Data-Driven MLB Predictions for 2026

MLB 2026 Spring Training Analytics: Sasaki's Pitch Development, Braves' Rotation Crisis, and Judge's Elbow Recovery

February 16, 2026

Spring training camp reports are often dismissed as noise by serious analysts. And most of the time, that's the correct posture. Exhibition results, batting practice home runs, and "best shape of his life" narratives rarely translate into meaningful regular-season outcomes. But occasionally, the early February and March intel contains genuine signal worth isolating. This week, three developments out of MLB camps carry real analytical weight for anyone modeling 2026 projections: Roki Sasaki's pitch arsenal expansion in Los Angeles, Atlanta's compounding rotation losses, and Aaron Judge's elbow recovery timeline in New York.

Sasaki's Pitch Mix Problem, and the Statistical Case for a Third Offering

Roki Sasaki's talent has never been in question. His fastball velocity and his splitter's devastating vertical break make him one of the most electric arms in baseball when everything is working. But the analytical case against Sasaki as a true frontline starter has always centered on one limitation: he's been essentially a two-pitch pitcher. Fastball, splitter, and not much else that hitters needed to respect.

The downstream effects of a two-pitch arsenal are well-documented in modern pitching analytics. When a starter lacks a reliable third offering, hitters can begin eliminating pitches from their decision tree during second and third plate appearances. They sit on the fastball, knowing the only alternative is the splitter. This forces the pitcher into predictable patterns. For Sasaki specifically, the result was falling behind in counts too often and being forced back to the fastball when he couldn't get hitters to chase the splitter. That sequence led to longer at-bats, elevated pitch counts, and shorter outings, exactly the outcomes you'd predict from a pitcher whose arsenal lacks the complexity to keep hitters off-balance through multiple trips through the order.

This is why the spring training reports about Sasaki's pitch development carry legitimate analytical significance. He's actively throwing cutters and sliders, with a stated focus on what he describes as the "gyro-spin slider." He's also incorporating a two-seamer to complement his four-seam fastball. From a pitch design perspective, each of these additions addresses a specific gap in his arsenal. The cutter gives him a weapon that looks like the fastball out of his hand but moves laterally away from right-handed hitters, something the splitter does not do. The gyro-spin slider would give him a breaking ball with a different shape and velocity profile than the splitter, forcing hitters to distinguish between three off-speed pitches instead of one. And the two-seamer adds horizontal arm-side movement, giving left-handed hitters a pitch that runs in on their hands rather than one that drops straight down.

If even one of these pitches develops into something Sasaki can throw with confidence in any count, the statistical model for his 2026 projection changes significantly. A third reliable pitch should, in theory, reduce his times-through-the-order penalty, lower his pitch counts, and extend his outings. For the Dodgers' rotation depth models, Sasaki going from a 5.0-inning average to a 6.5-inning average per start would meaningfully reduce the workload on their bullpen and improve their run prevention across an entire season. The Dodgers coaching staff clearly understands this, which is why the development focus this spring is specifically on arsenal expansion rather than mechanical refinement.

Atlanta's Rotation Losses: Compounding Risk in the Projection Models

The Braves entered spring training with legitimate rotation depth questions, and the early camp news has made those questions dramatically worse. Two developments in particular are forcing significant downward revisions in any honest projection model for Atlanta's 2026 season.

First, Spencer Schwellenbach has been placed on the 60-day injured list with right elbow inflammation. The 60-day IL designation is important from a modeling perspective because it establishes a hard floor for his absence. He cannot return before roughly mid-May at the earliest, and elbow inflammation in young pitchers historically trends toward extended timelines rather than aggressive returns. The Braves were projecting Schwellenbach as a rotation contributor in 2026, meaning his innings now need to be absorbed by pitchers who were not part of the original plan. Replacement-level innings are, by definition, less valuable than the innings the player they're replacing was projected to provide.

Second, Hurston Waldrep's situation introduces additional uncertainty. Manager Walt Weiss disclosed that an MRI revealed "loose bodies" in Waldrep's arm and that he may require surgery to remove them. The encouraging finding is that no ligament damage was detected, which rules out the most catastrophic outcome. Waldrep is scheduled to consult with Dr. Keith Meister for a final determination. Even in the optimistic scenario where the procedure is straightforward and recovery proceeds on schedule, Waldrep's spring ramp-up is disrupted. He'll be behind in his pitch count buildup, his timing will need recalibration, and his early-season readiness becomes a genuine question mark.

