Last week, Eric Chavez signed a 6-year, $66 million contract extension with the Oakland A’s, locking him up through his 32nd birthday. Aaron Gleeman had a fine analysis of the deal on Friday, pointing out that Chavez is already the second-best third baseman in the game today, and he’s just entering his prime.
I don’t disagree with any of that; Chavez is an outstanding player, and it’s easy to see a bright future for him. One thing concerns me, though… Something you might call “Andruw Jones Syndrome.” Jones is a year older than Chavez, and was a wonderful player in his early-20s. He’s still a fine ballplayer, worth 25 Win Shares a year, but Jones’ game has just stopped developing. Every year we wait for him to break out, and every year he hits his customary .270 with 35 homers — good numbers, sure, but not the mega-star production that his early performance suggested.
Eric Chavez may be an even more striking case. He’s one of the most consistent hitters in all of baseball, batting around .280/.350/.515 in each of the last four years. Here are Chavez’s GPAs since 2000:
GPA 2000 .284 2001 .287 2002 .285 2003 .286
As Aaron said on Friday, “Of course, you can take consistency like that from a young player in two ways. One is that he is a good player and consistently good production is a great thing to have. The other way is that, despite being 22 years old in 2000 and 25 years old last year, he hasn’t really gotten any better offensively.”
One way to tackle this question is to find similar players in history. To do that, I looked for players whose Win Share value patterns through age-25 most closely resembled that of Eric Chavez, with increasing weight being given to more recent years. I found 10 third basemen who had reasonably similar value patterns to Chavez entering their age-26 seasons. The 10 comps:
Harlond Clift Eddie Yost Jim Ray Hart Red Smith Denny Lyons Bill Bradley Bill Madlock Robin Ventura George Brett Ken Keltner
Actually, Edgardo Alfonzo and Troy Glaus also had similar value patterns, but since both are still younger than 32, we can’t really use them in this study. Here is how Chavez’s Win Share totals compare to the group averages:
20 21 22 23 24 25 <26 Chavez 2 9 16 26 25 25 103 group 0 8 18 22 26 25 99
And here is what the group did from ages 26-32:
Age WS 26 21 27 22 28 16 29 12 30 14 31 12 32 10
This is a little odd. Most players peak around age-27, but this group peaks at 24 and begins to decline at 26. All 10 of the third basemen in this study had more than 20 Win Shares at both 24 and 25, but all except Brett also had their career years before their age-26 seasons.
The most-similar player to Chavez, according to this study, is Harlond Clift, the old St. Louis Browns' 3B from the 1930s. Few remember Clift now, but at the end of his age-25 season, he sported a career .291/.405/.496 line. Clift had 107 Runs Created Above Average (RCAA) and a .590 Offensive Winning Percentage; Chavez has 87 RCAA and a .594 OWP. Both players had 25 Win Shares in their age-25 season. While Clift remained a productive player from 26-29, both his batting average and home run numbers fell off considerably and he never again topped 25 WS. In his final year, at age 32, Clift batted just .211 with a .307 slugging percentage.
Interestingly, both players had trouble getting selected to the All-Star team. Chavez has yet to make the team, and despite being one of the best third sackers in the game, Clift had just one All-Star appearance, at 24.
From here on out, let's go in chronological order. The earliest player in the study is Denny Lyons, who starred in the American Association and the National League in the 1880s and 90s. The Lyons comparison is admittedly difficult because 1) it was a different sort of game back then, and 2) his games played totals are all over the place. Through age 25, Lyons had 251 RCAA, but he managed just 67 after that and was out of baseball by 32.
Bill Bradley started his career with the Cubs, but jumped to the American League in 1901 and was one of its earliest stars. His Win Shares increased every year up to age 25, when he peaked at 29 WS. Bradley followed that up with two more good years before quickly fading into mediocrity. He batted under .200 at both 31 and 32, at which point his career looked over. Though Bradley resurfaced in the Federal League a few years later, he never regained his hitting stroke.
I'm pretty sure Bill James pointed this out in the New Historical Abstract, but (can you believe this?) I somehow can't find my copy. Anyway, you can look it up for yourself, but Red Smith was the Shannon Stewart of the 1914 Miracle Braves. The Braves purchased Smith from Brooklyn on August 10, and immediately went on a 43-13 run (.768 winning percentage) to take the NL pennant. The 24-year-old Smith hit .314/.401/.449 in that run, capping a career year. On the last day of the season, Smith broke his leg while sliding and missed the World Series. He recovered to top the 20 Win Share plateau every year from ages 24-28, but he collapsed at 29 and was out of the majors before his 30th birthday.
