Chris Paddack counters for Miami with a 6.14 ERA through 14.2 innings, a figure shaped almost entirely by one disastrous outing on March 30 (4.0 IP, 8 ER vs Chicago). Strip that start out and his recent trajectory changes shape. He went 4.2 shutout innings against New York, then 6.0 innings and 2 earned runs against Detroit. His K/9 in 2026 sits at 8.59, and his walk rate has tightened in those later appearances. The underlying contact suppression appears to be returning. Which version shows up in Atlanta tonight is the real question on Miami's side of the ledger.

The batter-vs-pitcher splits are the sharpest part of this card and they cut in both directions. Xavier Edwards has posted a .571 AVG and 1.196 OPS against Elder in 8 career plate appearances, with his 2025 sample (6 PA) showing an even stronger 1.267 OPS. Otto Lopez adds a .500 AVG and 1.000 OPS in 6 career plate appearances against Elder, all from 2025. Both sit near the top of Miami's lineup. On Atlanta's side, Mike Yastrzemski is a different kind of matchup problem for Paddack, having hit .391 with a 1.353 OPS and 3 home runs across 25 career plate appearances against him spanning five seasons. Austin Riley, by contrast, is 0-for-3 in his 2025 sample against Paddack with a 0.000 OPS in those appearances. Michael Harris II is 0-for-4 in 2025 plate appearances against this arm with the same result. Paddack generates enough weak contact to neutralize Atlanta's middle-of-the-order threats, even as Yastrzemski represents a genuine threat every time up.

Atlanta enters at 7-4 at home with a team ERA of 2.93 and a bullpen sitting at 3.29, the best relief unit in this series. Miami is 2-6 away from home on the season, though you would not know it from this week: a 10-4 blowout in Game 1 followed by a competitive 5-6 loss in Game 2 where Dominic Smith delivered the decisive late-inning hit for Atlanta. Both bullpens are taxed. Truist Park plays at a neutral run environment with a runs factor of 1.0 and an HR factor of 1.02, so the park is not going to tilt this in any direction. Our model projects Atlanta 4.5, Miami 3.7, for 8.2 combined runs. The market is sitting at 9.0. That 0.8-run gap is not a rounding error.

• Bryce Elder's 1.02 ERA in 2026 is built on three clean outings against Cleveland, Arizona, and Oakland. In his two most recent starts against Miami (August 2025) he allowed exactly 5 earned runs each time across 6.0 and 5.1 innings.
• Xavier Edwards (.571 AVG, 1.196 OPS in 8 career PA vs Elder, including a 1.267 OPS in his 2025 sample) and Otto Lopez (.500 AVG, 1.000 OPS in 6 career PA vs Elder, all from 2025) sit near the top of Miami's order with the two strongest BvP edges in this game.
• Paddack's 6.14 ERA is almost entirely one start. His last two outings: 4.2 shutout innings vs New York, then 6.0 innings and 2 earned runs vs Detroit. Contact suppression appears to be improving.
• Mike Yastrzemski has hit .391 with a 1.353 OPS and 3 home runs in 25 career plate appearances against Paddack. Austin Riley is 0-for-3 with a 0.000 OPS in his 2025 sample against this arm, the weakest recent matchup among Atlanta's regulars.
• Atlanta's bullpen leads this series at 3.29 ERA, but both clubs have burned through relief innings over two games. Game 3 depth management will shape the late innings regardless of starter performance.
• Our model gives Atlanta a 60.5% win probability. The market implies 61.7%. That under-2% gap offers no exploitable edge on the moneyline in either direction.

The run line pairs with the total naturally. Our model projects a one-run Atlanta margin, and Marlins +1.5 at -155 covers that exact outcome. The moneyline has no play on either side: the market's implied 61.7% for Atlanta and our model's 60.5% are functionally identical. If you want to explore the contrarian case for Miami at +148, the argument lives in Edwards and Lopez, two hitters who have genuinely owned Elder in recent matchup data. But the model does not support betting against Atlanta's projected win probability at that price, and neither do I. The run line is the right way to get Miami exposure if you want it.

The prop card adds sharper edges. Yastrzemski's .391 average and 1.353 OPS in 25 career plate appearances against Paddack make him Atlanta's single most dangerous bat tonight, regardless of his current season line. On the Miami side, Edwards at +105 for over 1.5 total bases and Lopez at +120 for the same represent the clearest value against Elder's specific BvP weakness. This game is worth playing on the total, the run line, and the props above. Variance exists on every MLB game. Bet accordingly. For the full slate, see our MLB picks today, NRFI picks, and player props.