New York Mets at Washington Nationals: Final Score & Recap

The New York Mets edged the Washington Nationals 2-1 at Nationals Park on May 21, 2026, in a tightly contested game that the DiamondIQ model's estimate had opening as a 59 percent proposition in favor of the home side before flipping entirely by the final out. The Mets scored both of their runs in the third inning and protected that lead through eight frames, while Washington's lone run came on a James Wood double in the fifth that cut the deficit to one and, per the DiamondIQ model's estimate, added 13.5 percent to the Nationals' win probability. David Peterson and Luke Weaver combined to hold Washington largely in check, contributing 15.0 and 23.3 percent to the Mets' win probability respectively, with Weaver's eighth-inning work proving particularly decisive after Dylan Crews grounded out in a high-leverage spot that swung win probability 16.5 points in New York's favor.

The game's most volatile sequence unfolded in the bottom of the ninth with closer Devin Williams on the mound, producing three of the five biggest win-probability swings of the entire contest. Daylen Lile opened the inning with a double that added 25.1 percent to Washington's chances, representing the single largest positive swing for the Nationals all night, and momentarily made a home comeback feel plausible. Keibert Ruiz followed, and while his groundout ultimately registered as a positive for Washington's probability by 19.0 percent given how the baserunning situation resolved, the inning ended on José Tena's strikeout, a play that subtracted 25.9 percent from the Nationals' win probability and closed the door entirely. Brooks Raley contributed 9.7 percent to New York's win probability in relief to help bridge to Williams.

On the offensive side, Keibert Ruiz and Daylen Lile paced Washington in win-probability terms, finishing at plus-22.3 and plus-22.0 percent respectively, though neither translated that leverage into run-expectancy gains of note given their RE24 figures of essentially zero. Bo Bichette was the most impactful Mets bat by WPA at plus-9.6 percent, also adding 0.9 runs above expectation. New York finished with six hits against five for Washington, while the Nationals committed two errors to the Mets' one. The DiamondIQ model leans toward crediting New York's pitching staff, which limited damage across a game in which Washington's best opportunities arrived late and ultimately went unconverted.