Keith Hernandez made his famous play six times. Nobody counted the one he made 250 times.

Keith Hernandez is famous for one particular play. He made it six times in seventeen seasons.

The play that actually set him apart from every other first baseman who has ever lived, he made two hundred and fifty times. It has no nickname, no highlight reel, and nobody has ever written a word about it.

Neither of those is a failure of attention. The numbers at first base give out, and the reason sits in the first line of his fielding record. Eighty-four percent of the outs Keith Hernandez made there came off somebody else’s throw.

A first baseman plants a foot on the bag and catches what the shortstop and the second baseman and the third baseman throw him, and the scorer writes a putout next to his name. Hernandez has 17,936 of those across seventeen seasons. He fielded 2,823 of the balls himself. The rest arrived through the air.

So almost everything written down about a first baseman is really a record of the four men around him. Which is awkward, because the entire argument about Keith Hernandez is a defensive argument.

Baseball’s main career-value number, the one people reach for in Hall of Fame arguments, does not try to solve that. Baseball Reference’s version subtracts about ten runs from every first baseman before he has done a single thing, on the theory that his is the easiest job on the field. FanGraphs subtracts twelve and a half. Neither of them is measuring Hernandez. They are both measuring the position, and they disagree about it by a quarter.

The one system that counts what a first baseman does on bunts, which is where this ends up, does not start until 2003. Hernandez retired in 1990.

Where these numbers come from

A word about that first, because it decides what I can and cannot tell you.

Every play in a major league game going back to 1908 has been written down. Not a summary of the game, the actual play, one line at a time. Who was batting, whether he bunted, which fielder picked the ball up, who that fielder threw it to, and which runner was put out at which base. Volunteers at an outfit called Retrosheet have spent decades typing all of it in from old scorecards and newspaper accounts, and they give it away free.

What nobody ever wrote down is where the players were standing. There is no column for how far off the bag a first baseman was playing, or when he started running.

So when I say Hernandez fielded a bunt, I mean the line for that play names him as the man who picked the ball up. Where he came from to get there, nobody recorded.

Dwight Gooden cost him two putouts a game

The oldest measure of a fielder has a name, range factor, and it is as simple as it sounds. Count every out he had a hand in, divide by games played. Bill James put it in his first Baseball Abstract in 1977, though he was not the first to think of it. A scorekeeper for the Philadelphia Athletics was rating fielders that exact way in 1872.

Hernandez’s sits between 10.4 and 11.7 through nine seasons in St. Louis. Then it falls off a cliff. In 1984, his first full year with the Mets, it drops to 9.11 and never comes back.

Read that cold and you would guess he got old. He was thirty.

What changed was the pitching. The Cardinals staff he had stood behind for nine years struck out about four men a game. The Mets staff he joined struck out six and a half, on its way to Gooden and Darling and Ojeda. Every extra strikeout is a ball that never gets hit anywhere, and the Mets infield went from handling fourteen ground balls a night to twelve.

The proof is in which putouts vanished. He lost about two a game, and almost every one of them was a throw coming to him. The balls he went and got himself barely moved.

So range factor says 1984 was the worst fielding season of Keith Hernandez’s career. Count instead how often he was the first man to touch a ball hit into the infield, which is the part that is actually him rather than his pitchers, and 1984 turns out to be the best year he ever had.

What he did with the ball

An assist is a record of a decision. He had the ball in his hand and he picked where to send it.

He has 1,673 of them, more than any first baseman who ever played, and I took every one apart to see where the ball went.

Just over half, 885, are the dullest play in the sport. Ground ball to his right, pitcher running over to cover the empty bag, soft little toss. Everybody makes that play, and another 158 are the same thing with the second baseman taking the toss.

Five hundred and thirty-six went somewhere else. To second base, to third, or to the catcher.

Think about what that asks of him. A ground ball comes to him with a man on first. The out he is owed is six feet away, the batter running straight at him, and Hernandez is already standing on the bag. Nobody in the history of baseball has been second-guessed for taking it.

Throwing ahead of that runner means turning away from the certain out and going after the man already on base. The one with the head start. The one who actually scores.

When it works the rally is over. When it does not, both men are safe and fifty thousand people just watched him try.

Hernandez made that throw about fifty percent more often than the other first basemen playing his own seasons. Nobody in a hundred and eighteen years of box scores made it more.

The man in second place is Ferris Fain.

The play nobody told it to look for

Then I did something backwards on purpose.

If you already know what a player is famous for and you go measure that, you will find it. So instead of looking for bunts I made a list of every different kind of play a first baseman can be part of, ninety-two of them, and asked the identical question about each one. Did Keith Hernandez do this more than the first basemen of his day?

