MLB Picks Today: Predictions & Best Bets for July 22, 2026 | SharpSide

Wednesday presents a 17-game slate with several model-market divergences worth noting. Our model identifies a significant 2.6-run difference in the Washington Nationals and Colorado Rockies total, the largest single run-edge of the day. We also highlight key plays in Athletics at Arizona Diamondbacks and Baltimore Orioles at Boston Red Sox. This brief refreshes at 11:30 am Eastern based on new analytics, weather factors, pitching changes, and line movement, so featured matchups may update between morning publish and the refresh. Today's Best Bet is highlighted immediately below.

Today's Best Bet: New York Yankees ML

Today's Best Bet represents the single game and line SharpSide has the highest confidence in. We're targeting the New York Yankees on the moneyline at -171 against the Pittsburgh Pirates. Our model projects the New York Yankees to win by 1.1 runs (4.0–5.1), which is roughly in line with the market's implied margin. This is not necessarily an edge play against consensus, but rather a conviction play where our model's projected outcome for the Yankees is most decisive, reflecting strong confidence in their ability to secure a win. Best Bet: New York Yankees ML (-171)

Washington Nationals @ Colorado Rockies

The market consensus projects a much higher-scoring game between the Nationals and Rockies, with our model showing a substantial 2.6-run total edge against the market. Cade Cavalli starts for Washington, bringing a 3.21 FIP, which indicates he's performed better than his ERA suggests by stripping out defense, and a 0.77 HR/9, demonstrating strong home run suppression. Opposing him for Colorado is Gabriel Hughes, with a 2.20 FIP and an impressive 0.00 HR/9 across a small 14.3 IP sample, indicating elite early-career performance at preventing hard contact. Heading into today's rubber match, the Nationals took game one 7-3, but the Rockies responded with an 8-7 victory in game two. Colorado's offense has been experiencing some good fortune this season with a .307 BABIP, meaning the model regresses their projected runs down by 0.27, pushing the side away from Colorado and the total down. Both starting pitchers average short outings (Cavalli 5.00 IP/start, Hughes 4.78 IP/start), which heavily exposes both bullpens. The Washington bullpen has a 5.13 season relief ERA, worse than the league average of 4.10, and a taxed fatigue index of 0.23, leading to a model run nudge of +0.40 towards the opposing offense. Similarly, the Colorado bullpen has a 4.87 ERA and a taxed fatigue index of 0.32, also prompting a +0.40 run nudge for the opponent. There are no notable park or weather conditions for this game. How the model decided: The largest factor driving our lean is the model's regression of Colorado's offense due to their lucky .307 BABIP, which subtracts 0.27 runs from their projection. This, combined with Gabriel Hughes's excellent 2.20 FIP, significantly pushes the total down. The short average outings by both starters also expose a pair of struggling bullpens, factoring into the model's overall projection. Lean: Colorado Rockies ML (+116)

Athletics @ Arizona Diamondbacks

Our model projects a 0.7-run total edge over the market in the Athletics-Diamondbacks game, alongside a 1.9-run margin edge toward the Diamondbacks. Gage Jump takes the mound for the Athletics, showing strong underlying numbers with a 3.10 FIP and a very good 0.302 xwOBA-against, indicating weak contact allowed, over 52.3 innings pitched. Merrill Kelly starts for Arizona, but his 5.89 FIP and exceptionally high 0.403 xwOBA-against suggest he's allowing very hard contact and is due for negative regression. He also has a high 1.92 HR/9, indicating a propensity to give up home runs. Heading into the final game of this series, the Athletics took game one 5-2, with the Diamondbacks winning game two 6-5. The Athletics arrive with poor recent form, standing at 2-8 in their last 10, while the Diamondbacks are playing well, 7-3 over their last 10 games. Although the Athletics have a slightly lucky .296 BABIP, which causes the model to regress their offense down by 0.09 runs, the Diamondbacks have been unlucky on balls in play with a .275 BABIP, leading the model to add 0.21 runs to their offensive projection. This regression pushes the side toward Arizona and the total up. Arizona's bullpen is stronger than the league average with a 4.00 season relief ERA, but its 0.39 fatigue index suggests it is taxed, leading to a model run nudge of +0.40. There are no notable park or weather conditions. How the model decided: Merrill Kelly's poor 5.89 FIP and 0.403 xwOBA-against are significant drivers, pushing the opposing offense's run projection up. The Diamondbacks' unlucky .275 BABIP also adds 0.21 runs to their offensive projection, further influencing the total and side. The most substantial factor is Kelly's underlying poor metrics, which project more runs for the Athletics. Lean: Athletics ML (-100)

