Harmon Killebrew and “Versatility”
The recent death of Harmon Killebrew prompted many touching reminiscences about a man with seemingly no enemies, despite carrying around the nickname of “Killer” for most of his life. (He did not really need a nickname; both his first and last names are unique in major league history.) By all accounts, he was a gentle and loving person, who also happened to hit home runs more frequently than anyone of his time. He hit 45 or more round trippers six times in the 1960s, while no other American League batter did it more than once. For baseball fans of a certain age, no player will ever better personify the word “slugger.”
Another interesting thing about Killebrew, perhaps unique among Hall of Fame players: he was repeatedly shifted between three defensive positions throughout his career, getting 44% of his starts at first base, 33% at third base, and 22% in left field. While many players shift positions along the defensive spectrum as they age, moving from shortstop to third base, or from left field to first base, Killebrew’s managers shifted their star hitter, nearly to the end of his career, depending mainly on the other players on the team. (It would be as if Tony LaRussa started playing Albert Pujols at third base. Oh, wait …)
1954-58. Forced to start his big-league career early because of the bonus rule, Killebrew spent parts of five seasons as a little-used infielder for the Washington Senators.
1960. Harmon remained at third until mid-season, when Lavagetto decided he needed to get Reno Bertoia into the lineup (or Julio Becquer out of it) and shifted Killebrew across the diamond to first base.
1961. With the franchise now in Minnesota, Killebrew spent the first half of the 1961 season splitting time between first and third, until Sam Mele became the skipper in mid-season and kept Killebrew at first. (I am not going to recite lots of offensive statistics, so just go ahead and assume that Killebrew hit 45 home runs and batted .260 with a bunch of walks, since he did that every year.)
1962. Just prior to the start of the 1962 season, the Twins acquired Vic Power, a great defensive first baseman, and moved Killebrew to left field for the first time.
1963. Left field.
1964. Left field. Tony Oliva took over in right field in 1964, and Power was discarded early in the season, creating a perfect opportunity to get Killebrew back to first base. Instead Mele shifted Bob Allison and left Killebrew in the outfield.
1965. Killebrew moved to first base (and Allison to left), but Harmon began shifting to third often by mid-season so that the team could play Don Mincher against right-handed pitchers. In early August Killebrew hurt his arm during a collision (while playing first base), but returned in September and played all seven games—at third—in the World Series.
1966. He played all 162 games, moving between third base, first base, and left field depending on who else Mele wanted to play. The Twins also had Cesar Tovar playing all over the field, leaving Mele about seven million possible defensive alignments. Tovar played this role for several years.
1967. Mincher was traded to the Angels, allowing Killebrew to play a full season at first base (160 games) for the first time in his career.
1968. A full-time first baseman again, Killebrew ruptured his hamstring in the All-Star game stretching for a throw on Houston’s AstroTurf (which was blamed at the time for the injury). When he returned in September Rich Reese had taken over at first, so manager Cal Ermer put Harmon (recovering from a severe injury) back at third base to play out the season.
1969. New manager Billy Martin took one look at the 33-year-old slugger coming off major surgery, and decided to return Killebrew to the 3B-1B role, allowing Martin options at the other corner spot. Harmon started all 162 games (96 at third base, 66 at first), drove in 140 runs, and won the MVP award, while the Twins nabbed the inaugural AL West title.
1970. Martin was replaced as skipper by Bill Rigney, who made Reese more of a full-time player. Killebrew started 129 games at third, but still managed 26 back at first.
1971. Killebrew again played both corner spots, though Reese’s poor season (.219) gave Killebrew several long stretches at first, where he started 82 times.
1972. For the first time since 1958 (when he played just nine games in the field), Killebrew played just one defensive position, first base. He was 36 and had slowed down quite a bit, though he could still rake (138 OPS+).
1973-75. With the advent of the designated hitter, the elderly Killebrew seemed to have a ready-made role. Unfortunately, the Twins also had a hobbled Tony Oliva, who needed the role even more. Killebrew eventually made it to DH, but spent his final three seasons fighting injuries and ineffectiveness.
OK, so the question is: how much defensive value did Harmon Killebrew have? According to bWAR, Killebrew’s cumulative defensive value was -7.6 wins, meaning that his place on the field cost his teams nearly 8 games on defense when compared with a replacement level player. Killebrew was a big guy, not fast, and no one ever accused him of being a good glove man. On the other hand, one wonders whether he could have been better on defense had he been allowed to play one position (preferably first base) for 15 years.
More importantly, did Killebrew’s ability to play multiple positions, often day-to-day, provide additional value to his team? In 1969 Martin played Harmon at third base 2/3 of the time so that Rich Reese could play first base. In an otherwise undistinguished career, Reese hit .322 with 18 home runs (good for a 139 OPS+), Killebrew had his best year, and the Twins led the league in runs. According to bWAR, Harmon’s (mostly) third base play cost the team 1.3 games on defense. This might be true, and Harmon’s isolated value might have been better had he just played first base all season and let Frank Quillici or someone play third. In order to get Reese’s bat in the lineup (or Mincher’s, Power’s, or Bertoia’s), Killebrew was asked to play a position he could not play particularly well.
It seems to me that Killebrew’s “value” to the Twins might have been greater than his statistical record might show.
Another player shifted around the diamond throughout his career was Pete Rose. Unlike Killebrew, Rose did not move day-to-day—he stayed in one place for several years before moving on. Also unlike Killebrew, Rose was an outstanding defensive player for part of his career, before being asked to move again. Rose came up as a second baseman in 1963, then moved to left field (1967), right field (1968), left field (1972), third base (1975), and first base (1979). Let’s examine his move to third base.
Rose won two Gold Gloves in right field, where he had good range though only a fair arm. He was moved to left field in 1972 largely in deference to Cesar Geronimo, a great defensive player with a cannon. In left field, Rose was outstanding. How outstanding? According to the defensive runs metric used on baseball-reference.com, here are the best outfielders in baseball over the years 1972-74, in aggregate.
Pete Rose 52
Other than Rose, these are all center fielders. As a hitter, Rose trailed only Willie Stargell, Cesar Cedeno and Reggie Jackson in batting runs among outfielders, making him every bit as valuable as he was famous.
Nonetheless, in May 1975 Sparky Anderson moved Rose to third base. The effect on the Reds was to replace third baseman John Vukovich, hitting .211 with zero home runs, with left fielder George Foster, who would hit .300 with 23 home runs. Rose continued to hit as well as ever, and the team won 108 games and the World Series.
Over the 1975 and 1976 seasons combined, Rose had the sixth highest total of batting runs in the major leagues, but rather than being worth two wins per season on defense (as he had been in left field) he was now worse than replacement level. Meanwhile, George Foster became a star and the Reds won two championships. Anderson could have moved Foster to third base, but he thought Rose could handle it. Given what happened to the Reds, I am forced to conclude that Anderson knew what he was talking about.
So, what am I saying? I am not saying that there should a new statistic to measure flexibility, nor am I suggesting that the WAR values we have become familiar with are wrong, or should be adjusted. I am saying: assessing “value” is complicated.