The compounding effect of losing both arms simultaneously is what makes this so damaging in the projection models. Losing one young starter can be absorbed through depth. Losing two simultaneously forces a team to rely on its fifth and sixth pitching options for sustained stretches, and the expected run prevention from those arms is measurably worse. The gap between a team's number three starter and its number six option can easily represent half a run per nine innings or more, and over the 30 to 40 starts those replacement arms might collectively make, that gap translates into multiple additional losses over the course of a season. Atlanta's projected win total needs to come down, and it needs to come down more than the single-arm injury adjustment most models would apply, because the cascading effect of two simultaneous rotation losses is greater than the sum of the individual impacts.

Judge's Elbow Recovery: Health Data Points and the Workload Question

Aaron Judge has been fully cleared for spring training activities after the flexor strain in his right elbow that developed late in the 2025 season resolved without surgical intervention. The MRI showed no structural damage requiring repair, and the elbow healed through rest and conservative treatment. Judge reported no residual issues, stating that he's been throwing confidently, including throws to bases, with no limitations. He is scheduled to appear in four or five of the first nine Grapefruit League games, which represents a standard early spring workload for an established veteran.

From a health analytics perspective, the fact that the flexor strain resolved without surgery is the most significant data point. Flexor strains exist on a continuum. The mild end involves inflammation and micro-tears that heal with rest. The severe end involves significant tears that require surgical repair and months of rehabilitation. Judge's case falling on the mild end of that spectrum means there is no structural compromise to account for in projection models, and his 2026 baseline should reflect his full healthy-season talent level rather than a post-surgical recovery discount.

The one variable that introduces analytical uncertainty is Judge's plan to leave camp on March 1 to join Team USA for the World Baseball Classic. From a workload modeling perspective, the WBC represents a disruption to the normal spring training progression. Players typically use the six weeks of spring training to systematically build their pitch counts (for pitchers) and at-bat volumes (for hitters) toward regular-season readiness. The WBC compresses this timeline by inserting competitive, high-intensity games into the middle of the ramp-up period, and the travel and schedule demands of the tournament add physiological stress that exhibition games do not.

For Judge specifically, the question is whether his body, which has a documented history of durability challenges, responds differently to the WBC workload than it would to a conventional spring training buildup. The sample size of WBC participants who subsequently had injury issues during the regular season is too small to draw statistically meaningful conclusions. But the directional risk is real enough that any responsible projection model should at least flag it as a potential variance factor in Judge's first-half availability estimates. The upside, of course, is that Judge entering the season with competitive at-bats already under his belt could accelerate his early-season performance, reducing the "spring training hangover" effect that some sluggers experience in April.

The Preller Extension: An Organizational Signal Worth Noting

In a smaller but analytically relevant development, A.J. Preller agreed to a multiyear extension with the Padres as president of baseball operations. From a roster construction modeling standpoint, this matters because Preller's track record as one of baseball's most aggressive in-season dealmakers means the Padres' roster at the trade deadline could look dramatically different from their Opening Day roster. Any projection system that treats San Diego's current roster as static through the full season is likely underestimating their second-half ceiling. Preller's extension is an organizational signal that the Padres intend to compete, and that willingness to acquire at the deadline introduces upside variance that standard preseason models do not capture.

Synthesizing the Spring Training Signal

The challenge of spring training analysis is always separating signal from noise. Most of what happens in February and March camps is meaningless for regular-season projections. But the three developments examined here, Sasaki's deliberate arsenal expansion, Atlanta's compounding rotation losses, and Judge's clean bill of health paired with WBC workload considerations, all represent genuine inputs that should move projection models in specific, quantifiable directions. The Dodgers' pitching depth projections should improve if Sasaki's third pitch development shows progress. The Braves' win total projections need downward revision to account for the simultaneous loss of two young starters. And the Yankees' projections should reflect both the positive signal of Judge's full health and the uncertainty introduced by the WBC schedule. These are not hot takes or narrative-driven overreactions. They are data-informed adjustments based on new information, which is exactly what good projection models are designed to incorporate.

