To no one’s surprise, Mariano Rivera was elected to the Baseball Hall of Fame this week. The former Yankees’ closer became the first ever to gain unanimous enshrinement in the BBWAA election, receiving votes from each of the 425 writers who turned in ballots. They were recognizing his record 652 regular-season saves and microscopic 2.21 ERA, along with his 42 saves and ridiculous 0.70 ERA in post-season, helping the Yankees to five world championships during his tenure. Rivera appears to meet the highest standards of integrity, sportsmanship, and character as well.
I don’t really have a problem with Rivera’s election; more so with the likes of Bruce Sutter, Lee Smith, and Trevor Hoffman. In my opinion, the closer is the most overrated position in baseball. It is relatively easy to fill and succeed at, and it does not contribute nearly as much to team success as most people think. I wrote a chapter about this in Baseball Myths (Rowman & Littlefield, 2012), and will add to it here.
We now have eight relievers in the Hall of Fame, nine who have won Cy Young Awards, and four who have won MVPs. And relievers have also had their own special awards, like the Sporting News’s “Fireman of the Year” and the Rolaids Relief Man Awards. Why is it that the men who play the fewest innings are eligible for the most awards? Besides awards, there is plenty of money available to successful relievers. Rivera earned more than $169 million during his career, an average of about $132,000 per inning.
Do they deserve all this attention and cash? Can a relief ace – who toils for perhaps 70 innings per season – be as valuable as a starting pitcher (who pitches 210) or an everyday player (who plays in 1,400)?
Let’s look at just the pitchers. For the 70-inning closer to be as valuable as the 210-inning starter, you have to believe that the closer’s innings are three times as important as the starter’s. Bill James, after simulating thousands of games, concludes that “The real value of a run saved by a modern reliever is 73% to 89% greater than that of a run saved by a starting pitcher.” In other words, current relievers’ innings are not three times as important as starters’, but “only” about 1.8 times as important, partly because relief aces aren’t put in true pressure situations as much as they used to be.
Researcher Mark Armour summarizes today’s trend: “The modern closer is a role designed to avoid, at all possible costs, pitching in the most difficult situations.” It used to be that the relief ace was brought into the game when it was on the line – be it the fourth or the ninth inning. It could have been when the team was up a run or two, tied, or even down a run or two – his job was to maintain the team’s chance to win. And he often pitched two or more innings in a game. Now, the closer comes in almost exclusively at the start of the ninth inning, only when the team is ahead, often by two or three runs.
Saves were first officially defined in 1969. After the 1974 season, the save rule was deemed too tough, and softened to its current definition. A pitcher is credited with a save “when he meets all three of the following conditions: (1) He is the finishing pitcher in a game won by his club; and (2) He is not the winning pitcher; and (3) He qualifies under one of the following conditions: (a) He enters a game with a lead of no more than three runs and pitches for at least one inning; or (b) He enters the game… with the potential tying run either on base, or at bat, or on deck…; or (c) He pitches effectively for at least three innings.”
Over the ensuing decades, the save rule became a strategy. If a team is leading by three runs or less going into the ninth inning, the manager brings in his closer, no matter how well the current pitcher is doing. This has led to single-season save totals in the 50s and 60s. But, under the 1974 definition, today’s best closers, even Rivera, would be credited with only about 12-15 saves per year.
As Tim McCarver wrote in Baseball for Brain Surgeons and Other Fans, “Currently, when the weak middle men start to get battered around, few managers are bold enough to bring in their star closers because agents are insistent that their clients pitch only an inning or an inning and a third each appearance… [T]oday’s agents aren’t going to allow their closer clients to be used frequently in nonsave situations.”
Bill Felber, in the fourth edition of Total Baseball, analyzed whether this newer strategy was more effective. Felber examined all games in the 1952, 1972, and 1992 seasons, in which a team had a lead of one, two, or three runs after seven or eight innings – how often did the teams win? There was little difference between the three seasons, but the biggest difference was with a one-run lead after eight innings, the most critical closer situation. In 1952, those teams won 89.1% of the time, and in 1972, 89.4% of the time. But in 1992, with the multi-million dollar closer going against the other team’s mop-up guy, the teams won only 85.3% of the time in this situation.
Felber paid special attention to Dennis Eckersley, the American League’s Cy Young and MVP Award winner in ’92. Eck was credited with 51 saves that year – but only 12 times did he face the tying run, and 37 of those 51 saves were vultured while working with at least a two-run lead. His manager, Tony LaRussa, often said that he preferred to bring his ace reliever in only in the ninth inning with nobody on base. Eckersley was the best in the business – but the business was nowhere near as important as it used to be, or as people still believe it to be.
