Forty years on, the sentence is automatic. Bill Buckner lost the 1986 World Series.

He lost a game. Boston still had another one, and they led it 3-0.

Dwight Evans and Rich Gedman homered in the second inning of Game 7. Boston held that lead through five. Then Keith Hernandez singled two runs home, Gary Carter tied it, Ray Knight opened the seventh with a home run, and the Red Sox lost 8-5.

A man has carried a championship for four decades over one play, in a game his team led three separate times, in a Series his team still had a full night left to win.

The counting stopped at Game 6.

They got two outs and never got another one

Here is what happened in the bottom of the tenth, stated the way nobody has ever stated it.

Boston retired Wally Backman. Boston retired Keith Hernandez. Then Gary Carter came to the plate and the Red Sox never recorded another out in that game.

Four batters. Nineteen pitches. Three runs. No third out.

Nobody has ever counted that. Not the runs, everybody counts the runs. The length. How long a team spent hunting for one out with a championship sitting on the other side of it.

Baseball has no statistic for that, so I built one, ran it against every inning played since 1912, and pointed it at the bottom of the tenth.

What came back says two things that should not both be true. The inning was ordinary, the kind of thing that happens somewhere in baseball about once a week. And one thing about it has not happened again in a hundred and twenty-two years.

How long was it

I have been watching that inning since I was nine years old, and after forty years I realized I did not know one basic thing about it. How long was it?

Not how improbable. Everybody has that number. One percent, quoted since before I could drive. How long. How many batters did Boston chase before they gave up the last out of a championship? Was that a lot? Compared to what?

There is no statistic that answers that.

Chris Jaffe came closest. Writing at The Hardball Times in 2013, he refused to make the inning about Buckner. It wasn’t just the Buckner play. It was the entire sequence. Then his very next sentence described that sequence as a one-percent chance turned into a walk-off win. He saw that the sequence was the story, and reached for a percentage, because a percentage was all anybody had.

What baseball forgot to count

Here is the thing nobody tells you about baseball statistics. The entire field runs on a shortcut so elegant that it has been in place for fifty years, and so complete that it has quietly closed off an entire category of question.

Every moment of every game gets filed into one of twenty-four boxes. How many are out, and which bases have runners on them. Nobody on with one out. Man on second with two outs. Twenty-four combinations, and every pitch ever thrown in the history of the sport lands in one of them.

It works because situations repeat. A man on first with two outs in April behaves the same as a man on first with two outs in October, so you can pool a century of them and learn what tends to happen next. Run expectancy, win probability, leverage, every number you have ever seen on a broadcast sits on that foundation.

Now look at what the shortcut throws away.

Two outs with a runner on first is the same box whether he singled, walked, or reached because a shortstop threw the ball into the dugout. The route is erased. It has to be. The whole power of the system comes from refusing to care how you got somewhere.

Which means baseball can tell you precisely how much the tenth inning moved Boston’s odds. It cannot tell you what the inning felt like. It has no memory of length, of grinding, of a team throwing pitch after pitch at a man who will not go away. Nobody built that, because nobody needed it.

I needed it.

So I built it

Three point six million half-innings, every one played between 1912 and 2025. For each of the twenty-four boxes, count how many more batters come to the plate before the inning ends.

The answer for a fresh inning, nobody on and nobody out, is 4.274 batters.

That number is the reason to believe the rest. Baseball already knows a half-inning runs about 4.3 batters. It has known for decades. My table, built from scratch, landed on it to three decimal places.

Now the box that matters. Two outs, bases empty. The situation Gary Carter walked into with the season over.

1.47 batters. Two thirds of the time the defense escapes on the very next man.

Boston faced four.

The inning, laid against the table

Each row below is a Mets batter. When he stepped to the plate, Boston was sitting in one of the twenty-four boxes. The third column is how many more batters an ordinary defense faces from that box before the inning is over, counting the man at the plate. The fourth is how many Boston actually faced.

