MLB Picks Today: Predictions & Best Bets for August 19, 2026 | SharpSide

Wednesday's 15-game MLB slate presents several intriguing matchups, with the SharpSide model identifying significant divergence from market consensus in key areas. Our analysis pinpoints three matchups where this divergence is most pronounced: the Miami Marlins at Philadelphia Phillies, the Washington Nationals at Texas Rangers, and the Arizona Diamondbacks at Boston Red Sox. The largest single run-edge the model identifies today is a 1.4-run advantage for the Miami Marlins against the Philadelphia Phillies. This brief refreshes at 11:30 am Eastern based on new analytics, weather factors, pitching changes, and line movement, so the featured matchups may update. Today's Best Bet is highlighted immediately below.

Today's Best Bet: Tampa Bay Rays ML

SharpSide's highest-confidence play today is on the Tampa Bay Rays to win their home matchup against the Toronto Blue Jays. The model projects the Tampa Bay Rays for a moderate 1.1-run victory, with a final score of 4.5–3.4. This prediction is broadly consistent with market expectations, indicating this is a conviction play rather than one driven by a significant market mispricing. The model is most assured of the Tampa Bay Rays' likelihood of winning this contest. Best Bet: Tampa Bay Rays ML (-208)

Miami Marlins @ Philadelphia Phillies

The SharpSide model diverges from consensus in this matchup, projecting a 4.6–4.1 victory for the Miami Marlins, while the market favors the Philadelphia Phillies 4.7–3.8. This represents a 1.4-run shift in projected margin favoring the Marlins. Miami starter Sandy Alcantara brings a strong 3.79 FIP and an impressive 0.294 xwOBA-against, indicating he allows weak contact. In contrast, Philadelphia's Aaron Nola has struggled, with a 5.04 FIP and a concerning 2.08 HR/9, suggesting he is prone to giving up home runs. Alcantara's average of 6.81 IP per start further limits exposure to the Marlins' bullpen, which holds a 3.98 season relief ERA, better than the league average of 4.10. Nola's average of 5.24 IP per start means the Phillies' bullpen, with its 4.37 season relief ERA and a taxed fatigue index of 0.20, will be exposed for more innings. Philadelphia's offense, with a .284 BABIP, has been unlucky on balls in play and is regressed up by 0.09 runs, while Miami's offense, with a .299 BABIP, has been lucky and sees a regression of -0.13 runs. The series continues at Citizens Bank Park, where the Phillies have won both prior games (6-5 and 6-4). Both teams have middling recent form, with the Marlins 6-4 and the Phillies 7-3 over their last 10 games. How the model decided: Alcantara's superior FIP and xwOBA-against pushed the Marlins' run projection up and the Phillies' down. The Phillies' taxed bullpen and Nola's shorter average outing further contributed to the model's lean. The largest driver was the significant gap in starter quality favoring Alcantara. Lean: Miami Marlins ML (+109)

Washington Nationals @ Texas Rangers

The model projects the Washington Nationals to defeat the Texas Rangers 4.0–3.8, a stark contrast to the market's 4.2–3.3 projection for the Rangers. This indicates the model is 1.1 runs more favorable to the Nationals than the market is. Washington sends Cade Cavalli to the mound, who boasts a solid 3.46 FIP and a 9.94 K/9, suggesting he suppresses runs effectively and generates strikeouts. Texas counters with Kumar Rocker, whose 4.37 FIP and 1.24 HR/9 indicate a higher susceptibility to runs and home runs. Cavalli's 5.50 IP/start is slightly above average, while Rocker's 5.00 IP/start is short, exposing the Texas bullpen more significantly. The Rangers' bullpen, with a 4.32 season relief ERA, is slightly worse than league average. The Nationals' bullpen has a 4.84 season relief ERA but enters the game rested with a -0.14 fatigue index. Both teams enter with neutral BABIP luck, suggesting no significant offensive regression in play. The Rangers secured a 5-0 victory in Game 1 of this series. Both the Nationals and the Rangers have struggled recently, holding 4-6 records over their last 10 games. Park conditions are neutral with no notable factors. How the model decided: Cavalli's stronger FIP and K/9 pushed the Nationals' projection up and the Rangers' down. Rocker's shorter average outing exposing the Rangers' weaker bullpen was also a significant factor. The largest driver was the discrepancy in projected starter performance favoring Cavalli. Lean: Washington Nationals ML (+117)

