few howls of outrage arose when the cardinals signed adam kennedy last winter. few shouts of joy went out either; it wasn't that type of a signing. kennedy seemed like a sensible, low-risk acquisition who'd stabilize a position that had lacked a regular tenant for most of four years --- since the middle of 2003, when fernando vina went out with an injury. since then hart, womack, grud'k, luna, miles, and belliard had taken turns at the position, and junior smiley had played himself out of a crack at it; kennedy was expected to be as good as the best of that lot and far superior to the worst.
hah.
the other day, when i compared scott rolen's actual performance to his expected performance, i estimated the deficit at about 30 runs, or 3 wins. kennedy's corresponding deficit this year was nearly as large --- 21 runs, or 2 wins (and that's adjusted for playing time). but the kennedy slump seems to have come from out of nowhere; he's never had a season remotely like this one, whereas rolen in 2007 essentially returned to his 2005 level of production. while we didn't expect scotty to wash out, we at least knew there was a certain risk of his shoulder acting up again. kennedy, by contrast, has been a model of consistency throughout his career. it was thought that, at 31, he might be susceptible to a slight downturn, but not a soul predicted that he'd be one of the worst position players in the majors.
out of curiosity, i checked out the list of comparables at kennedy's baseball-reference.com page, and out popped a familiar name: tony pena, the catcher the cardinals traded for right before the 1987 season. paid a pretty penny for him, too --- two of the guys they gave up, andy van slyke and mike lavailliere, were regulars on the pirates' division winners of 1990-1992. but pena was the best player at his position in the league --- a perennial all-star who provided gold-glove defense and an above-league-average bat. he was 30 years old when he came to st louis, a year younger than kennedy; compare the two seasons (pena's numbers are pro-rated to equal kennedy's 279 at-bats; he actually batted 384 times):
| AB | R | H | 2B | 3B | HR | RBI | BB | | | AVG | OBP | SLG | OPS+ | |||
| kennedy 2007 | 279 | 27 | 61 | 9 | 1 | 3 | 18 | 22 | | | .219 | .282 | .290 | 50 | ||
| pena 1987 | 279 | 30 | 60 | 9 | 3 | 4 | 33 | 27 | | | .214 | .281 | .307 | 55 |
the 1987 cardinals won the pennant despite pena's performance, so it doesn't particularly irritate. the following year pena bounced back partway, posting a .263 / .308 / .372 line (94 ops+, above average for a catcher), and he stayed at about the same level for 1989, then went off to the american league for the last 8 years of his career.
interestingly enough, a couple of the other high-ranked comparables on kennedy's list also crashed in their age 31 season. remember now, the comps aren't updated until the end of the season, so this list reflects the guys most similar to kennedy through age 30 --- they were already similar before this year's slump. the #1 guy on the list, johnny logan, played shortstop for the milwaukee braves in the 1950s; his number 2 comp is julio lugo, who signed a big-bucks contract last winter and, dollar for dollar, has stunk at least as badly as kennedy this season. comparisons, again pro-rated for 279 at-bats:
| AB | R | H | 2B | 3B | HR | RBI | BB | | | AVG | OBP | SLG | OPS+ | |||
| kennedy 2007 | 279 | 27 | 61 | 9 | 1 | 3 | 18 | 22 | | | .219 | .282 | .290 | 50 | ||
| logan 1958 | 279 | 27 | 63 | 11 | 0 | 5 | 27 | 21 | | | .226 | .286 | .326 | 68 | ||
| lugo 2007 | 279 | 35 | 66 | 18 | 1 | 4 | 37 | 23 | | | .239 | .296 | .355 | 70 |
logan bounced right back at age 32 with one of his best seasons ever --- .291 / .369 / .411 (115 ops+); it remains to be seen what trajectory lugo will follow. now warming to the inquiry, i decided to search the lahmann database for seasons that were similar to kennedy's, to see out how those players fared the following year. i set the following parameters:
average: .225 or less
slugging: .340 or less
ray knight, 1984: .237 / .279 / .299 in 371 at-bats
like kennedy, knight was 31 years old when he had his awful year. he was better than kennedy in his 20s (two all-star appearances and a top-5 mvp finish), and there was a sizeable ballpark factor involved in the collapse of knight's bat --- the previous year he'd moved to the astrodome, one of the most hitter-unfriendly ballparks ever. in the midst of his 1984 slump knight got traded to the mets, for whom he played even worse the following year (.218 / .252 / .328). but knight did have one good season left in him --- 1986, when he approached his career highs in batting, obp, homers, and rbis. knight was the guy who scored the winning run when the ball rolled through billy buckner's legs in game 6 of the '86 series; in game 7 he hit a solo homer to break a 7th-inning tie and propel the mets to the world title.
tony bernazard, 1984: .221 / .290 / .287 in 439 at-bats
bernazard was 27 years old and in his 4th full year as a big-leaguer; the previous 3 seasons he'd been an above-average hitter (tony la russa himself made bernazard a regular) with a good glove at 2b, ie a very valuable player. he changed teams / ballparks in 1984 and posted an ops+ of 60, but turned in the two peak years of his career in 1985-86. he was significantly younger than kennedy, but otherwise a reasonably similar case.
vance law, 1986: .225 / .298 / .325 in 360 at-bats
interestingly enough, he was tony bernazard's teammate under la russa on the early-1980s white sox. law was with montreal by 1986, 29 years old and coming off a career year (123 ops+). primarily a 3d baseman for la russa, he shifted to 2d at montreal. after his 1986 slump (ops+ of 73), he returned to above-average levels of production in his age-30 and -31 seasons.
vinny castilla 2000: .221 / .254 / .308 in 331 at-bats
castilla is not very comparable to kennedy; entirely different type of player, one whose success was highly ballpark-dependent. but he slumped at a similar age to kennedy (32) and, like kennedy, after switching teams and leagues --- it was vinny's first year with the devil rays. his ops+ in 2000 was just 42, but he returned to league-average status the following year after being traded back to the nl. at age 34 he faltered badly again, but then settled in as a useful player for another 3 years.
in all of these cases (and in the pena and logan cases above), the player's dreadful year did not signal an irreversible loss of skill; all the guys in question came to back to have good seasons. i suspect that the same will be true for adam kennedy. he prob'y was playing on a bad knee all year; he also had switched leagues and didn't know the pitchers. while the prospect of opening next year with kennedy at 2b seems rather depressing at the moment, he has a reasonable chance of surprising us in a pleasant way in 2008.