Joey Loperfido Jesus Sanchez Trade Statistical Breakdown: OPS Splits, wRC+ Analysis, and What the Advanced Numbers Say About This Astros Blue Jays Outfield Swap

February 14, 2026

On February 13, the Houston Astros reacquired outfielder Joey Loperfido from the Toronto Blue Jays in exchange for outfielder Jesus Sanchez. At first glance, this is a straightforward swap of underperforming outfield depth pieces. But when you pull the numbers apart and examine the statistical profiles of both players, a more nuanced picture emerges. This is a trade where both organizations are making calculated bets on opposing sides of the same coin: sample size versus true talent level.

The Loperfido Statistical Profile: A Tale of Two Seasons

Joey Loperfido's MLB career spans 122 games between Houston and Toronto, and the year-over-year splits are dramatic. In 2024 with the Blue Jays, he struggled to a .197/.236/.343 slash line with 2 home runs and 9 RBI across 43 games. The .343 slugging percentage suggested a player who simply was not making enough quality contact to survive against major league pitching. The .236 on-base percentage, in particular, pointed to an approach that was overly aggressive and easily exploitable.

Then came 2025, and the numbers flipped entirely. In 41 games with Toronto, Loperfido posted a .333/.879 OPS line with 4 home runs and 14 RBI. That .879 OPS represents a staggering improvement. To put that in context, an .879 OPS over a full season would place a hitter comfortably in the top 20 percent of all MLB hitters offensively. The average jumped 136 points from .197 to .333. The power output, while modest in raw terms at 4 home runs, was produced in roughly half the games of his 2024 campaign, suggesting a significant per-game power improvement.

The Critical Question: Which Loperfido Is Real?

This is where the analytics get interesting and where the Astros are making their bet. The 41-game 2025 sample is small enough that regression models would naturally pull his projections back toward a mean somewhere between his two seasons. If you weight both campaigns equally by games played, you get a combined 84-game line that falls somewhere in the .260 to .270 range with modest power, which is a useful fourth outfielder profile. But the Astros are not weighting equally. They're betting that the 2025 version represents genuine development, not statistical noise. The fact that Loperfido was originally a seventh-round pick in Houston's 2021 draft means the organization has proprietary scouting data from his developmental years that no public model can replicate. They saw something in his swing mechanics, his approach, or his physical maturation that made them willing to bring him back. That's a data point worth taking seriously, even if it doesn't show up in a spreadsheet.

The Sanchez Statistical Profile: Volatility Personified

Jesus Sanchez presents one of the more perplexing analytical profiles in baseball right now. His Miami Marlins career, spanning roughly 1,300 plate appearances from 2023 onward, produced a .253/.319/.428 slash line. That translates to a wRC+ that sits in the range of a league-average to slightly above-average offensive contributor, with the power component (.428 SLG) providing the primary value. As a left-handed hitting outfielder with that kind of slugging ability, Sanchez's Miami numbers painted the picture of a solid everyday player with upside.

Then Houston happened. After being acquired from Miami at the 2024 trade deadline for Ryan Gusto and two minor leaguers, Sanchez's 2025 numbers with the Astros were alarming: .199/.269/.342 in 160 plate appearances across 48 games. That .342 slugging percentage is almost 90 points below his Miami mark. The 71 wRC+ means he was producing at roughly 29 percent below league-average offensive output. The 4 home runs and 12 RBI in 48 games represent a significant power drop-off from the rates he established in south Florida.

Decomposing the Houston Struggles: Environment or Decline?

The central analytical question for Toronto is whether Sanchez's Houston numbers represent a true talent decline or an environmental adjustment issue. His career .239 average provides a useful anchor point that sits between his Miami highs and his Houston lows. Several factors could explain the Houston regression without requiring the conclusion that Sanchez simply got worse at hitting. A change in lineup position, different hitting coaches, unfamiliarity with American League pitching staffs, or even the psychological pressure of performing for a contending team after coming from a rebuilding one, all of these variables can temporarily suppress a player's production without reflecting a genuine skill change.

