The following article is part of my ongoing look at the candidates on the BBWAA 2013 Hall of Fame ballot. For a detailed introduction to JAWS, please see here.
When I first cobbled together the system that became JAWS, just two relievers had been elected to the Hall of Fame: Hoyt Wilhelm in 1985 and Rollie Fingers in 1992. Since then, that number has more than doubled via the elections of Dennis Eckersley (2004), Bruce Sutter (2006) and Rich Gossage (2008). In theory, the larger group made it at least somewhat easier to sketch out a standard for relievers, but within the stathead community, a debate raged over how to properly value reliever contributions.
In particular, the focus has fallen on the concept of incorporating win expectancy and leverage — the quantitatively greater impact on winning and losing that a reliever has at the end of the ballgame than a starter does earlier — into valuation metrics. Smart minds have come down on both sides of the issue, and as they have, the JAWS methodology for examining relievers has evolved even more frequently than any other part of the system in an attempt to keep pace. Baseball Prospectus' Colin Wyers argued against the use of leverage and win expectancy to measure reliever contributions, objecting on the grounds that it was rewarding a player for things that were out of his control before he showed up (kind of like using RBIs to measure value). The Baseball-Reference.com version of WAR does contain a leverage adjustment, but it also measures relievers against a higher replacement level than starters, since they tend to allow fewer runs per nine innings.
They also throw far fewer innings, so the result is that even among elite relievers, WAR totals are much lower than for even average starters, particularly as the job of closer has evolved into a one-and-done night. Mariano Rivera, who holds the all-time lead in saves, has a career WAR of 52.7, a number that would tie him with Orel Hershiser and Frank Tanana for 74th all-time among starting pitchers. Trevor Hoffman, who has just seven fewer saves than Rivera, has been roughly half as valuable, with a career WAR of 27.0, which would rank 286th among starters, a hair lower than Livan Hernandez, Jason Schmidt and Johnny Podres — nobody you'd elect to the Hall of Fame. The peak scores of Rivera and Hoffman would stack up even less impressively if measured against starters, tied for 232nd for the former, and tied for 468th for the latter.
With the ongoing changes to JAWS, my verdict on Lee Smith, the top reliever on this year's ballot, has changed several times. Over the course of an 18-year career, the physically intimidating 6-foot-5, 220 pounder spent time closing for eight teams and earned All-Star honors seven times. He's been on the BBWAA ballot for 10 years, and has drawn a fairly consistent amount of support, debuting at 42.3 percent and staying within six percentage points of that in either direction before rising to 50.6 percent of the vote last year.
|Avg HOF RP||38.6||27.1||32.8|
A native of Jamestown, La., Smith was discovered by Negro Leagues legend Buck O'Neil, who spent decades as a scout for the Cubs. Chicago chose Smith in the second round of the 1975 draft, and spent the first four years of his minor league career trying him as a starter, but control problems, including walk rates exceeding 7.0 per nine in some years, led the team to change course. By September 1980, Smith's control had improved enough for the Cubs to bring him to the majors, and he spent the entirety of the strike-shortened 1981 season with the team. The Cubs briefly tried him as a starter in June 1982, then moved him into the closer role in the second half of the season. He converted all 15 save opportunities that came his way from July 23 onward.
The next year, Smith earned All-Star honors and led the NL with 29 saves while pitching to a 1.65 ERA in 103 1/3 innings, the type of performance that would alas become nearly obsolete with the Eckersley-driven move to the one-inning closer by the end of the decade. Four of Smith's saves were of three innings or more, 14 were of at least two innings and 19 were longer than one inning. His 4.7 WAR set a career high that he would never come close to matching, and ranked fifth in the league among all pitchers.
Over the next four years, Smith compiled a 3.25 ERA and averaged 33 saves, finishing in the NL's top five in each year, and helping the Cubs to their first postseason berth in 39 years in 1984. Alas, he surrendered a walkoff homer to the Padres' Steve Garvey in Game 4 of the NLCS, thereby evening the series at two games apiece; San Diego would win in five. That blip aside, from 1983-1987 Smith was the majors' second-most valuable reliever in terms of WAR, with his 14.3 trailing only Dan Quisenberry.
In December 1987, the Cubs traded Smith to the Red Sox for Al Nipper and Calvin Schiraldi. He pitched reasonably well in Boston, though again he struggled in the postseason, taking the loss in Game 2 of the 1988 ALCS against the A's. His workload continued to wane as the industry model moved toward a one-inning closer; in 1989 he threw just 70 2/3 innings, the sixth season out of seven he had thrown fewer than the year before.
Early in the 1990 season, Smith was traded to the Cardinals, and the next year, he broke Sutter's NL saves record with 47. Accompanied by a 2.34 ERA, he finished second in the Cy Young balloting behind Tom Glavine. He led the league in saves again with 43 in 1992, and in 1993 he put together a 46-save season that made him the first man to reach 400 career saves, taking over the all-time lead in the category from Jeff Reardon along the way.
He had been traded to the Yankees that August and moved on to Baltimore for the 1994 season and later made stops in Anaheim, Cincinnati and Montreal. The innings began to take their toll, and his managers limited his usage to about 50 frames a year, one inning at a time, to keep him effective. He spent his last two seasons, 1996 and '97, in a set-up role, with diminishing returns, finally retiring in 1998 after being released by the Royals in the spring.
From a traditional standpoint, Smith's case starts with his status as the number three guy on the all-time saves list; Hoffman broke his record in 2006, and Rivera took the crown in 2011. His 169 long saves — those of at least four outs — ranks fourth behind Fingers (201), Gossage (193) and Sutter (188). His 1,022 games pitched ranked third when he retired, behind only Wilhelm and Kent Tekulve, but he has since been bumped down to 11th.
Beyond that are Smith's seven All-Star selections and amazing string of consistency that followed him to virtually every stop on his 18-year ride. Until his abbreviated final season, his ERA+ was always better than league average, and was 32 percent better for his career. On the down side, his teams never went further than the League Championship Series, and he was bombed for an 8.44 ERA in four post-season appearances, taking two losses.
In terms of JAWS, the five closers in the Hall of Fame are strewn among the top 28 relievers, with Fingers ranking the lowest among that Cooperstown quintet. That ranking is at least somewhat skewed by pitchers' value as starters -- Eckersley, for instance, made 361 career starts -- but Smith, who made six starts and none after 1982, nevertheless ranks 16th in JAWS. Because of the way Eckersley's WAR total skews the standards, it's perhaps more instructive to note that Smith tops only two of the five enshrined relievers in career WAR and only one in peak WAR:
Even if one excludes Eck's numbers in calculating a standard, Smith would fall short of the averages of the other four (32.7 career, 24.4 peak, 28.6 JAWS) — a different result than I've gotten in the past, when Sutter was the low man on the totem pole and Smith fit behind Gossage and ahead of Wilhelm and Fingers. While it's fair to suggest that given the small sample size of relievers in the Hall, the voters are still in the process of fleshing out the standard, even if one imagines the inclusions of both Rivera and Hoffman in the not-too-distant future (the latter becomes eligible in 2016), Smith gets no closer to a standard that excludes Eckersley's numbers.