The NLDS is tied 1-1, and Game 3 moves to Truist Park tonight, October 6, 2026, with first pitch at 6:00 p.m. ET on FS1. Yoshinobu Yamamoto starts for the Los Angeles Dodgers and Chris Sale starts for the Atlanta Braves. The series so far has run on thin margins. Los Angeles took Game 1, 5-3. Atlanta answered 3-2 in Game 2, a game decided by a fifth-inning wild pitch, after which the Braves' bullpen retired 14 consecutive Dodgers hitters.
Our pick is Under 6.5.
The number
Our model engine projects a game total of 5.0 runs. The posted total is 6.5, priced over -102 and under -120. The engine's reference market figure is listed at 8.0, which does not match the current 6.5 line. We grade against the 6.5 actually on the board, where the model's projection sits 1.5 runs below the number.
We need to be clear about what is not available. This pick carries no published model probability, no market-implied probability comparison, and no formal edge figure. Those fields are n/a for this game. The engine's game-level confidence reading is 55.9, and that should not be read as a win probability for the Under. For reference, a -120 price requires the Under to land about 54.5% of the time to break even. Without a published probability, we cannot state how far above that threshold the pick sits. Please size your expectations accordingly.
Why the model leans this way
The projection itself. The engine projects Atlanta to win 3-1, with a margin of 0.32. That score totals four runs, and the model's full total of 5.0 allows for ordinary scoring noise above it. Both figures sit below 6.5.
A neutral park. Truist Park carries a park factor of 1.0. The venue neither adds nor removes runs from the projection, so the Under lean comes from the matchup rather than the setting.
Close, unremarkable offense ratings. Atlanta rates 27.7 and Los Angeles 26.0. The 1.7-point gap is small, and neither lineup grades as an outlier that would push the projection upward.
Pitching. Sale has allowed a 0.78 WHIP this postseason, with 12 strikeouts against one walk in 7.2 innings. That profile limits traffic on the bases, which is the main way an Under survives. His 3.52 ERA shows he has been hittable for runs, but the strikeout-to-walk ratio is the steadier signal. On the other side, we could not confirm details of Yamamoto's most recent outing, so we make no specific claim about his current form.
Bullpen context. Atlanta's relievers just retired 14 Dodgers in a row to close out Game 2. That is a single game, but it is recent evidence that the late innings can stay quiet.
Series scoring. The two games have produced 8 runs and 5 runs. The 3-2 game in the series' only tight-pitching matchup so far is the template this pick needs.
Streaks. Los Angeles has won 3 straight. Atlanta has lost 1 straight. The model's projection already accounts for this form and still lands at 5.0.
The case against
There are concrete ways this pick loses.
Game 1 already cleared the number. The 5-3 opener totaled 8 runs. A repeat of that game beats the Under by 1.5 runs. These lineups have shown they can put up a seven-run combined total in this exact series.
Dodgers form and power. Los Angeles went 100-62, closed 8-2 over its last 10 and 15-5 over its last 20, and arrives on a 3-game win streak. Max Muncy has two home runs this postseason and Freddie Freeman has three hits. Two swings can account for half the margin here.
Atlanta's bats at home. The Braves went 52-29 at Truist Park and swept the Dodgers there in August. Michael Harris II leads the club with 8 hits, Ozzie Albies has 2 home runs and 5 RBIs, and Dominic Smith is hitting .400. A lineup that comfortable at home can produce a crooked inning against anyone.
Sale's run prevention is imperfect. A 3.52 ERA over 7.2 innings means roughly three earned runs. The low WHIP has not fully translated into run suppression, and the sample is small.
Thinned Dodgers depth. Gavin Stone, Ben Casparius, Brusdar Graterol, Will Klein and Jake Cousins are all on the 60-day IL. If Yamamoto exits early, Los Angeles reaches deeper into its pitching staff, and middle innings become a real risk to this pick.
The price and plain variance. At -120, every loss costs more than a win returns. Totals swing on a single walk, error or wild pitch, and Game 2 was decided by one. We publish no probability for this pick, so treat it as a lean rather than a measured edge. Over a large sample, a pick like this loses a substantial share of the time, plausibly four games in ten or more. Every result, win or loss, goes on our public ledger.
21+ only. Treat every prediction as entertainment, set limits in advance, and seek help if it stops being fun.