Save evolves from stat to game-changer

April 12th, 2017

One day in 1959, a 33-year-old baseball writer named Jerome Holtzman got angry. Holtzman had lived a hardscrabble Chicago life. He'd spent much of his childhood in an orphanage, an experience he described, surprisingly, as delightful.

"There was a ballfield beside it, and we had ice cream every Friday night," Holtzman said. "My wife, Marilyn, says there has to be something wrong with someone who likes an orphanage."
Holtzman had, through sheer effort, worked his way up at the Chicago Sun-Times from copy boy to prep sportswriter to baseball scribe -- with a stint in the Marines in between. What Holtzman had developed through his developing years was an instinct for a fight and a furious sense of right and wrong. In later years, for instance, he would rail against the correction of historic baseball stats. To him, the numbers had become canonized, they were supposed to be left alone. When it was calculated that Honus Wagner's hit total was actually 3,420 and not the 3,430 that had been in the record books, Holtzman called it "statistical grave robbery."
This exchange, recorded in Alan Schwartz's excellent "The Numbers Game," between Holtzman and Baseball Encyclopedia editor Rick Wolff is illustrative of Holtzman's passion:
Holtzman: You don't understand. It's baseball history.
Wolff: Don't you want it to be accurate?
Holtzman: But who gives you the right to CHANGE these things?
That was later, though. In 1959, Holtzman was just beginning as a baseball writer. His passion already burned. And Holtzman knew, just knew, that Pittsburgh reliever Roy Face's 18-1 record was a farce. Face finished seventh in the National League MVP Award voting that year because of that gaudy won-loss record, ahead of Frank Robinson among other legends. Face was being celebrated in a way that made Holtzman fume. He went to the box scores and found that Face had allowed the tying or go-ahead run himself in 10 of his 18 victories. The win stat, Holtzman realized, lied. Relievers needed something else to define them, a whole other statistic.
That's when Holtzman invented the save.
* * *
The save has fundamentally changed baseball in the past 50 years. This was not Holtzman's intention, of course. He just wanted a quick and easy statistic that measured a relief pitcher's contribution. Holtzman had no idea that his little invention would create a whole new kind of ballplayer, earn unsuccessful starters and pitching specialists hundreds of millions of dollars, inspire a generation of young men to throw 100 mph and get managers to reinvent how they use their pitching staff.