The 160 plate appearance sample size also matters significantly here. Statistical stabilization points for most batting metrics require considerably more than 160 plate appearances to be considered reliable. Batting average typically needs 500 to 600 plate appearances to stabilize. Slugging percentage requires a similar threshold. Even strikeout rate, one of the quicker-stabilizing metrics, needs roughly 150 plate appearances to be reasonably predictive. Sanchez's Houston sample sits right at the edge of that minimum threshold for the most stable metrics and falls well short for the noisier ones. Drawing firm conclusions about his true talent level based on this sample alone would be a statistical error.

Projecting the Trade Value: A Numbers-First Assessment

When you strip away the narratives and focus strictly on what the data supports, this trade looks like a reasonable bet for both sides. Houston is acquiring a player whose most recent 41-game stretch produced elite offensive numbers (.879 OPS), while accepting the risk that the sample is too small to be fully predictive. Toronto is acquiring a player whose 1,300 plate appearance track record with Miami suggests a .253/.319/.428 true talent level, while accepting the risk that the Houston struggles exposed a genuine limitation.

The expected value calculation for both teams likely comes out positive, which is what makes this trade analytically clean. If Loperfido's 2025 gains are even partially real, and regression pulls him down from .333 to something like .280 with an OPS in the .770 to .800 range, Houston has a useful everyday contributor. If Sanchez's true talent sits halfway between his Miami line and his Houston line, Toronto is looking at a .230/.290/.390 hitter with 15-home-run power, which is a perfectly serviceable fourth outfielder. Both organizations are buying perceived upside at a discount because the most recent data point for each player was either unusually good (Loperfido) or unusually bad (Sanchez). The numbers tell an interesting story here. Neither team is reaching. Both are making calculated bets based on different interpretations of incomplete data, which is exactly what smart front offices do.

Paul Goldschmidt Returns to Yankees on $4 Million Deal: A Statistical Deep Dive Into His Declining Production and What the Advanced Numbers Say About His 2026 Platoon Value

February 7, 2026

The New York Yankees have re-signed Paul Goldschmidt to a one-year, $4 million contract, bringing back the 38-year-old first baseman for a second season in the Bronx. But this is not the Goldschmidt the Yankees signed last year. That version came with a $12.5 million price tag and the expectation that he would be the everyday first baseman. This version comes at less than a third of the cost and is projected for a short-side platoon role after Ben Rice outplayed him for playing time during the 2025 season. Goldschmidt reportedly passed on more money from other teams to stay in New York. From a pure analytics standpoint, the question is straightforward: what does a $4 million investment in a declining 38-year-old actually buy you in projected value?

The Decline Trajectory in Numbers: Goldschmidt's statistical decline has been one of the more dramatic falloffs in recent baseball history. This is a player who won the 2022 NL MVP with the Cardinals, slashing .317/.404/.578 with 35 home runs and a league-leading 178 OPS+. He was one of the best hitters on the planet just four seasons ago. In 2023, the numbers dipped to .268/.365/.441 with 25 home runs. Still productive, but the power was fading and the batting average dropped nearly 50 points. In 2024, his final year in St. Louis, the floor fell out: .245/.302/.414 with 22 home runs in 154 games. The OPS+ cratered to 100, meaning he was exactly league average at the plate. That is a shocking number for a former MVP.

Then came 2025 in New York. Goldschmidt signed a $12.5 million contract to be the everyday first baseman, and the results were worse than anyone projected. By midseason, the Yankees had seen enough and started giving Ben Rice more at-bats at first base. Rice, younger and cheaper, outproduced Goldschmidt in nearly every offensive category. The message was clear: the everyday role was no longer Goldschmidt's.

The Platoon Split Argument: The analytical case for bringing Goldschmidt back rests almost entirely on his platoon splits. Throughout his career, Goldschmidt has been significantly better against left-handed pitching than right-handed pitching. In his prime years, this split was manageable because he was so good against both sides that it did not matter. But as his overall production has declined, the platoon gap has become the critical variable.

Against left-handed pitchers during his career, Goldschmidt has historically posted an OPS roughly 80 to 100 points higher than against right-handers. For a player whose overall bat can no longer support everyday at-bats, limiting him to the favorable side of the platoon becomes the most efficient way to extract remaining value. A platoon-only role means Goldschmidt would face roughly 180 to 220 plate appearances against lefties, shielding him from the right-handed matchups where his bat speed decline is most exposed.

