The Mets Are Allowing Half the Runs They Were — and It Still Might Not Matter

Here is the uncomfortable truth buried inside the most encouraging number the 2026 Mets have produced all season: allowing 2.9 runs per game over your last ten games is genuinely good pitching and defense. It is not a mirage on its face. But a team sitting 45-62, eighteen games back in the division and eleven out of the last wild card with 55 games left, doesn't get to treat a ten-game defensive stretch as a turning point. It gets to treat it as a question.

The question is whether this is real. The prior twenty games told a different story — 5.5 runs allowed per game, a number that explains a lot of the 45-62 record and most of the season's damage. A 2.6-run improvement in allowed runs is not a rounding error; that's the difference between a rotation that's bleeding and one that's actually pitching. Christian Scott's 3.00 ERA and Nolan McLean's 3.32 suggest there is genuine starting pitching here worth watching. But the bullpen has been shredded by injuries — Clay Holmes, Tylor Megill, Dedniel Nunez, Reed Garrett, and Justin Hagenman are all on the 60-day IL simultaneously — and Sean Manaea has been stretched out to nearly six innings per outing recently, up sharply from his earlier baseline. That workload jump is worth watching closely. When a team leans that hard on one arm to patch a decimated relief corps, the efficiency numbers tend to catch up.

The one-run game record makes the defensive improvement even harder to cash in on. The Mets are 10-20 in games decided by a single run — a win rate that trails their already-modest overall percentage. What that tells you is that even when the Mets keep games close, they lose them. Better run prevention gets you to close games; it doesn't guarantee you win them. Those two facts together — a dramatic recent drop in runs allowed paired with a chronic inability to finish tight contests — describe a team that has found a way to compete but not yet a way to win.

And then there is the Juan Soto situation, which colors everything. The Mets' only All-Star this season, a hitter slashing .296/.397/.574 before going down, is now on the IL with a grade 2 left calf strain. He expects to return this season, but a grade 2 strain is not a two-week inconvenience; it is a legitimate setback for a player who was already carrying the offensive load for a lineup that doesn't have much behind him. The run-prevention improvement means nothing if the offense goes quiet while Soto heals.

My read is this: the last ten games are worth acknowledging precisely because they were so unexpected from a team that had been hemorrhaging runs. McLean and Scott are real. The defensive tightening may reflect genuine stabilization rather than a soft schedule gift. But the Mets are not a team in the middle of a meaningful playoff chase — they are a team auditioning pieces for 2027 while the standings make the math increasingly brutal. A ten-game defensive stretch is evidence, not salvation. The box score says they're pitching better. The standings say it started about sixty games too late.