But this is something that gets to the heart of baseball ... and time-travel movies: The tiniest changes can have enormous effects.
Holtzman was not the first person to tinker with the save concept, by the way. Allen Roth, the famed statistician for the Brooklyn Dodgers, played around with the idea back in the early 1950s, when his team had superb pitchers like Joe Black, Jim Hughes and Clem Labine, who mostly pitched relief.
But it was Holtzman who gave the save shape and, more, bite. He was a prominent writer for The Sporting News, appearing in 1,000 consecutive issues at one point, and he used the Bible of Baseball as a pulpit to spread the word about the save. Starting in the early 1960s, Holtzman started giving out a "Fireman Award" to the best reliever -- based on reliever wins and his invention, the save. And he preached the value of the save whenever he could.
Holtzman only rarely explained his rules for the save -- after doing it a couple of times, he expected his readers to know it. He felt like his save rule was so simple and logical that it sold itself. It was a two-part rule, with two exceptions to the second part.
• Save | Save opportunity | Relief win | Blown save | Save percentage
Part 1: A save is only awarded when a team wins a game and can only go to a pitcher who does not get the win.
Part 2: A relief pitcher shall not get a save unless he faces the tying or go-ahead run ... with two exceptions.
Exception 1: A relief pitcher does qualify for a save if he comes in with a two-run lead in the final inning and pitches a perfect inning.
Exception 2: A relief pitcher does qualify for a save if he come in with a three-run lead and pitches two or more innings and finishes the game without giving up the lead.
You will notice that there are a few differences between Holtzman's rule and today's save rule. In the Holtzman rule, a reliever did not need to finish the game to get a save. In fact, more than one pitcher would qualify for a save in any given game, at which point the official scorer (or Holtzman himself) would decide which pitcher deserved the save more.
There were other quirks; Holtzman's "perfect inning" was unclear. Did he mean a scoreless inning? A 1-2-3 inning? What if a fielder made an error? Holtzman also made it so the reliever could create his own save opportunity. If a reliever came in with a four-run lead and then loaded the bases with walks, he was in line for the save if he got out of the jam.
The Sporting News -- and the Baseball Writers' Association of America -- worked hard to make Holtzman's save an official statistic. In 1964, the BBWAA appealed to both leagues to give the save a try for a year, sort of a test run. The American League tried and quickly abandoned it.
The NL wouldn't even try.
* *
After the 1968 season, the famed Year of the Pitcher, there was a bit of a panic that baseball was losing its mojo. Rules were changed and amended in a furious attempt to pump some life back into the game. Mounds were lowered. The strike zone was adjusted. Talk of a designated hitter began. And both leagues decided to adopt the save rule.
Thing is, it wasn't Holtzman's rule. The doyens of baseball thought that was too complicated. And so they instituted the simplest possible save rule: A reliever earns a save if he enters the game with a lead and finishes the game having protected that lead.
That's it. It was a rule that nobody liked. The traditionalists didn't even want a save. And the people who did want the save thought this was a stupid version of it. The problem, as you probably noticed right away, is that there is nothing in the rule about how long the pitcher had to pitch or how close the game needed to be. A reliever could come into a game with his team leading, 30-0, get one out to end the game and he was credited with a save. Baseball stuck with this rule for five seasons.
Then in 1974, the rule was changed -- but still not to the rule we know today. The 1974 version of the save rule stated that for a pitcher to get a save he had to either.
1. Enter a game with the tying or go ahead run on the bases or at the plate and preserve the lead.
2. Pitch three effective innings while preserving a lead.
It was incredible; this rule was even worse than the 1969 version.The influential baseball writer Leonard Koppett tore it apart in a Sporting News column months before the season even began. The 1974 rule managed to take some of the worst parts of previous rules and mash them all together. By this rule, a reliever did not have to finish the game meaning, again, more than one pitcher could qualify for the save. Also a pitcher could get eight straight outs with a two-run lead and not get a save, while a pitcher could pitch "effective" baseball for three innings with a 20-run lead and get the save.
That version only lasted one year. Then, finally, in 1975, the modern save rule was put into place:
• Now the reliever has to finish the game, meaning only one pitcher can qualify for a save.
• Before the relief pitcher needed to face the tying run ... now the rule is that if the tying run is on deck, it's considered a save opportunity.
• Now a pitcher, given a three-run lead, has to only pitch one inning to get the save.
This last part of the rule was invented just for the 1975 ruling (Holtzman had insisted that a pitcher with a three-run lead pitch two innings to get a save). And it is the one-inning, three-runs-or-fewer save that would alter how baseball is played.
Question: Do you think the save has changed baseball?
Cubs general manager Theo Epstein: "Yeah, definitely. Yeah. I don't think that's a unique phenomenon. Almost every human endeavor, there are these frameworks created, some based on common sense, some based on merit, some based on logic, others based on accidents or maybe even impure motivations, that define your future in significant ways."
* *
In 1974, just as the save rule was being modified and codified, Mike Marshall became the first relief pitcher to win the Cy Young Award. In '77, two years after the save rule was finalized, Sparky Lyle became the first American League relief pitcher to win a Cy Young Award.
Two years later, Bruce Sutter won the NL Cy Young Award. Two years after that, in a strike season, Rollie Fingers became the first reliever to win the Cy Young Award and the MVP Award. Three years after that, Willie Hernandez became the second.
What was happening here? There were a couple of things. One, relief pitchers were becoming a more integral part of the game. And two, voters and baseball analysts had a relief-pitching statistic they trusted. Over the next decade, three more relief pitchers -- Steve Bedrosian, Mark Davis and Dennis Eckersley -- all won Cy Young Awards, essentially because of their imposing save numbers. Eck won the 1992 AL MVP Award despite pitching just 80 innings. Well, he had 51 saves.