Arizona Diamondbacks @ Boston Red Sox

The SharpSide model projects a lower-scoring affair for the Arizona Diamondbacks at Boston Red Sox, predicting a 4.1–3.7 victory for the Red Sox. This total of 7.8 runs is 0.7 runs lower than the market's consensus total of 8.5 runs. Arizona's Brandon Pfaadt comes into this game with a 3.80 FIP and a solid 0.293 xwOBA-against, indicating good control and weak contact allowed. Boston's Payton Tolle is even stronger, featuring a 3.21 FIP and an excellent 0.255 xwOBA-against, showcasing a talent for suppressing contact quality and runs. Both starters average over 5.5 innings per start, with Pfaadt at 6.07 and Tolle at 5.67. The Red Sox bullpen, despite a strong 2.93 season relief ERA, is taxed with a 0.32 fatigue index, potentially leading to more runs. The Diamondbacks' bullpen is close to league average with a 4.21 ERA and is mildly fatigued. Arizona's offense is missing core hitter Ketel Marte, leading to a ~0.18-run reduction in their projection. Their .280 BABIP suggests they have been unlucky on balls in play and are due for an upward regression of 0.15 runs. Conversely, Boston's .298 BABIP indicates some luck, leading to a -0.12-run offensive regression. Park conditions are neutral for this game. The Red Sox have dominated the first two games of this series, winning 11-1 and 9-4. Both teams enter with struggling 4-6 records over their last 10 games. How the model decided: Tolle's elite FIP and xwOBA-against against a Diamondbacks lineup missing Ketel Marte were significant factors in suppressing the total. The Red Sox' lucky BABIP, due for downward regression, further contributed. The largest driver for the UNDER lean was the combination of strong starting pitching from both sides, particularly Tolle. Lean: UNDER 8.5 total (-114)

First Five Innings: Arizona Diamondbacks @ Boston Red Sox (F5 UNDER)

The First Five Innings total for the Arizona Diamondbacks at Boston Red Sox presents another strong UNDER opportunity. Our Bayesian starter model projects a low 2.77 runs for the F5, significantly below the market's implied 4.42 runs. This substantial -1.65 run edge is driven by the strong pitching matchup. Brandon Pfaadt, with his F5-only ERA of 2.65, faces Payton Tolle, who boasts an F5-only ERA of 3.18. Both pitchers have demonstrated an ability to limit early-game scoring. Additionally, the Red Sox' offense has benefited from a lucky BABIP this season, leading the model to regress their early-game run projection downward. Lean: F5 UNDER 4.0 total

How the model decided

Today's sharpest divergences from market consensus are driven by a combination of factors across the slate. In the Marlins-Phillies game, the model significantly favors Miami due to the marked difference in starting pitcher quality, with Sandy Alcantara's strong underlying metrics outperforming Aaron Nola's. The model also accounts for bullpen fatigue and Nola's shorter average outings, exposing a taxed Philadelphia bullpen. For the Nationals-Rangers matchup, Cade Cavalli's strong FIP and strikeout rate against Kumar Rocker's higher FIP and home run rate push the model towards Washington, exacerbated by Rocker's short average workload. The UNDER lean in the Diamondbacks-Red Sox game is largely due to the excellent FIP and xwOBA-against numbers from both Brandon Pfaadt and Payton Tolle, as well as the Red Sox' offensive BABIP regression and the Diamondbacks' missing core hitter. The model also considered advanced Sabermetric data, systems trends, and other analytics.

How to read this

This analysis reflects the SharpSide model's projected outcomes and divergences from the market, not financial advice. All decisions to place wagers are at the sole discretion of the reader.

Projections update after publish. This brief is locked in at the 11:30 am ET refresh, but the model keeps recalculating all day as lineups are posted, weather shifts, pitching changes are announced, and lines move — a side or total highlighted here can flip before first pitch. Always confirm the current number on the live projections page (which locks 30 minutes before each game's first pitch) before placing a bet.