As Armour says, “Once LaRussa did it with Eckersley, everyone else said, ‘Hey, I don’t want to have to think during a game either, I want one of those guys.’ … There is still this perception that pitching in the ninth inning with a two-run lead takes a certain mental toughness, whereas entering in the eighth inning of a tie game does not.”
Retrosheet founder David Smith did a similar (but much more exhaustive) study to Felber’s, presenting it at SABR’s 2004 Convention and summarizing the results in the Fall 2009 Baseball Research Journal. Smith examined the scoring patterns in 122,906 games. As the late David Vincent summarized, Smith “researched late-inning leads … from 1944 to 2003, and an additional 14 seasons prior to that span. What he found is that the winning percentage for teams who enter the ninth inning with a lead has remained virtually unchanged over the decades. Regardless of the pitching strategy, teams entering the ninth inning with a lead win roughly 95 percent of the time. That was the exact rate in 1901 and that was the rate 100 seasons later. In fact, the rate has varied merely from a high of 96.7 percent in 1909 to a low of 92.5 percent in 1941. But I know what you’re thinking. That study applies to all leads, including big ones. But what about the slim leads, the ones defined as ‘save situations’? Glad you asked. Because Smith looked at those leads as well. And what he found is winning rates for those leads have also remained constant – one-run leads after eight innings have been won roughly 85 percent of the time, two-run leads 94 percent of the time and three-run leads about 96 percent of the time.”
James points out that “Each run saved in a tie game has more than eight times the impact of a run saved with a three-run lead. If you use your relief ace to save a three-run lead in the ninth inning, you’ll win that game 99% of the time. But if you don’t use your ace in that situation, you’ll [still] win the game 98% of the time.”
Another baseball author, Gabriel Schechter, has done extensive research on relief pitching. In one project, he compared the first ten seasons of the save rule (1969-78) with the ten most recent seasons at the time (2000-09), checking every game in which the starting pitcher finished the seventh inning with a lead of one, two, or three runs. In the former period, the starter finished the game 51.5% of the time; in the latter, only 9%. Yet the results, in the form of team winning percentages, were almost identical: .850 in 1969-78, .846 in 2000-09.
“The way I see it, the save rule has devolved into a rationale for managers approaching pitching deployment in an ass-backwards way,” Schechter wrote in 2006. “It used to be that the manager relied on his starter to go as far as he could, using relievers only when necessary. Today, when [then-Yankees’ manager] Joe Torre (to use just one example out of a possible 30) goes to the ballpark, he is hoping that he will wind up using Mariano Rivera in the 9th inning. He’s also hoping that he’ll use his best set-up guy in the 8th inning. And he’s depending on several other relievers, including the lefty-lefty specialist, to get him through the 7th inning. In other words, today’s manager is managing from the end of the game backward toward the start, instead of the other way around as it was 30-40 years ago. The starting pitcher is no longer expected, trained, or conditioned to give his manager more than six or seven innings. I think that’s a big mistake in philosophy. When we ask for less, we get less; when we accept mediocrity, we foster mediocrity. The study I presented at the [2006 SABR] convention demonstrated that using a bevy of relievers after the 7th inning to do the job that the starter used to do by himself does not generate more wins. All it generates is jobs for more relievers with less talent than the pitchers they replace.”
If you think about it, relief pitchers are, by and large, guys who aren’t good enough to be starting pitchers. If somebody had the ability to give you 210 quality innings, why would you use him for only 70? You wouldn’t, which should remind you that a reliever is just a role-player: important, but not to be confused with the best pitchers in the game. Does anyone think that guys like Randy Johnson, Pedro Martinez, or Justin Verlander wouldn’t have been dominant as closers, if management were foolish enough to waste them in that role?
Occasionally, a team has tried to convert a star reliever into a starter, with poor results:
• After a monster year in the White Sox’ bullpen in 1975 (a league-leading 26 saves and a 1.84 ERA over 142 innings), Goose Gossage became a starter in 1976. He went 9-17 with a 3.94 ERA, which ranked 39th of 43 ERA title qualifiers. He never started another game and carved out a Hall of Fame career as a reliever.
• After three strong seasons in the Braves’ bullpen, Steve Bedrosian was given 37 starts in 1985 and went 7-15 with 111 walks and a 3.83 ERA, which ranked 34th of 37 ERA qualifiers. He returned to the pen for the rest of his career and won the 1987 Cy Young Award.
• Between 1999 and 2002, the Reds’ Danny Graves averaged 30 saves with an 8-5 record each year. Converted into a starter in 2003, Graves went 4-15 with a 5.33 ERA. Back in the bullpen in 2004, Graves saved 41 games and made the All-Star team.