But the save had a much larger effect than awards handed out or relievers getting previously unimaginable recognition. Davis signed what was then a huge deal with Kansas City after winning the 1989 NL Cy Young Award; teams were now beginning to spend for saves. But more, teams began to adjust their strategies to get more saves. How do you think Eckersley got 51 of them? He came into the eighth or ninth inning of just about every game that the A's led -- and the A's were a dominant team in those days. In 19 of Eck's saves, the A's won by three runs or more.
Eck begat the most prolific generation of closers in baseball history with Mariano Rivera and Trevor Hoffman both topping 600 saves for their soon-to-be Hall of Fame careers. It wasn't just the save that motivated teams, of course. One-inning pitchers threw harder. Their ERAs were much lower. And managers found safety in having a closer to pitch the ninth; nobody could question them for using their best reliever in the ninth inning.
"In baseball -- and I'd say in business, too -- where there's so much unknown, we hold on to things that are definable, things that we think are knowable," Epstein said. "A lot of them make sense. But some of them we just give in to our need for the tangible, something that we can count."
* *
Bill James -- because he's Bill James -- put together a little chart that showed top the relievers' usage from the 1950s to today. He found, first of all, that with every passing decade, relievers' outings have gotten shorter:
Average outing of top relievers by decades:
1950s: 6.12 outs
1960s: 5.25 outs
1970s: 5.19 outs
1980s: 4.72 outs
1990s: 3.67 outs
2000s: 3.27 outs
2010s: 3.07 outs
You see the shrinkage there -- the biggest drop was from the 1980s to the 1990s. Why? That's the decade of Eckersley, the decade that managers started to believe in the one-inning closer.
Now, here are the decade-by-decade percentages for how often a top reliever's appearance ended up with a save (by today's rules):
1950s: 32 percent
1960s: 38 percent
1970s: 39 percent
1980s: 44 percent
1990s: 53 percent
2000s: 56 percent
2010s: 63 percent
You can see clearly how baseball has been gravitating toward the save. Closers are now used almost exclusively for one inning and with the team ahead by three runs or fewer. This is, word for word, in line with the save rule. Holtzman came up with a concept to measure baseball. Instead, his rule had come to define it.
* *
In the postseason last year, Cleveland manager Terry Francona broke a lot of the unwritten rules. He rarely used either of his top relief pitchers -- and -- in pure save situations. Instead, Francona asked both of them (particularly Miller) to pitch whenever they were needed and for as long as they could be effective. In Allen's 10 postseason appearances, he pitched more than an inning six times. Miller pitched more than one inning in all 10 of his appearances.

Miller was the superweapon -- Francona used him at various times in the fifth, sixth, seventh, eight and ninth innings. But Francona was not reluctant to use Allen in the seventh inning, if the situation called for it. It was a shattering of the protocols that have built up for more than 50 years, ever since Holtzman came up with his save idea.

Now, people around baseball argue about what Francona did. Most acknowledge it can be effective as a short-term strategy, but it cannot work over a long season. Why not? The reasons vary. Some say it is because relief pitchers are creatures of habit and need to know their roles -- you're the closer, you're the setup guy, you pitch the sixth inning, etc. Others say that relievers will wear out with the overuse that is unavoidable in a Francona-like free-for-all. And still others say that pitching the ninth inning is just different from every other inning and requires a different and rare mind-set.
But are any of those things true? Yes, the game has changed, and the relievers in baseball today have grown up in the era of the one-inning closer. But are relief pitchers today incapable of pitching an extra 20 innings a year? Are they unwilling to give up the cheap save -- one inning, team up by three runs? Are the best so inflexible that they cannot muster the resolve to pitch the team out of a jam in the sixth inning or seventh inning? Are managers and GMs so convinced of ninth-inning intensity that that they will always insist on having their highest-paid reliever pitch only in that inning?
All this is worth watching over the next few years. In the meantime, the save reigns over baseball. It is worth considering this: What if baseball had made the rule so that you needed to pitch two innings to get the save. Would the game look different today?
"Probably," Epstein said. "But you can say that about a lot of things. What if whoever invented batting average -- [Albert] Spalding or whoever ... it was -- had decided to count walks as hits? The whole history of the game would have been completely different."