Baltimore Orioles @ Boston Red Sox

The model identifies a considerable 1.7-run total edge against the market in the Baltimore Orioles and Boston Red Sox game, leaning toward the over. Dean Kremer starts for the Orioles, and his 6.18 FIP and alarming 3.12 HR/9 over 26.0 innings suggest he's due for significant negative regression and allows a high rate of home runs. Jake Bennett takes the mound for the Red Sox, showing a solid 3.03 FIP and a promising 0.269 xwOBA-against over 53.7 innings, indicating he's effective at limiting hard contact. This is game two of a three-game set, with the Boston Red Sox taking game one 6-5. The Orioles enter this contest in strong form, with a 7-3 record over their last 10 games, while the Red Sox are exceptional, riding a 10-0 run. Kremer averages a short 5.00 IP per start, heavily exposing the Orioles bullpen. While their 4.15 season relief ERA is close to league average, their -0.22 fatigue index indicates they are rested, leading to a model nudge of -0.40 runs. The Red Sox bullpen has a strong 3.09 season relief ERA, below the league average of 4.10, and a relatively rested fatigue index of -0.12, resulting in a model nudge of -0.19. The park and weather conditions provide a score of 15, indicating no notable factors. How the model decided: Dean Kremer's significantly high 6.18 FIP and 3.12 HR/9 are the primary drivers pushing the total up, indicating his vulnerability to allowing runs. While the Orioles bullpen is rested and provides a slight run suppression, the market underestimation of the game's total is heavily influenced by Kremer's poor underlying metrics. The Red Sox' strong recent form also factors into the overall projection. Lean: OVER 9.5 total (-112)

First Five Innings: Washington Nationals @ Colorado Rockies (F5 UNDER)

The model projects a F5 total of 3.58 runs for the Nationals and Rockies, significantly lower than the market-implied 5.98 runs. This 2.40-run edge points toward an F5 UNDER. Cade Cavalli's blended F5-only ERA is 3.78, while Gabriel Hughes's blended F5-only ERA is 3.76, even after shrinking his small sample size toward league average. Colorado's offense has also been lucky on balls in play with a .307 BABIP, which prompts the model to subtract 0.27 runs from their projection, further reinforcing the under. Lean: F5 UNDER 5.5 total

Batter Prop: Heliot Ramos Over 1.5 Total Bases

Sharp money has moved the line on Heliot Ramos's Total Bases Over 1.5 against the Kansas City Royals. The price has shifted from +112 to -103, a 3.6 probability-point move toward the OVER since opening. This aggressive line movement across various books historically flags significant institutional action, indicating smart money entering the market on Ramos to exceed 1.5 total bases. Lean: Heliot Ramos Over 1.5 Total Bases (-103)

How the model decided

Today's largest divergences from market consensus were primarily driven by a combination of starter quality metrics like FIP and xwOBA-against, alongside BAPIP luck regression and bullpen fatigue. In the Nationals-Rockies game, the model's run projection was heavily influenced by Colorado's lucky BABIP regressing their offense downward, coupled with strong underlying metrics from Gabriel Hughes. For the Athletics-Diamondbacks matchup, Merrill Kelly's high FIP and xwOBA-against were significant factors, projecting more runs for the opposing offense, while the Diamondbacks' unlucky BABIP added runs to their side. The Baltimore-Boston total was predominantly pushed OVER due to Dean Kremer's very high FIP and HR/9. The model also considered advanced Sabermetric data, systems trends, and other analytics.

How to read this

This analysis reflects SharpSide's proprietary model's view compared to current market odds and should be used for informational purposes only. It is not financial advice.