Batter	Boston’s situation when he came up	Batters an ordinary defense faces from here	Batters Boston faced	Difference
Backman	nobody out, nobody on	4.27	6	+1.73
Hernandez	one out, nobody on	2.88	5	+2.13
Carter	two out, nobody on	1.47	4	+2.53
Mitchell	two out, man on first	1.41	3	+1.59
Knight	two out, first and second	1.44	2	+0.56
Wilson	two out, first and third	1.44	1	−0.44

Read that last column from top to bottom. It climbs to Carter, the point at which Boston was furthest past what the situation was supposed to ask of them, and then drains away as the runs come across.

The Wilson row is negative, and that is not a rounding error. He came up with the defense owing about a batter and a half, and there were none left. Nobody followed him. The inning did not end. The game ended underneath it.

The answer was disappointing

So how unusual is four batters from two outs and nobody on?

It happens 3.38 percent of the time. One inning in thirty.

That is a Tuesday in Cleveland. It is the fourth inning of a game nobody will ever bring up again. Somewhere in baseball it happens about once a week, fifty times a season, and not one of those innings has ever been called famous.

Forty years of documentaries, and the measurement says the inning itself was nothing special.

So I went looking for what was.

The only one

Three outs from a championship. Nine men standing on the grass who have been playing since February to get to this one turn on the field, and all that is left is to collect three outs before the other team scores three runs.

That has happened one hundred and twenty-two times. Every World Series since 1903, every team that walked out there in the ninth inning or later with a lead and a championship on the far side of it.

A hundred and twenty of those teams got as far as two outs. The other two never did, losing on a walk-off with a single out recorded, Kansas City in 1985 and Arizona in 2001.

So I took the hundred and twenty and counted what happened after that second out.

Almost always, nothing at all. The typical team throws four more pitches. Four. It faces one more batter, gets him, and the dugout empties onto the field.

Boston threw nineteen.

Of the ninety-five clinching innings where somebody wrote the pitch count down, exactly one ran longer, and it belongs to the 1931 Cardinals, who needed twenty-three to finish off Philadelphia. Boston also faced four batters after that second out, which ties the most anybody has ever faced. The other two are those same 1931 Cardinals and a game from 1906 that nobody kept a pitch count for.

So it was a long inning. Second-longest of its kind ever played. That is a real fact and it is still not the thing I went looking for.

This is. It has been sitting in the record since the Coolidge administration, waiting for somebody to go and get it.

Of the hundred and twenty times a team reached two outs with a championship three outs away, the third out failed to arrive exactly once.

Every other team eventually got it. Several lost the game anyway, on a run that crossed the plate before the out came. Texas blew a World Series lead twice in the same night in 2011, in the ninth inning and then again in the tenth, and still recorded the third out both times.

A hundred and twenty-two years of this. One team went out for the last out of a World Series and never found it.

It took two ordinary things

Two things had to be true at the same time, and on their own, neither one is anything special.

Here is the first. A pitcher gets the second out and then cannot get the third. Base hit. Another base hit. The count runs full, the manager walks out, the inning will not close. Every fan has watched that happen and forgotten it by the next morning.

It occurs about two hundred and sixty times a season, which is more than once a day somewhere in baseball. Almost all of them are in June, in a game nobody will mention again, with nothing riding on any of it.

Here is the second. An inning ends without anybody recording that third out, because the winning run crosses the plate first and the game simply stops. Twenty-seven times in the history of the World Series a team has reached two outs and never gotten another one.

Now ask for both at once. A long grinding struggle after the second out, and then no third out at all.

Across every turn a team has taken on defense since 1912, three and a half million of them, that combination has come up sixty-four times. Once every other season. You could watch every pitch of every game, all thirty clubs, for two full years, and expect to see it happen one time.

And in the World Series, where the stakes make it unforgettable, the list shrinks again. Of those twenty-seven innings that reached two outs and produced no third, the 1986 tenth is by a wide margin the longest. It is also the only one of them with a championship sitting on the next pitch.

One at bat in four hundred and sixty-one

Mookie Wilson fouled off five pitches after the count reached 2-1. Six in total, counting the first one he saw.

