Mets vs. Giants Preview — September 6, 2026
By the end of a season like this one, most baseball is content. The Mets are 64-78, the Giants 59-84, and this Sunday matinee at Citi Field carries the unmistakable scent of two clubs playing out the string. The standings say one of these teams is merely bad and the other is worse — and the betting market has behaved accordingly.
The Mets opened -149 on the moneyline and have been steamed to -172. The total has been dragged down a full run, from 9 to 8. Sharp money and public money appear to be in rare agreement: New York wins a low-scoring, home-finale-type game against a team traveling cross-country with nothing to play for.
Earl's model disagrees with the price, the total, and parts of the narrative. And in that disagreement is where the value hides.
Our model projects a home margin of just 0.46 runs with the Mets scoring 4.5 and San Francisco 4.1 — a projection that makes the Mets the favorite to win outright but lands them well short of the -1.5 run line. The model projects a total of 8.59, 0.59 runs above the market's 8, producing a soft but real lean to the Over.
The model's highest-conviction play of the day is on the run line: San Francisco +1.5, carrying an expected value of +11.3%. That edge is worth examining closely, because everything about the surface statistics — the Mets at home, the Giants' ghastly road record (26-48), the injury report — screams the opposite direction.
The Pitching Mismatch That Isn't
The naive read of this game is a clean starting-pitcher advantage for New York. Christian Scott takes the ball for the Mets — a right-hander with a 3.80 ERA and 1.31 WHIP over 20 starts. The Giants counter with Cesar Perdomo, a left-hander with 4.0 career innings pitched at the major league level and a sparkling 2.25 ERA that means absolutely nothing at that sample size.
| Pitcher | ERA | WHIP | IP | K | BB | K/9 | BB/9 |
|---|---|---|---|---|---|---|---|
| Christian Scott (R) | 3.80 | 1.31 | 94.2 | 113 | 44 | 10.74 | 4.18 |
| Cesar Perdomo (L) | 2.25 | 1.00 | 4.0 | 4 | 1 | 9.00 | 2.25 |
On paper, Scott should dominate. But strip the ERA off and look at the underlying profile: Scott is striking out 10.74 batters per nine innings while walking 4.18 per nine — 44 free passes in 94.2 frames. This is a pitcher who misses bats at an elite rate but also creates his own traffic. Against a San Francisco lineup that, in Earl's model, has been more productive of late than its season line suggests, that walk rate is a wobble point.
Perdomo, meanwhile, is effectively a mystery box. No recent starts are on record, and his role is undefined — either a fresh call-up or a hybrid/bulk arm. What we know is confined to four innings, during which he allowed three hits and one home run. The Giants have not deployed him as a traditional starter, which strongly suggests this is an opener or bullpen-day scenario. That is not inherently a red flag: a scheduled parade of unfamiliar arms can neutralize a scouting report that was built for one pitcher. The Mets' hitters prepared for Perdomo; they did not prepare for the three or four arms behind him.
The deeper point: this game is far more likely to be decided by bullpens than by starting pitchers, and both bullpens are compromised.
Two Roster Apocalypses, One Wildly Worse
Here's where the public narrative gets it wrong. The casual bettor looks at the Giants' record and assumes a roster of healthy underachievers. In reality, San Francisco is running out a lineup assembled from the wreckage of a 60-day injured list that reads like a starting nine.
Matt Chapman (3B) is out. Willy Adames (SS) is on the 10-day list. Marcelo Mayer (2B) is out. Harrison Bader (CF) is out. Casey Schmitt and Victor Bericoto are both out, and both of San Francisco's listed catchers — Jesus Rodriguez and Daniel Susac — are on the 60-day list. That is not a depleted lineup. That is a pitching-change-induced batting practice squad in disguise.
The crippling is not limited to position players, either. The Giants' pitching staff has been shredded: Adrian Houser, Keaton Winn, Hayden Birdsong, Joel Peguero, José Buttó, Randy Rodríguez, Rowan Wick, Matt Gage and Carson Whisenhunt are all on the sidelines. Yet the model still projects San Francisco to score 4.1 runs against Scott and the Mets' bullpen.
