arrived in Seattle as a frontline arm, the kind of pitcher you build a rotation around, and for stretches he has been exactly that. But a 5.24 ERA and a 1.43 WHIP across 24 starts this season tells a different story — a pitcher who is leaving too many runners, surrendering too much contact, and giving opponents too many opportunities to punish a defense that has itself been fraying. Miller's situation is distinct: his 4.17 ERA is survivable in isolation, but 4-10 in 22 starts means the offense hasn't been able to bail him out either, and at this point in the season, wins and losses carry real weight.
The top of the rotation is genuinely good. Logan Gilbert has been a workhorse through 35 starts, and Bryan Woo's 13-10 record with a 3.93 ERA makes him the staff's most reliable presence. George Kirby sits at 9-11, which is serviceable. Together those three give Seattle a credible foundation. But a five-man rotation runs five times through the order, and the fifth slot — and increasingly the fourth — has been where games go to die. In a pennant race where every start matters, a rotation that leans this heavily on three arms while two others bleed losses is structurally fragile.
The defensive deterioration compounds everything. Seattle has allowed 6.9 runs per game over the last ten contests, up sharply from the 4.7 they were surrendering in the prior stretch. That kind of run-prevention collapse doesn't happen in a vacuum — it lands hardest on pitchers who are already struggling to generate clean innings. Castillo's profile, with a WHIP that suggests he's putting runners on at a high rate, is exactly the type that gets punished when the defense behind him softens. The two problems are feeding each other.
The 8-2 record over the last ten games is real, and I wrote as much earlier in the week. But the Mariners have been winning despite the rotation's back end, not because of it. At some point in the next 34 games, Castillo or Miller will draw a start in a game Seattle cannot afford to lose. The question isn't whether that moment is coming — it's whether either pitcher has shown enough in 2026 to suggest they'll rise to it. The record says no. The ERA says no. The WHIP says no. September's winning streak has bought this team time. It hasn't fixed the thing most likely to end their season.