MLB Picks Today: Predictions & Best Bets for July 3, 2026 | SharpSide

The Friday 13-game MLB slate presents several divergences from market consensus, with our model identifying a significant total run edge of 1.8 runs in the New York Mets @ Atlanta Braves matchup. We also highlight the Pittsburgh Pirates @ Washington Nationals and Minnesota Twins @ New York Yankees games where the model sees actionable value. In addition, a specific F5 total and a batter prop are registering sharp movement.

New York Mets @ Atlanta Braves

The SharpSide model projects 1.8 fewer runs in this matchup compared to the market, and 0.4 more runs toward the home side. Christian Scott, with a 3.2 ERA and 10.60 K/9, takes the mound for New York, while Grant Holmes, carrying a 3.96 ERA and 8.03 K/9, starts for Atlanta. The Mets' offense has been unlucky on balls in play, evidenced by their 0.270 BABIP, prompting our model to regress their offense up by 0.20 runs. Similarly, the Braves, with a 0.284 BABIP, see their offense regressed up by 0.06 runs. Christian Scott's average of 4.50 IP per start means the Mets' bullpen, despite a 3.41 season relief ERA, will be heavily exposed, contributing to the model's -0.25 run nudge for the Mets. This is a series opener of a four-game set, coming after both the Mets (2-8 in their last 10, scoring 3 runs in their last game) and the Braves (2-8 in their last 10, scoring 5 runs in their last game) have struggled recently. The park and weather conditions are neutral, with a score of 0. The significant gap in projected run totals, driven primarily by pitching and bullpen assumptions, is the largest factor in the model’s lean.

Lean: UNDER 9.5 total

Pittsburgh Pirates @ Washington Nationals

Our model projects 0.5 more runs in this contest than the market consensus, with a significant 1.1 additional runs toward the Washington Nationals. Mitch Keller, with a 4.87 ERA and 6.96 K/9, starts for Pittsburgh. Foster Griffin, posting a 2.93 ERA and 8.97 K/9, gets the nod for Washington. The Pirates' offense has benefited from good fortune, reflected in their 0.320 BABIP, which leads the model to regress their offense down by 0.30 runs. The Nationals, conversely, have a neutral BABIP of 0.289. The Pirates bullpen, with a 4.37 season relief ERA and a fatigue index of 0.12, receives a +0.24 run nudge from the model, indicating potential struggles. This game marks the series opener of a three-game set. Both teams are 5-5 in their last 10 games, with the Pirates winning their last outing by scoring 6 runs, and the Nationals winning theirs by scoring 10 runs. There are no notable park or weather conditions. The substantial sabermetric regression of the Pirates' offense, combined with bullpen concerns, was the primary driver of the model's projection.

Lean: Washington Nationals RL -1.5

Minnesota Twins @ New York Yankees

The SharpSide model projects 0.8 more runs in this game than the market, but 0.4 fewer runs toward the home side. Mike Paredes, with a 4.26 ERA and 4.62 K/9 over a modest 25.3 IP, starts for the Twins. Gerrit Cole, with a 4.06 ERA and 8.12 K/9 over a modest 37.7 IP, starts for the Yankees. Paredes’ average of 4.22 IP per start will expose the Twins' bullpen, which holds a 5.37 season relief ERA, although the model only applies a -0.28 run nudge. The Yankees' bullpen, despite a 3.19 season relief ERA, receives a +0.39 run nudge from the model. The Twins' 0.287 BABIP suggests some offensive unluckiness, leading to a +0.03 run regression upward. The Yankees, with a 0.277 BABIP, are also due for offensive regression upward by +0.13 runs from the model. This is the series opener of a three-game set. The Twins are 5-5 in their last 10 games, having scored 8 runs in their last outing, while the Yankees are 2-8 in their last 10, scoring only 2 runs in their last game. Park and weather conditions are neutral, with a score of 17. The model's projection here is predominantly influenced by the combined offensive regression upward for both teams and concerns regarding both bullpens.

Lean: OVER 9.5 total

First Five Innings: New York Mets @ Atlanta Braves (F5 UNDER)

The F5 total for the New York Mets @ Atlanta Braves game shows a significant divergence. Our model projects a 3.69 run total for the first five innings, against a market-implied 4.94 runs. Christian Scott's F5-only ERA, blended with his season ERA, comes to 3.47, while Grant Holmes' blended F5-only ERA is 4.31. Both offenses have been unlucky on balls in play this year; the Mets (0.270 BABIP) see their offense regressed up by 0.20 runs, and the Braves (0.284 BABIP) see theirs regressed up by 0.06 runs. However, the early-game damping and park factor (0.930 multiplier) contribute to a lower F5 expectation from our model.

Lean: F5 UNDER 4.5 total

Batter Prop: Jose Altuve Over 1.5 Total Bases

The market for Jose Altuve's Total Bases Over 1.5 in the Tampa Bay Rays @ Houston Astros matchup has seen notable line movement. The price moved from +149 to +127, representing a shift of 3.9 probability points toward the OVER. This aggressive movement, observable across multiple books, is historically indicative of sharp money entering the market. While we do not speculate on the underlying reasons for this specific shift, such consensus-moving action typically flags a high-conviction play.

Lean: Jose Altuve Over 1.5 Total Bases

How the model decided

Today's largest divergences were driven by a combination of factors. The New York Mets @ Atlanta Braves game's UNDER lean was heavily influenced by a cautious stance on starting pitching and adjusted bullpen performance, significantly lowering projected run totals compared to market consensus. In the Pittsburgh Pirates @ Washington Nationals game, the model's strong lean toward the Nationals on the run line stemmed from a substantial negative BABIP regression for the Pirates’ offense. Finally, the Minnesota Twins @ New York Yankees OVER was a result of upward BABIP regression for both offenses, in addition to nuanced bullpen evaluations. The model also considered advanced Sabermetric data, systems trends, and other analytics.

How to read this

This analysis represents the SharpSide model's projection against current market consensus. These are statistical opportunities identified by our systems, not financial advice.