George Brett, Willie McCovey and Eddie Murray were three of the game’s all-time greats. Bill James ranked Brett as the 30th greatest player ever, followed by Murray as the 61st and McCovey in the 68th spot. James awards Murray 437 career win shares, followed by Brett at 432 and McCovey at 408. James’ ranking makes sense and recognizes Brett’s additional value as a third baseman, rather than a first baseman like the other two. If, however, we just focus on offensive production, the debate raises some interesting questions not only about peak versus career figures, but about how valuable different kinds of peaks are.

Eddie Murray’s consistent production meant that for a period of about 14 years, 1977-1990, he could be counted upon to produce at a level that was somewhere between all-star and middling MVP candidate, but he never put together a truly outstanding season. Brett had higher peaks than Murray, but they were unpredictable. His three best seasons were 1985, 1980 and 1976. McCovey had an awesome peak from 1966-1970, but outside of those years, and one or two others, he was good, but rarely even performed at a solid all-star level. George Brett sprinkled about four dominant seasons over a 20 year career that was otherwise made up of seasons that were merely good to excellent.

Eddie Murray was an almost freakishly consistent player. He had 25 win shares or more, if we adjust for the 1981 strike, every year between 1978-1985. Brett, again adjusting for the 1981 strike, never managed to put four 25 win share seasons together in a row. His best was three in a row from 1975-1977. McCovey’s best in this category was also a three year stretch, from 1968-1970.

Murray had a similar level of consistency with adjusted OPS+, having an OPS+of 130 or better every year between 1979-1986. Brett’s longest run in this category was 1979-1983. McCovey, however, managed a seven year run from 1965-1971. Murray had the longest run in this category despite having the lowest career OPS+ of the three players.

Murray’s impressive stretches of production were not punctuated by any truly outstanding years. Based on win shares, the three best seasons enjoyed by any of these players were McCovey’s 1969 seasons (39 win shares) followed by Brett’s 1985 season (37 win shares) and Brett’s 1980 season (36 win shares). Murray’s best season, 1984 (33 win shares) ranks behind two of Brett’s seasons and three of McCovey’s. Brett’s four best seasons are responsible for 32% of his career win shares and McCovey’s for 34%. For Murray, the proportion is only 28%.

McCovey and Brett were clearly better peak performers than Murray, but their peak’s were different. Brett’s four best years were 1976, 1979, 1980 and 1985, with 1976-1980 his best five year run. During this five year run Brett accrued 118 win shares, good for 27% of his career total. His OPS+ during this period ranged from 123 to 203. This is an impressive peak, but Brett’s best year was 1985 several years after this peak. Further, Brett had nine seasons in his career where he topped 25 win shares, but they were spread over a 16 year period. In six of the seasons in which he had 25 win shares, he fell below this number the following year.

McCovey’s peak was much more compressed than Brett’s. All of his four best seasons were in a five year period from 1966-1970. During this period he accrued 164 win shares, good for an extraordinary 40% of his career total. His worst OPS during this period was 159. It was really a six year peak because McCovey also had a very good year in 1965 with 29 win shares, so this six year period accounted for 47% of his career win shares. McCovey hit 226 home runs during this period, which given that most of these years were in the pitching dominated late 1960s while McCovey was playing half his games at old Candlestick Park, is barely short of amazing. Murray’s best five year run, 1981-1985, adjusting for the strike in 1981, by contrast only produced 152 win shares or 34% of his career total.

Another way to see McCovey’s peak is that outside of a six year run in the middle of his career, McCovey rarely played enough or hit well enough to be a truly great player over the course of any given season. In the remaining 16 years of his career, actually 15 because his first and last years together combine to equal one full season, McCovey averaged 20 home runs and only 14 win shares. Only twice in his career, outside of this six year run, did McCovey get 25 win shares in a season. In only one other season did he hit more than 30 home runs. A caveat to this brief analysis is that McCovey’s early career numbers were depressed because he did not really become a full time player until 1963, his fifth year with the Giants.

McCovey and Brett were both great peak players, but their peaks were distributed differently. Brett’s three best years were scattered over ten seasons, McCovey’s over four. It is not difficult to find other players, who while not as good as these three, have had even more dissimilar distributions of peak seasons. The practical question this raises is which is more valuable. The question, while somewhat abstract, because career paths cannot be predicted with great certainty, is still important when thinking about building teams, particularly when making decisions about free agents. Obviously, answering this question in a useful way would require a much bigger study, but recognizing the different types of peaks is a useful start.

Lincoln Mitchell