The Baseball Save: Why the Stat Exists, Who Invented It, and Why It's Still Controversial

Most baseball stats have obvious names. Strikeouts strike out batters. Home runs go home. But "save" is a word that implies a rescue, a near-miss, something going wrong that gets fixed. So why does a pitcher who enters a game with a comfortable lead and throws three clean innings get credit for one? The answer involves a Chicago sportswriter, a Hall of Fame closer, and decades of rule tinkering that never fully solved the original problem.

Jerome Holtzman and the Problem He Was Trying to Fix

The save was invented by Jerome Holtzman, a Chicago sportswriter for the Sun-Times, in the late 1950s and early 1960s. Holtzman was frustrated by how wins were distributed in baseball. Relief pitchers who came in to protect a lead, sometimes throwing three or four critical innings under pressure, had no meaningful statistic to show for it. If they didn't give up the lead, the starting pitcher took the win. If they came in during a tie game and their team scored, the reliever got the win, which felt backward.

Holtzman started tracking saves informally and eventually convinced The Sporting News to publish them in 1960. The stat caught on enough that MLB officially adopted it in 1969. The original definition was loose: a reliever got a save if he finished a game his team won and wasn't the pitcher of record for the win.

That definition was too loose. Pitchers were getting saves in lopsided games with no real pressure involved, which defeated the entire point.

How the Rules Evolved

MLB tightened the definition in 1974 to something closer to what we use today. The modern save rule has three qualifying conditions, and a pitcher only needs to meet one of them:

1. He enters with a lead of three runs or fewer and pitches at least one inning.
2. He enters with the tying run on base, at bat, or on deck.
3. He pitches effectively for at least three innings, regardless of the score.

That third condition is almost never invoked in the modern era of specialized relievers. Most saves now are one-inning affairs, often the ninth inning, often against the bottom of the order.

The 1974 revision helped, but it introduced a new quirk: a pitcher can earn a save while allowing the lead to shrink to one run, as long as he entered with a lead of three or fewer and didn't lose it. Meanwhile, a closer who enters with a four-run lead in the ninth gets no save even if he throws a perfect inning. The stat penalizes managers for building big leads before turning it over to the bullpen.

The Rollie Fingers Effect

Rollie Fingers is probably the reason saves became culturally important. He won the 1981 AL MVP and Cy Young Award, making him the first reliever to win MVP in the American League, and the save was the shorthand fans and writers used to explain why. He retired with 341 saves, which was a record at the time.

After Fingers, teams started building their bullpen strategy around protecting the closer's save opportunities. The "closer" role as a defined, specialized position is largely a product of the save stat's existence. Before saves, relievers were often used in high-leverage situations across multiple innings and multiple days. The save, paradoxically, encouraged managers to save their best reliever for the ninth inning even when the eighth was the more critical moment.

Dennis Eckersley, who refined the one-inning closer role in Oakland in the late 1980s, became the template. When Eckersley recorded 48 saves in 1988 with a 2.35 ERA, the model was set. Every team wanted that guy. The save stat was now shaping roster construction and in-game strategy.

The Mariano Rivera Complication

Mariano Rivera holds the all-time record with 652 saves. He's also widely considered the best reliever in baseball history. But here's the uncomfortable part: Rivera's save total was partly a product of playing for the Yankees, one of the most offensively dominant teams of his era. More wins mean more save opportunities. A reliever on a losing team might be just as good and accumulate a fraction of the saves.

This is why advanced metrics people push back on saves as a measure of pitcher quality. A save doesn't tell you how difficult the situation was, how many runners were inherited, or what the win probability was when the pitcher entered. ERA and FIP apply equally to starters and relievers. Saves apply only to the pitcher who happened to finish a game the team won while meeting one of three criteria.

Why It's Still Around

For all its flaws, the save endures for a few reasons. It's simple to explain and easy to follow in a box score. It attaches narrative meaning to relief pitching in a way that ERA alone doesn't quite capture. And closer contracts are still heavily tied to save totals, which means players and agents have financial reasons to defend the stat.

The sabermetrics community has offered alternatives. Win Probability Added (WPA) measures how much a pitcher shifted his team's odds of winning on each pitch. Leverage Index (LI) quantifies how high-stakes a given game situation was. Neither has displaced saves in public conversation, partly because they require more explanation and partly because baseball fans are, as a group, somewhat resistant to discarding a century of statistical tradition.

Holtzman's original goal was to give relief pitchers credit for work that went unrecognized. He solved that problem. The stat he created just turned out to measure something slightly different from what anyone intended, and now it's too embedded in contracts and Hall of Fame debates to quietly retire.

A Few Facts Worth Knowing

Lee Smith held the all-time save record for about a decade before Trevor Hoffman passed him in 2006 and Rivera passed Hoffman in 2011. Smith, interestingly, waited until 2019 to get into the Hall of Fame, largely because voters weren't sure whether his career save total was enough on its own.

The lowest-leverage save in recent memory is probably any appearance where a closer enters in the ninth with a three-run lead, gets two quick outs, and allows a solo home run. That's technically a save. That's also a pitcher giving up a run in a low-pressure situation while getting full closer credit.

You can record a save without throwing a single pitch if the previous pitcher is removed due to injury or ejection and you inherit the lead. The rule cares about finishing the game, not about what you did to finish it.

One more: the blown save, the unofficial opposite of the save, doesn't count against a pitcher's ERA in any special way. A reliever can blow three saves in a week and still post an ERA that doesn't reflect that specific failure pattern. Which is another reason the stat is hard to use in isolation.

Holtzman was inducted into the writers' wing of the Baseball Hall of Fame in 1990. His save is still with us.