Nothing about bunts. Nothing suggested, nothing ranked. Ninety-two identical questions.

What came back has never had a nickname, and there is a good reason nobody remembers it. The inning does not end. Nobody scores. The batter ends up safe at first base, so from the stands it looks like the defense came out behind.

He fields a ground ball near the bag. The shortstop abandons his position and sprints to cover second. Hernandez turns and throws there, and the man coming from first is out before he gets down.

Two hundred and fifty times. A first baseman of those years would have done it about a hundred and forty. Out of ninety-two things he could have been good at, that was the one that set him furthest from everybody else, and it was not close.

Every version of it sits up there with it. The one where the shortstop fires straight back to first for the double play, which the New York Post called unparalleled in 2022. The one where the pitcher takes that return throw. And the bunt version, which he ran at six times the normal rate.

The play that is missing

Start with something so ordinary it barely registers. A bunt rolls toward the left side, the third baseman charges in, picks it up, throws across to first. Every team in baseball runs that play dozens of times a season.

In the games Keith Hernandez played first base, it happened 47 times. Over that many games anywhere else in the league, a third baseman would have made it about 76 times.

Close to four in ten of those plays never happened.

Something was getting to those bunts first, and the second number says what. A first baseman of that era handled about a fifth of the bunts his team faced. Hernandez handled a quarter of his, 276 of the 1,202 hit while he was out there.

His share went up. The third baseman’s went down. Nobody wrote down where either man was standing, but somebody was picking up balls that were not supposed to be his.

The Hal Chase play

There is one sentence about his defense that everybody repeats. It shows up word for word on more than one biography site, which means somebody wrote it once and everyone else copied. It says he revived Ferris Fain’s practice of charging to the third base line on bunts.

Picture that. The third-base line runs from home plate out to third base. A bunt dropped along it is the third baseman’s ball in every coaching manual ever printed. A first baseman going to get one has left his own base, run past the pitcher’s mound and crossed most of the infield to take a ball that belongs to somebody else.

Behind him the whole infield rotates. The second baseman sprints over to cover first, so the bag is not abandoned. The third baseman stays home on his own bag waiting for a throw instead of charging the way he normally would. Four men end up somewhere they do not usually stand, all at once, on the strength of one man’s decision to go.

Get it right and the lead runner is erased. Get it wrong and nobody is out anywhere, and a bunt that should have cost the other team an out has handed them a free runner instead.

It was not Fain’s invention either. His own managers called it the Hal Chase play, after a first baseman who quit in 1919, and they wished he would stop doing it. Jimmy Dykes, who managed him in Philadelphia, said Fain tried it about six times in a season and that the best throw Fain made hit a customer in the sixth row.

So how do you count a play like that?

Every bunt in those seventeen seasons has its own line in the record, and that line names the fielder who picked the ball up and says which runner was put out where. So you go looking for the lines where the fielder is Hernandez and a runner was forced at third base.

There are six of them. In seventeen years.

Whether six is a lot takes one more step, and it is the step that makes the whole thing work. There were 1,202 bunts hit while Hernandez was standing at first base. Take every other first baseman in the game across those same years, work out what fraction of their bunts ended with the lead man forced at third, and apply that fraction to his 1,202. What you get is the number a completely ordinary first baseman would have produced with exactly the chances Hernandez was given.

It comes to two, maybe three.

In his twelve full seasons he got five, where an ordinary first baseman would have got two or three. A gap that small turns up by accident about one year in nine, which is another way of saying it is not a gap.

Do the identical arithmetic for Ferris Fain and he lands fourteenth, three spots behind Hernandez. Neither man can be told apart from an average first baseman at the play they are both supposed to have owned.

The one it does pick out has not come up once. Jeff Bagwell forced the lead runner at third on a bunt nineteen times, where six was expected. That is not a small gap. The odds are about three in a hundred thousand, and nobody has ever written it down.

One thing this counting cannot do. It only sees the times it worked. A charge that ends with the batter out at first, or with everybody safe, leaves no sign the man even tried, and thirty-eight of the bunts Hernandez fielded ended with everybody safe. How often he went for it is a question nobody can answer, for him or for Fain.

The night Gary Carter played third base

The famous one only happened because of a fistfight.

July 22, 1986, Riverfront Stadium. Dave Parker dropped a fly ball to let the Mets tie it, and an inning later Ray Knight punched Eric Davis at third base and both benches emptied. When the umpires were done, the Mets had lost Knight and Kevin Mitchell, and Darryl Strawberry was already gone for arguing balls and strikes. Davey Johnson had nobody left on the bench. He sent Gary Carter out to play third base, a position he appeared at three times in his entire career, and put a relief pitcher in the outfield.

