Pittsburgh Pirates at New York Yankees: Prediction, Odds & Preview
DiamondIQ Model — Win Probability
The model leans NYY (55%). DiamondIQ model v2: season records, home-field advantage, the starting-pitcher quality gap (PitchIQ), and a calibration adjustment fit and validated on four seasons of backtest — plus live game state once underway.
The Matchup
The Pittsburgh Pirates (46-45) travel to Yankee Stadium on July 21, 2026, to face the New York Yankees (50-40) in an interleague matchup that pits a .505-winning-percentage club against one sitting ten games above .500. The DiamondIQ model's estimate gives New York a 55.1% win probability and Pittsburgh a 44.9% chance, a lean that reflects the Yankees' home-field advantage and superior record more than any single personnel edge. Notably, both starting pitchers are listed as TBD, which strips the model of its usual PitchIQ starting-pitcher quality differential — meaning that component of the v2 formula is effectively neutral, and the gap in win probability is being driven primarily by team record and home field.
With starters unannounced, the bullpen picture takes on added relevance as a potential tiebreaker. The two pens are nearly indistinguishable by BullpenIQ score — Pittsburgh sits at 49 and New York at 48 — though the Yankees carry six fresh arms against the Pirates' four, with each side listing two pitchers in heavy usage over the past three games. The closers are worth tracking as well: David Bednar handles save situations for New York, while Gregory Soto fills that role for Pittsburgh. Neither bullpen enters in dominant form, so sustained starter length, whoever those starters turn out to be, would be meaningful. Both rosters are navigating notable IL absences: the Yankees are without Aaron Judge and Giancarlo Stanton from their lineup, while the Pirates are missing Oneil Cruz in the field and Spencer Horwitz at first base.
First-pitch conditions at Yankee Stadium project as mild — overcast, 77 degrees, a light five-mile-per-hour wind blowing right to left, and a 30% precipitation probability that introduces some scheduling uncertainty without guaranteeing disruption. There is no market data in today's feed to assess divergence from the model's 55.1-44.9 split. The one thing to watch entering this game is starter confirmation: once Pittsburgh and New York name their pitchers, the PitchIQ gap — currently zeroed out — could shift the model's lean meaningfully in either direction, and the model leans toward New York on record and venue alone until that information is available.