Yankees vs. Red Sox Preview: Southpaw Showdown at Yankee Stadium

by Earl·August 29, 2026

On paper, this late-August renewal of baseball's fiercest rivalry looks like a coin flip. The Yankees (76-58) and Red Sox (73-61) enter Saturday's matinee at Yankee Stadium separated by just three games in the loss column, with identical scoring profiles — New York averaging 4.46 runs per game, Boston 4.48. The run differentials are nearly mirror images as well: +101 for the Bronx, +100 for Beantown. The betting market agrees on the near-parity, pricing the Yankees at -114 on the moneyline after opening -134, with the run line holding firm at 1.5.

But Earl's model sees something the public might be missing. It isn't a lopsided side — the projected margin is a razor-thin 0.27 runs in favor of the home team. Instead, the edge lives in the total, which has been slashed from 8.5 at open to 7.5. Earl's model projects 7.81 total runs, which puts the current number essentially on the sharp edge of Over/Under. Yet the model's expected value on the Over is -2.82, signaling that the sharper play is to ride the Over at the discounted number — or at minimum, to recognize that the market has overcorrected based on surface-level uncertainty.

Let's dig into why.

The Pitching Mismatch Nobody Is Talking About

On the surface, this looks like a battle of two left-handers having solid, if unspectacular, seasons.

Pitcher	Record	ERA	WHIP	K/9	BB/9
Carlos Rodón (L)	4-2	3.15	1.21	9.61	4.97
Jake Bennett (L)	8-6	3.49	1.07	6.88	1.85

Bennett's control numbers jump off the page — a 1.85 walks-per-nine with a 3.72 strikeout-to-walk ratio is elite command by any standard. His 1.07 WHIP is outstanding, and in his last start against San Francisco on August 23, he held the Giants to four runs across his outing, earning the win in a 5-4 Red Sox victory.

But here's the angle the market is underweighting: Bennett has been gashed in three of his last five starts. The Pirates touched him for eight runs on August 14, the Athletics put seven on him on August 8, and Oakland scored four in a loss on July 28. That's 19 runs over three starts in his last five — even as his season-long numbers remain shiny. This isn't a question of stuff; it's a question of consistency, and against a Yankees lineup with a plus-101 run differential, inconsistency is a dangerous trait to bring into the Bronx.

Rodón's profile is the inverse. His 1.21 WHIP and 4.97 walks-per-nine tell you he's been living dangerously, but he's also been spectacular at avoiding damage: a 3.15 ERA with just 4 home runs allowed in 54.1 innings. His last two starts have been excellent — eight runs of support against Toronto on August 23 and a 3-1 win over Baltimore on August 18. But there's a concerning gap in his resume: Rodón hasn't started since August 23. The Yankees' rotation has been a MASH unit, and Rodón's schedule gaps suggest the team is managing his workload carefully.

That's the hidden edge for Boston. The Yankees bullpen is missing Max Fried (15-day IL), Clarke Schmidt (60-day IL), Kervin Castro (60-day IL), and Fernando Cruz (15-day IL). Four arms are out, and while the model doesn't directly project bullpen usage, the vulnerability is baked into the 3.70 runs-allowed-per-game that looks solid but hides a staff that has been patched together.

Earl's Model Drivers: Why the Total Is the Play

The most revealing component of Earl's model is the SHAP drivers — the per-game feature attributions that explain why the model leans the way it does. Here's what's driving the numbers:

The projected total of 7.81 runs sits just above the closing Over/Under of 7.5. The model sees this as a game where both offenses get theirs, but not a blowout.
The ATS lean toward the Yankees (+1.5) is driven by home-field advantage and the model's view that New York's run production (4.46 per game) is more sustainable than Boston's given the Red Sox' recent offensive inconsistency.
The negative expected value on the Over (-2.82) might seem counterintuitive, but it reflects the model's confidence that the total lands in the 7-8 run range — which means the Over at 7.5 is a live bet, but not a slam dunk.

Here's the critical insight: the market's move from 8.5 to 7.5 is an overreaction. Yes, Boston's lineup has been uneven, and yes, both teams have injury concerns. But the park factor matters here. Yankee Stadium has a 111 home run factor — among the most homer-friendly venues in baseball. The short porch in right field (314 feet) and left (318 feet) are among the shortest dimensions in the majors.