WAR Projection and Dollar-Per-Win Analysis: Here is where the analytics get interesting for the $4 million price tag. A full-time player producing at league-average levels is worth roughly 2.0 WAR over a full season. In a reduced platoon role of 200 to 250 plate appearances, Goldschmidt's projected WAR falls somewhere in the 0.5 to 1.0 range, depending on how effectively the Yankees can deploy him in his high-leverage matchups against lefties.

At $4 million for 0.5 to 1.0 WAR, that puts the cost-per-win in the $4 to $8 million range. In the current free agent market, where a win is valued at approximately $8 to $9 million, this deal is either right at market rate or slightly below depending on where the actual production lands. Compare that to last year's $12.5 million for roughly the same output, and the Yankees clearly got smarter about how they allocated dollars to Goldschmidt's declining skill set.

The Age 38 Season in Historical Context: The data on first basemen playing at age 38 or older is not encouraging. Bill James' age curves show that first basemen typically experience their steepest offensive decline between ages 36 and 38, with power being the first tool to go. The historical batting average for first basemen at age 38 in the modern era hovers around .240 to .250 with significantly reduced extra-base production. Goldschmidt's trajectory aligns almost perfectly with this curve.

Albert Pujols at age 38 hit .245 with 19 home runs. Miguel Cabrera at 38 hit .256 with 15 home runs. Paul Konerko at 38 hit .244 with 12 home runs. The pattern is consistent: even elite first basemen at this age become average-to-below-average hitters. The question is not whether Goldschmidt will decline further. The question is whether the rate of decline makes a $4 million platoon investment worthwhile compared to alternatives.

Defensive Metrics and Baserunning Value: One area where Goldschmidt still provides quietly positive value is his defense at first base. His fielding metrics over the past two seasons have remained slightly above average, with positive Outs Above Average numbers despite his age. He is not the Gold Glove defender he was in his Arizona days, but he remains a reliable glove at first. This matters because in a platoon role where the offensive contribution is marginal, every fraction of a win from other sources becomes meaningful.

On the basepaths, however, the data is less forgiving. Goldschmidt's sprint speed has declined steadily, and his baserunning runs above average have been negative in each of the last three seasons. He is no longer a threat to steal bases or take extra bases on hits to the gap. In a platoon role, this is less of a concern simply because he will be accumulating fewer at-bats and therefore fewer baserunning opportunities, but it is another area where the aging curve is doing what it always does.

Why Goldschmidt Passed on More Money: The fact that Goldschmidt turned down larger offers from other teams to stay with the Yankees at $4 million is the most telling detail in this entire story. It suggests two things. First, the other offers likely came with expectations of a larger role, perhaps everyday at-bats, which would expose his declining skills across a full season of plate appearances. Second, Goldschmidt understands that being a part-time player on a contending team in New York provides something that a full-time role on a lesser team cannot: October baseball. For a player with only one World Series appearance in his career, with the 2023 Cardinals, the chance to compete for a championship apparently outweighed additional salary.

The Analytical Bottom Line: At $4 million, the Yankees are not paying for the Paul Goldschmidt who won an MVP in 2022. They are paying for a veteran platoon bat who provides 0.5 to 1.0 WAR against left-handed pitching, league-average defense at first base, and invaluable clubhouse presence for a team with World Series aspirations. The advanced numbers indicate further decline is virtually certain. His bat speed, exit velocity trends, and age-related performance curves all point in the same direction. But the structure of this deal, reduced salary for a reduced role, shows that the Yankees' analytics department learned from the $12.5 million overpay last season. This is a mathematically sound allocation of resources for a team that already has its core lineup in place and simply needs a reliable left-handed pitching specialist off the bench and in the short side of a first base platoon. At $4 million, the downside is minimal and the upside is a veteran who can still punish lefties in October.

Isaac Paredes Agrees to $9.35M Arbitration Deal With Astros: WAR Projections, Statcast Profile, and What the Advanced Numbers Say About His 2026 Outlook

February 5, 2026

The Houston Astros and infielder Isaac Paredes have avoided arbitration, agreeing to a one-year, $9.35 million deal that includes a $13.35 million club option for 2027. The agreed salary split the difference between Paredes' $9.95 million ask and Houston's $8.75 million offer. From a pure analytics standpoint, this is one of the more interesting arbitration outcomes in baseball this year, because Paredes' traditional numbers and his underlying metrics tell slightly different stories.