Twelfth inning. Nobody out, runners on first and second. Carl Willis, a pitcher, squares to bunt.

Hernandez comes in front of the mound, picks it up near the line, and throws to Carter at third for the force. Carter, who has never made this play in his life, turns and fires across the diamond to Tim Teufel covering first.

A 3-5-4 double play, in the twelfth inning of a fourteen-inning game, with a catcher at third base because of a brawl.

Hernandez told the Post about it years later. That play was really easy for me, because I was just playing it aggressively. He was already in, the turf was fast, and the only thing worrying him was whether Carter could make the throw. Carter made it perfectly.

Watch it.

Second

In 2022 the Post asked Hernandez where he ranked among defensive first basemen. I would have to say I’m in the top five, he said. I have to believe that.

So I went and checked.

I scored the five things a first baseman does with the ball, and I scored every man not against all of history but against the other regulars playing his own season, so nobody is punished for being born in the wrong decade. Everybody since 1908 who played enough to count.

Hernandez comes out second, behind Todd Helton, by a margin too thin to mean much. His 1979 season in St. Louis is the twenty-eighth best single year anyone has ever had. I moved the bar for who counts as a regular up and down, and he never fell below third.

Then something I was not looking for. A researcher who has never touched any of this ran thirty first basemen through a scoring system in 2023, and Hernandez finished second behind Frank Thomas. He ran the same thirty through a completely different system, and Hernandez finished second again, behind Jeff Bagwell.

Three ways of counting. Three different men in first place. Hernandez second in every one.

They are not measuring the same thing, and two of them are scoring his bat while mine never touches it. But he guessed top five, and he was selling himself short.

The bunts that never came

MLB.com wrote a few years ago that Hernandez being out there often talked teams out of bunting at all.

That is the best thing anybody can say about a defender, and I cannot check it.

He never sat down, which is the problem. To find out whether teams bunted less against him you need the games he missed, and in a full season those add up to three or four bunts, sometimes nine. Nothing can be built on nine bunts.

Underneath that is a problem no database will ever solve. Say a manager in 1985 looks out, sees Hernandez creeping in, and decides not to bunt. Nothing happens. The hitter swings away. That decision leaves no mark anywhere. Not in the box score, not in the play-by-play, not in any record that has ever been kept. The sign was never flashed, so there is nothing to count.

Everything baseball writes down is a record of what happened. If Keith Hernandez frightened anybody out of bunting, all the evidence is in the things that did not.

Call it the Hernandez

For seventeen years Keith Hernandez stood at first base and kept making a choice almost nobody else made.

A ground ball came to him with a man on first. The out he was owed was six feet away and could not miss. He looked past it, found the shortstop arriving at second, and went after the harder one.

He did that more than any first baseman in the box scores. He turned the throw back into a double play more than the men around him did. And he handled a quarter of every bunt hit while he was out there, in a game where a first baseman was expected to handle a fifth.

None of it ever got a name, because nobody was counting. On a scorecard it is two digits, a 3 and a 6, and nobody has ever said them out loud.

It deserves a name. Call it the Hernandez.

Where this comes from

Everything counted here comes from the Retrosheet play-by-play, 1908 through 2025, parsed one play at a time and checked against the box scores, with no mismatch on first-base putouts in any decade.

Everything not counted comes from somewhere else, and here is where.

Range factor was published by Bill James in the 1977 Baseball Abstract, which he had devised two years earlier. The 1872 Philadelphia Athletics scorekeeper who was already doing it that way is reported by American Heritage.

The positional penalties are published by Baseball Reference and FanGraphs in their own explanations of how they calculate wins above replacement. Baseball Reference has been quoted at both 9.5 and 10 runs per full season depending on when you look, FanGraphs at 12.5.

The reprinted sentence about Ferris Fain appears word for word on allamericanspeakers.com and baseballbiography.com. An exact-phrase search turns up no others and no original author.

The Hal Chase name and the Jimmy Dykes quote come from a Pitcher List article about Fain, which cites Dykes speaking to Pat Harmon of the Cincinnati Post and Times-Star in 1967. I have not seen the 1967 original. The Hal Chase name also appears in Fain’s Baseball Reference bullpen entry, independently.

Both Hernandez quotes and the word unparalleled are from “The Defense Rests,” New York Post, July 3, 2022.

Carl Willis as the batter on the 1986 bunt comes from MLB.com’s own list of Hernandez’s career moments. The twelfth inning, the two runners and the double play come from my own copy of the play-by-play.

The claim that his presence discouraged bunting is MLB.com’s, from that same list.

The 2023 study ran in the Baseball Research Journal.