We read all 15 of Sunday's MLB games: in 12, the winner's biggest inning matched or beat the loser's whole night

A box score tells you who won. It rarely tells you the game turned on one half-inning. We read the complete linescore of every one of the 15 games on Major League Baseball's Sunday slate — Aug. 23, 2026, all 30 clubs, all final — inning by inning, all 141 runs. One pattern holds across almost the entire card: in 12 of the 15 games, the winning team's single biggest inning matched or beat what the losing team scored in the whole game — five of the 12 exactly equal, seven larger. The Philadelphia Phillies made it literal. They were shut out for eight innings and trailed the St. Louis Cardinals 4-0, then scored all six of their runs in the bottom of the ninth — capped by Kyle Schwarber's two-out, two-run home run — to win 6-4. Their biggest inning, six, was larger than everything St. Louis managed all day, four. The blowout said it louder. The Chicago Cubs beat the Seattle Mariners 19-2 behind a nine-run eighth inning and two grand slams; that one frame outscored Seattle's whole night by seven. In Houston, the Athletics spotted the Astros a 6-0 lead, answered with a six-run fifth, and won 7-6 — their biggest inning exactly matched Houston's entire output, and a later solo homer did the rest. Add it up: 80 of the 141 runs scored across the slate — 56.7% — came in each team's single biggest inning. More than half of a full day of major-league offense arrived in 30 concentrated bursts, one per team, the rest of the innings nearly silent.

Strip the numbers and this is a truism — of course crooked numbers win baseball games. The finding isn't that big innings matter; it's their size, visible only by reading the whole card at once. On this Sunday the single best inning wasn't a contributor to the result, it was the result: in 12 of 15 games it alone matched or beat everything the other side did across nine. Five of those 12 winning innings exactly equaled the loser's full total; seven exceeded it. The losers often did the textbook thing and lost anyway. Houston scored in three separate innings and still fell, because none of the three was as big as the Athletics' one. Three teams won the other way, and they are the tell. Kansas City beat Detroit 11-7 needing two crooked innings — five runs, then four — because neither alone cleared Detroit's seven. Boston edged San Francisco 5-4 with a top inning of three against the Giants' four. And Atlanta beat Milwaukee 4-2 without ever scoring more than one run in an inning: four separate single runs, the anti-big-inning win, and the only game on the board where accumulation, not a burst, carried the day.

The honest caveat is sample. This is one day — 15 games, a single Sunday — not a season, and 'the winner's biggest inning matches or beats the loser's total' will not hold near 80% across 2,400 games; regression pulls it toward the sport's long-run rate. A skeptic can also call the framing close to tautological: the team that erupts for a big inning tends to be the team that wins, so the two naturally travel together. Both are fair. What the complete-slate read adds is the magnitude on a given night — and the three visible exceptions, including Atlanta's four single runs, that prove the burst isn't the only path, just the one this slate overwhelmingly rewarded.

From a single ballpark, Sunday was fifteen separate stories — a walk-off in Philadelphia, a laugher in Chicago, a comeback from six down in Houston. Read as one card, they fold into a single sentence: on this day the biggest inning was not part of the game, it was the game. In twelve of fifteen, one team's best frame outweighed everything the other side did across nine innings — and of the three games where it did not, Atlanta stands apart as the only winner without a burst at all: four separate single runs, the one place patience, not a crooked number, wrote the result.