Sporting News columnist Todd Jones, who was credited with more than 300 saves during his pitching career, makes no pretense about the value of his former role. In the January 19, 2009 issue, he wrote, “Relievers are in the bullpen for a reason: They can’t start… Most relievers either throw hard or have a trick pitch. Their command doesn’t need to be as sharp as a starter’s. Because relievers usually see a hitter only once a game, they don’t have to worry as much about setting him up. By the time the hitter knows a reliever isn’t locating well or doesn’t have his good stuff, the inning often is already over. Another reason relieving is easier than starting: It’s much easier to have one clean, scoreless inning three days a week than it is to go six or seven good innings once every five days. Think of it this way: Do you have a better chance of beating Tiger Woods on one hole or outlasting him for 18? I’ll take my chances on one hole. That logic worked for me for 16 seasons.”
Jim Kaat, who started 625 big league games, and relieved in 273 others, expresses similar sentiments: “As a starter, you’re a pitcher. As a reliever, you’re a thrower… Most of the time, your relievers are specialty-pitch pitchers. They come in, and they throw that one pitch, time after time, and they only have to do it for an inning or two… As a starter, you’ll have four pitches and you’ve got to learn to use them all and change speeds.”
Former Cy Young Award-winner Mike Marshall also minimizes the modern role of closer. “If you pitch fewer innings, [the hitters] don’t get to see what you do as often, and it’s hard for them to make adjustments,” said Marshall to Sports Business Daily. “So, pitching 80 innings one inning at a time with a lead? That’s a walk in the park. Billy Beane made a point. He said that if you want to get something for nothing, find a guy that can throw a little bit good and throw strikes, use him in a closing role and pump up a lot of saves, and then you can sell him for something very valuable because that’s not a very difficult man on your team to replace. He’s right! It’s the easiest gig in baseball.”
Journalist John Shiffert analyzed relief aces in the 2001-05 time-frame, noting that the Yankees were the only team to have the same closer (Rivera) in each of those five years. Pitcher after pitcher stepped into the role for different teams – the average club had 4.9 different closers in the five years – and the majority of the closers had success. Shiffert concluded that “In fact, practically anybody can close.”
The Dodgers’ Greg Gagne received deification for his record of 84 consecutive saves converted between 2002 and ’04 (disregarding his blown save in the ’03 All-Star Game, which cost the NL home-field advantage in that year’s World Series). Gagne is a fine example of a relief star who was a failure as a starting pitcher: after going 10-13 with a 4.70 ERA in parts of three seasons as a starter, he was demoted to the pen, and suddenly emerged as baseball’s best closer. But his record streak, as remarkable as it was, encompassed just 87 innings, less than ten full games worth. I’d put it in the same class as records for consecutive errorless chances at various positions – notable, but not the kind of thing that gets a guy to Cooperstown.
Getting back to Mariano Rivera, many people thought he was the Yankees’ key player during their 1996-2001 dynasty, cementing his Hall of Fame résumé. But who was building up all those leads for Rivera to protect? The Yankees were successful because they scored an average of 883 runs per season during that period, and allowed only 732, as compared to the league averages of 827-827. Yes, Rivera was part of the reason the Yankees allowed fewer runs than average, but only a part. The Yankees had 8,680 innings on defense in those six years, and Rivera pitched 467 of them. Per season, that’s 1,447 innings for the Yanks, 78 for Rivera – 5.4%. He had eight other guys helping him while he was out there, and he contributed zero to the Yankees’ offense.
People used to say “What are the Yankees ever going to do when Mo leaves?” Rivera had his usual 44 saves in 2011, and the Yanks won 97 games. In 2012, Rivera went down with a season-ending injury in early May. So Rafael Soriano stepped into the role, saved 42 in five months, and the Yankees won 95. Rivera came back for his final season in 2013, and saved 44, but the Yanks dropped to 85 wins. Then in ’14, David Robertson stepped into the role, saved 39, and the Yankees won 84. Somehow, even when the greatest reliever of all time was lost, life went on with hardly any difference in the team performance.
And here is why. Rivera had 652 saves and 80 blown saves, an .891 percentage. A blown save does not necessarily lead to a loss; the Yankees probably won about 93% of those 732 games. That sounds pretty good, until you consider the historic expected win percentages in save situations: a one-run lead gives a winning percentage of about 84%, a two-run lead 93%, and a three-run lead, 97%. Rivera’s save situations were pretty evenly split between one- (36%), two- (33%), and three-plus- (31%) run leads. Based on teams’ average winning percentages in those situations, we would have expected the Yankees to win about 91% of those 732 games, regardless of who was finishing them. So, having the greatest closer of all time, the way they used him, was worth a two percent increase in team win probability – about one extra win per year.
Again, I’m not picking on Mariano Rivera. He didn’t decide how he was going to be used, and with the possible exception of Hoyt Wilhelm, he was the best ever at what he did.
But what he did is far overrated by most baseball observers.
Dizzy Deane 