That projection is less about the Giants' offense and more about what the Mets are sending out behind Scott. New York's own relief corps is missing Devin Williams (15-day), Reed Garrett (60-day) and Cionel Pérez (15-day), with Tylor Megill also shelved. When both bullpens are running on replacement parts, late-inning leverage evaporates — and that is precisely the environment in which undersized totals get blown open.
The market dropped this total from 9 to 8, betting on a crisp, quick, pitcher-friendly matinee. Earl's model sees a combined 8.59, fueled by two second-unit bullpens and a starting pitcher who walks four batters per nine innings.
What the Steam Is Really Telling You
Line movement is often treated as a wisdom signal. When the Mets go from -149 to -172, the conventional interpretation is that sharp money knows something about New York. But look closer at the shape of the movement. The moneyline steamed while the total fell a full run — those two moves are in tension. A bettor who is confident the Mets win by multiple runs rarely also bets the Under; if you believe in Christian Scott, you want the Over to arrive via Giants helplessness. The simultaneous steam toward New York and the Under suggests the market is pricing efficiency, not conviction — a home team against a decimated visitor, at a pitcher-friendly park, in a game that means nothing to either club.
That is a narrative price, and narrative prices are what Earl's model is designed to exploit.
| Market | Model Lean | Line | Edge |
|---|---|---|---|
| Run Line | SF +1.5 | -1.5 | +11.3% EV |
| Moneyline | NYM | -172 | -10.3% EV |
| Total | Over 8 | 8.0 | +0.82% EV |
The moneyline number is the giveaway. Even though the model likes New York to win, at -172 the expected value is negative 10.3%. The model projects a coin-flip-ish game separated by less than half a run, and the market is asking you to pay a 63% premium on that coin flip. That is not an edge; that is a donation.
The Play
The structure of this game — a right-hander who issues free passes, an unproven lefty opener, two gutted bullpens, and a Giants lineup with nothing left to protect — points to a close, sloppy game. Earl's model sees San Francisco staying within the run line more often than the market implies, and the Mets winning by less than the spread even in the most common outcome.
That is why the recommended approach here is layered:
- The primary play is San Francisco +1.5. The model's predicted margin of 0.46 runs against a price on the Mets laying 1.5 gives the Giants cover the majority of the time, win or lose. At even money, that is a positive-expectation ticket.
- A modest lean to the Over 8 carries a thin edge (+0.82%), but it pairs logically with the run-line play. If the Giants keep it close, they are likely scoring; if the Mets blow it open, they are likely doing so in a game that sails past 8 anyway.
- Stay away from the Mets moneyline at -172. Earl's model has New York winning outright, but the price has already consumed the value. You do not need to bet a favorite to believe in a favorite.
Why the Public Is Wrong
The public sees 59-84 against 64-78, glazes over the injury report, and assumes the Mets roll a bad team at home in a September getaway game. What the public misses is that the Mets are a 33-40 home team — barely above replacement level at Citi Field — and that New York's offense has not been producing at its season rate recently. The model explicitly flags the Mets' recent run-manufacturing as a concern, noting the Giants have been the more productive lineup of late despite their record.
Both clubs have run differentials that scream mediocrity: the Mets at -50, the Giants at -73. Strip the names off the jerseys, and this is nearly a coin flip between two sub-.500 clubs — one of which happens to be playing at home. On Sunday, the market is pricing the Mets like a contender. Earl's model prices them like a team that wins by half a run.
In September, with nothing on the line, the difference between a 64-win team and a 59-win team is almost always the schedule, not the roster. The Giants' roster is healthier than their record implies — or at least, it was, before injuries ate the lineup. Even in its current state, the roster is good enough to hang around against a pitcher who walks a batter every other inning.
The model's number is 0.46 runs. Take the side that doesn't need to be correct by much, at a price that pays you even money to be right half the time. That is the only real edge on the board in the final game of this series.