I checked every postseason half-inning where a team took the field in the ninth or later, leading, one win from clinching a series. There are 396 of them. Inside those, 512 plate appearances began with two outs, and 461 of them carry a pitch record.

Here is the whole distribution of foul balls in those 461 at bats.

Fouls	Times
0	256
1	130
2	55
3	12
4	4
5	3
6	1

That last row is Mookie Wilson. One at bat out of 461, in the recorded history of teams standing three outs from a series win.

Five or more has happened four times. Billy Herman did it against Red Ruffing in 1938 and grounded out, and the Yankees finished a sweep. Reggie Jackson did it against Steve Howe in 1981 and reached on an error, and Los Angeles clinched two batters later. Ramon Martinez did it against Billy Wagner in 2006 and flied out, and the Mets finished off the Dodgers.

Four men, four at bats, across eighty-eight years. Wilson is the only one whose team went on to win.

What’s actually true

The ordinary part and the extraordinary part are the same part.

An inning that runs four batters past two outs happens once a week somewhere. Six foul balls in an at bat is nothing. Three singles, a wild pitch and an error is a box score you could pull from a game in April. Every ingredient of the most famous inning in the sport is common, and the measure I built to test it says so plainly. One in thirty.

What has never happened, in 122 chances across 122 years, is a team going out for the last out of a World Series and never getting it.

A probability table cannot produce that. It tells you the odds from a spot and forgets how anyone arrived there. You get the rest by going back through a century of play-by-play, one inning at a time, and adding things up until the number appears.

Nobody had. The data sat on a volunteer archive’s servers for decades, free to anyone, and the answer was four.

Four batters. Nineteen pitches. Three runs.

And one out that never came.

Appendix. Every other number on that inning

Everything else anybody has counted about the bottom of the tenth, and what each number is actually measuring. Renders collapsed.

One percent. You see it everywhere, and it comes from win probability, which is simpler than it sounds. A computer finds every game where a team was in roughly this spot and reports how often that team won. Boston sat near 99 before Carter came up. History, looked up. Not a count of anything.

Forty-one against forty. Writing for SABR in 2013, John R. Crook measured how far each play moved Boston’s odds. Bob Stanley’s wild pitch took 41 points off. Buckner’s error took 40. The pitch scored the tying run and the error scored the winner, but by how close Boston came to surviving, the wild pitch was the heavier blow. The error is the one that got a man’s name attached to it for life.

Seven. Pitches thrown with two outs and two strikes, each one ending the Series if it is a strike. One to Ray Knight, who singled on it. Six to Wilson, including the 2-2 pitch that got past Rich Gedman and tied the game. Boston was not near the title on that pitch. They were one strike from it.

Eleventh of ninety-seven. By two-strike pitches thrown, the tenth ranks well up the clinch list but not at the top. Toronto threw thirteen trying to close out Atlanta in 1992. The Dodgers threw twelve in 1981. The median is four.

Nine of 122. Games in which the team facing elimination came back and won anyway. 1910, 1911, 1912, 1985, 1986, 1997, 2001, 2011 and 2025. Texas did it twice inside the 2011 game, once in the ninth and once in the tenth, so the occurrences number ten while the games number nine. Only one of the nine began with two outs and nobody on base, and it scored three runs, more than any of the other eight.

One of thirty-nine. Take the shape of the rally by itself, two outs and nobody on, four men reaching in a row, three or more runs scoring. Thirty-nine occurrences in World Series history. Exactly one started from a two-run deficit, and it is the only one in extra innings.

Eleven thousand and ninety-eight. That same shape across all of baseball since 1903. Fifty-three times per thousand games, a rate that has barely moved since the 1920s.

The closest cousin. Game 4 of the 1941 World Series, when Mickey Owen dropped a third strike with two outs in the ninth and Brooklyn gave up four runs across seven batters and twenty-eight pitches. Longer than 1986 by every measure. Brooklyn was ahead in that game and trying to even the Series. They were not three outs from a title.