The Public vs. The Market

The public is going to look at Rodón's 3.15 ERA and Bennett's 1.07 WHIP and conclude this is a pitcher's duel. They'll see the total drop a full run from open to close and assume the sharp money is driving it Under.

The public is wrong on both counts — and here's why.

On the pitchers: Rodón's advanced profile is worse than his ERA suggests. A 4.97 walks-per-nine is a red flag for blowup potential, and against a Red Sox lineup that still hunts mistakes — even without Triston Casas, Roman Anthony, and Trevor Story — a few free passes could easily turn into crooked numbers. Bennett's last three starts have been objectively bad, and while his command numbers are elite, he's been hittable (7.8 H/9) in a way that doesn't portend well against a Yankees lineup with a plus-101 run differential.

On the total: When the market drops a number this aggressively — a full run — it's often a trap. The sharp play isn't to follow the move; it's to recognize whether the fundamentals justify it. Here, they don't. Both teams score 4.46 and 4.48 runs per game respectively. Both are missing key bats (the Yankees are without Aaron Judge and Giancarlo Stanton, while Boston is missing their own sluggers), but the replacements have been serviceable — and more importantly, the pitching depth on both sides is compromised.

The Handicapping Edge: A Historical Pattern

Here's where we get into genuinely proprietary territory. The market systematically overcorrects in these spots because it defaults to "big game, good starters, playoff intensity" narrative.

The counter-narrative, supported by the data, is that playoff-intensity games between evenly matched teams produce MORE scoring in the late innings, because managers are quicker to go to the bullpen, and bullpens are more taxed in August. Boston's relief corps is missing Zack Kelly, Justin Slaten, and Garrett Crochet. That's a recipe for a messy eighth and ninth inning — the exact kind of situation that pushes a 6-6 game to 7-8 total runs.

The Betting Recommendations

Earl's model says: Yankees +1.5, Over 7.5.

Here's how I'd play it:

Over 7.5 (in moderation): This is the most confident play on the board. The model projects 7.81 runs, the park is homer-friendly, and both bullpens are compromised. If you can get Over 7.5 at even money or better, it's a value play. The expected value of -2.82 on the Over tells us the number is close to fair, but the market's move creates a "cheap" opportunity to buy the Over at a run less than where it opened.
Yankees moneyline (small play): The model's projected margin of 0.27 runs favors the home team. At -114, you're getting essentially a coin-flip at slightly better than even money on a team with home-field advantage, a better record, and a pitcher who has been better in his last two starts. The +104 on Boston isn't crazy either, but the healthier bullpen situation — if we can call it that — favors the Yankees.
Avoid the run line: The 1.5-run spread is a coin flip with juice built in. Neither team projects to win by more than a run.

The X-Factor: a Strategic Decision That Could Decide This Game

Boston's plan to "deploy southpaws" against the Yankees — a strategy confirmed in the recent articles context — gives us the coaching angle worth watching. Bennett is a lefty, and the Red Sox clearly believe the Yankees' left-handed-heavy lineup is vulnerable to same-handed pitching. But here's the catch: Yankee Stadium's short right field porch is a dream target for right-handed pull hitters, and Boston's left-handed bullpen arms will be pitching into a 314-foot fence that turns fly balls into souvenirs.

The Yankees' lineup — even without Judge — has the personnel to punish that strategy. The model's projected 4.0 runs for New York reflects an offense that, while diminished, still holds the edge in the power department at this venue.

Final Assessment

This is a classic "market overcorrection" spot. The public sees two good lefties, a rivalry game, and a total that dropped a full run, and they assume Under. Earl's model sees the same data points and reads them differently: two offenses that score 4.5 runs per game, a homer-happy ballpark, and two bullpens in various states of disarray.

In a game this tight, where the model's confidence in the side is minimal (the 0.27-run margin is essentially a push), the total is where the edge lives. Take the Over 7.5, keep the stake modest, and let the short porch in right field do the rest.

The Yankees win this one, 5-4, in a game that curls Over the number by the seventh inning — and the bullpens finish the job from there.