The Traditional Stat Line: In 2025, Paredes slashed .254/.352/.458 with 20 home runs in 102 games for the Astros. Those are solid numbers. Good enough to earn an All-Star selection as a replacement for Jose Ramirez. Good enough to justify the $9.35 million salary without much debate. But the surface numbers are where the easy analysis ends.

The Statcast Profile, Where It Gets Complicated: Paredes' 2025 Statcast numbers reveal a profile with real tension between his outcomes and his underlying batted-ball data. His average exit velocity of 87.4 mph is league average at best. His hard-hit rate of 33.6% ranked well below the MLB mean. His barrel percentage of 6.3% is modest for a player whose primary value comes from his bat. These are not numbers you typically associate with a 20-homer hitter.

Here is where it gets analytically interesting. Paredes posted a .353 wOBA against a .320 xwOBA. That 33-point gap suggests he outperformed his batted-ball quality in 2025, meaning some of those results were driven by sequencing, timing, and a healthy dose of situational fortune rather than raw impact. His BABIP ran higher than expected given his exit velocity profile, which raises the question: can he replicate that production in 2026?

WAR Projection and Value Assessment: Despite the concerns in the Statcast data, Paredes still projects as a 2.5 to 3.0 fWAR player for 2026 based on his plate discipline, positional versatility, and proven ability to produce in a lineup slot that generates counting stats. He walks at an above-average rate, he does not chase excessively, and he can play both third base and second base. At $9.35 million, that puts his cost per projected win in the $3.1 to $3.7 million range, which is well below the free-agent market rate of roughly $8 to $9 million per win.

The $13.35 million club option for 2027 includes a fascinating wrinkle: it converts to a mutual option if Paredes finishes top-10 in MVP voting. This is Houston's hedge. If Paredes breaks out into a true star-level performer, they keep him at a discount. If he regresses toward his Statcast-implied true talent level, they maintain the flexibility to decline.

Regression Risk Versus Breakout Potential: The bear case for Paredes in 2026 is straightforward. His 2025 wOBA outran his xwOBA by a significant margin. Players with that kind of overperformance tend to come back to earth the following season. If you adjust his batting line to match his Statcast-expected metrics, you are looking at something closer to .240/.340/.420 with 15 to 17 homers. Still a useful player, but closer to a 2.0 fWAR contributor than a 3.0.

The bull case is equally compelling. Paredes was limited to 102 games in 2025 due to a hamstring injury. Before going down, he was slashing .259/.359/.470 with 19 homers in 94 games, a pace that projected to roughly 27 homers over a full season. A healthy 162-game sample could push his WAR above 3.5, especially if he maintains his All-Star-caliber first-half production. His plate discipline, including a walk rate that consistently sits in the 10 to 12 percent range, gives him a high floor even if the power dips slightly.

Context Within the Astros Lineup: Houston finished 87-75 in 2025, a competitive but not dominant record. Paredes slots into a lineup that includes some of the game's best hitters, meaning he will see better pitches hitting in front of and behind other dangerous bats. Lineup protection is one of those factors that traditional metrics struggle to capture but that can meaningfully impact a hitter's outcomes.

The Astros' decision to structure this deal with a club option rather than pursuing a multi-year extension tells you something about how their analytics department views Paredes. They believe in his floor but want more data before committing long-term. Paredes becomes a free agent after the 2027 season regardless, so this is effectively a two-year audition window where the data will either confirm him as a cornerstone piece or reveal him as a productive-but-replaceable contributor.

The Bottom Line on the Numbers: At $9.35 million, Paredes is being paid roughly at market rate for an arbitration-eligible player with his track record. The analytics suggest mild regression risk from his 2025 output, but even the pessimistic projection gives Houston a player worth his salary. The optimistic projection, one where he stays healthy for 150+ games and his first-half 2025 performance proves to be his actual talent level, makes this deal look like a bargain. For a front office that runs its decisions through expected value calculations, the math checks out either way. This is a fair deal that both sides can live with, and the club option gives Houston the asymmetric upside that smart organizations always want to capture.