A Level Series, an Ace on the Mound
The National League Division Series swings back to Truist Park knotted at one game apiece, and the momentum belongs to Atlanta — for whatever that's worth against a Los Angeles pitching staff that has looked almost unhittable.
Game 2 was the bullpen game the Braves won, and it validated manager Walt Weiss's gamble: hold Chris Sale back, let a patchwork relief corps handle the Dodgers, and trust that Los Angeles' bats could be kept quiet. They were. Now the ace takes the ball, and the series' most compelling subplot arrives with him — a rematch of an August 27 pitchers' duel that ended 1-0 in Atlanta.
The Dodgers counter with Yoshinobu Yamamoto, who was passed over for Game 1 (Tarik Skubal got that nod) and then bypassed again in Game 2 when Blake Snell took the ball. That sequencing drew outside skepticism. Yamamoto's answer comes Tuesday night, and it arrives carrying the weight of his historic World Series performance from last fall.
What Earl's Model Says
Our model's formal outputs for Game 3:
| Market | Model Pick | Model Expected Value |
|---|---|---|
| Moneyline | Los Angeles Dodgers | +9.25 |
| Run Line | Atlanta +1.5 | -9.39 |
| Over/Under | Over 6.5 | +1.99 |
Predicted score: Dodgers 3.8, Braves 3.5 — a total of 7.25 and a margin of -0.28 runs.
Read that margin again. The model's projected winner is Los Angeles, but by less than a third of a run. That's the entire tension of this game compressed into one number, and it dictates how you should treat every market on the board.
Sale's Ceiling — And the Regression Flag
Chris Sale's regular season was vintage: 14-9, 2.16 ERA, 0.98 WHIP, 200 strikeouts in 162.2 innings. His 11.07 K/9 and 1.72 BB/9 produced a 6.45 strikeout-to-walk ratio, and his percentile ranks are a wall of red — 94th percentile in strikeout rate, 97th in walk rate, 96th in chase rate, 85th in whiff rate.
The slider is the finishing weapon. Opponents hit .186 with a .247 slugging percentage against it across 268 plate appearances, with a 38.2% whiff rate and a 42.2% strikeout rate. The four-seamer (41.4% usage) is the setup pitch, and the two make up more than 81% of his arsenal.
The flag: Sale's xERA is 2.97, a full 0.81 runs above his actual ERA. He has also posted a 75th percentile hard-hit rate against a 37th percentile barrel rate — a profile that says he's generating lots of medium-contact and getting away with it. Against a Dodgers lineup at full strength, that gap is the single biggest risk in this game.
Yamamoto Is the Better Bet by the Underlying Numbers
Yamamoto's surface line — 14-9, 2.53 ERA, 0.87 WHIP — is a shade behind Sale. His underlying profile is a shade ahead.
He allowed just 5.79 hits per nine innings across 185 frames. His xOBP ranks in the 95th percentile and his xwOBA in the 90th, and his xERA of 2.94 sits only 0.41 runs above his actual ERA — less regression baked in than Sale's.
The split-finger is a genuine problem pitch: .158 batting average, .271 slugging, 33.8% whiff rate over 210 plate appearances. He throws it 26.6% of the time, pairs it with a four-seamer at 26.2%, and mixes a cutter, curveball, sinker and slider behind it. Six pitches, none of them a giveaway.
His recent form carries one blemish — a 1-5 loss to San Diego on September 23 — but the two starts before that were a 4-0 shutout of Cincinnati and a 14-1 rout of the same club.
The Over Is Where the Real Edge Lives
Here's where we separate from the conventional read. Both starters own ERAs that outperform their expected marks, which is precisely why the market has squeezed this total down to 6.5 — the kind of number you only see when two aces are advertised.
But the venue argues the other way. Truist Park carries a 103 overall park factor and a 109 home run factor, with a 1,050-foot altitude — unusually high for the Southeast — and short porches at 325 feet to right and 335 to left. This is not a pitcher's park.
Meanwhile, both bullpens are compromised. Atlanta's injured list is a rotation graveyard: Spencer Schwellenbach, Spencer Strider, Hurston Waldrep, Joe Jiménez and Joey Wentz are all on the 60-day, with Bryce Elder, Martín Pérez and Reynaldo López on the 15-day. Los Angeles is down Will Klein, Ben Casparius, Brusdar Graterol, Gavin Stone, Jake Cousins and Eric Lauer on the 60-day, plus Blake Treinen on the 15-day.
Neither manager has a full deck behind his starter. The model's +1.99 expected value on the Over isn't a bet against Sale or Yamamoto — it's a bet against the depth behind them and the park they're throwing in. Trust the number: the model's quantified output of 7.25 total runs clears the 6.5 line, even though its own narrative summary leaned under. We side with the math.
Why the Moneyline Is Playable but Razor-Thin
At -108 on both sides, the market is telling you this is a coin flip with a small tax. That's consistent — there has been zero line movement from open to current across the spread, total, and both moneylines. No sharp money has tipped the number either way.
The model's +9.25 EV on the Dodgers is the largest edge on the card, but recognize what it's built on: a 0.28-run projected margin and Los Angeles' +202 run differential versus Atlanta's +120, against a Braves club that went 54-30 at home. This is a lean, not a lock. If you play it, play it small and understand you're backing the better underlying team in a game the model itself calls nearly even.
The Run-Line Trap
Our model lists Atlanta +1.5 as its run-line pick — and simultaneously grades that play at -9.39 expected value. That's not a contradiction, it's a price problem. When the projected margin is 0.28 runs, the plus-1.5 has real win probability attached, but the juice being charged for it erases the edge.
Pass on the run line. Not because the projection is wrong, but because you'd be paying for protection you don't need at a cost the model says is negative.
The Card
- Dodgers moneyline (-108) — model's strongest EV at +9.25. Small stake, coin-flip awareness.
- Over 6.5 — +1.99 EV, the cleanest structural play: suppressed total, hitter-friendly park, two depleted bullpens.
- Braves +1.5 — model pick, negative EV. Lean only, no ticket.
Bottom Line
The public sees two aces and a low number and reflexively takes the under and the home dog with the runs. The model sees a 7.25-run game in a 109 home-run-factor park with two bullpens held together by tape, and it sees a Dodgers team with a 202-run differential winning by a nose.
That's the edge: Over 6.5, with a small Dodgers moneyline ticket attached. Atlanta's path to a 2-1 series lead runs through Sale out-dueling Yamamoto the way he did on August 27 — and the model says that particular lightning doesn't strike twice in the same month.