Ken Keltner's name has been immortalized in the famous "Keltner List," a series of questions that help to assess a player's Hall of Fame credentials. Keltner was a fine defender, and aside from the Keltner List, he's probably best-known for his good defense in the game that ended Joe DiMaggio's 56-game hitting streak. With the stick, Keltner was okay, though his only really excellent offensive seasons were at ages 22, 27, and 31. He missed all of his age-28 season because of World War II, and had only two good years after turning 26. One of those years, 1944 (he was 27), was against lesser competition, as most of the "real" major leaguers were fighting the war.
Harlond Clift is next on the list, but we've already discussed him. Next comes Eddie Yost, aka "The Walking Man." Yost was a regular at 20 and cleared 120 walks seven times from ages 23-33. Still, after age 27 he had just one 20-Win Share season, with 27 WS at age 32.
Nine of the 10 players in this study were still good players at age 27. The exception is Jim Ray Hart, who is probably the worst-case scenario for Eric Chavez. Hart was decidedly better than Chavez from 22-25, and at 25, he earned 29 Win Shares. Unlike Chavez, though, Hart was a poor defender. He had an subpar (but still decent) year at 26, but hurt his shoulder at 27 and never regained regular status. After a 1-for-19 start to the 1974 season, Hart was done at just 32 years of age.
Before turning 26, Bill Madlock had a gaudy .337 batting average and a .658 OWP. From 26-32, Madlock batted just .308, though his OWP remained over .600. That's the second-highest OWP of any player in the study (after Brett) from ages 26-32. Still, for Madlock it was a decline; the four-time batting champ was still good, but he never developed as a hitter.
One player in this study - George Brett - got better after turning 26. He is the hope for Eric Chavez. Brett seemed to have peaked at 23-24, and had a down year at 25. He then matched his career year at 26 before setting the league on fire at 27, batting .390 and earning 36 Win Shares in just 117 games. Brett did what Chavez must do - he boosted his average, improved his plate discipline, and increased his power output. He went from "star" to "Hall of Famer," and finished with over 3,000 hits and 300 homers in his career.
The most-recent Chavez comp, Robin Ventura, has a lot of similarities to Eric. Both were (are?) left-handed-hitting third basemen, and both were excellent defenders. Both were the 10th overall pick in the draft (Ventura in 1988, Chavez in '96). From 22-25, Chavez had 92 Win Shares, Ventura 91. Ventura remained a good player into his mid-30s, but he never "kicked his game up a notch," so to speak. He earned 30 Win Shares at ages 24 and 31, but 21 WS was a more typical Ventura season. Ventura, like all but one of the other players in this study, never developed.
Going into this study, I expected to find that the players most-similar to Chavez stepped up their games at 26 and 27, becoming MVP-level players through about age 32. That sort of result would only confirm my suspicion that Billy Beane had made a great move in locking up Chavez at $11 million/year.
Now, however, I'm not quite so sure. Even if Chavez doesn't decline at all, is he worth 11 million dollars a year if he doesn't improve? One Win Share is worth (roughly) the league minimum, and three Win Shares are worth around $1 million. Chavez's 25 Win Shares, then, are worth somewhere between eight and nine million dollars. If we assume that one must overpay to get a player who is among the best at his position, $10M-$11M seems reasonable. But, it is by no means "a great deal" if Eric Chavez continues to churn out "only" 25 Win Shares a year. And, surprisingly, the historical evidence says that a decline is more likely than an improvement.
I don't think a study like the one above is any sort of sure evidence that the Eric Chavez signing wasn't a great deal. He could become George Brett, which would make $11 million annually a bargain price. If, however, Chavez is merely the next Robin Ventura, that $11 million could prove to be an eyesore for the cost-conscious A's.
One last thing: That 6-year, $66 million extension doesn't start until next year. Chavez will make $5.2 million this season. So from this point forward, the A's are paying Chavez $71.2 million over the next seven years. I know it's a minor difference, but when I first read about the deal I assumed (incorrectly) that it would begin this year.
It may be obvious, but in baseball, not all dollars are equal. $11 million to the Oakland A's is a heck of a lot different than $11 million to the Red Sox or the Dodgers (and let's not even get started on the Yankees). Obviously you have to overpay for a star, but when you're on a tight budget already, the financial hit is a harder one. Signing Eric Chavez means locking up a dependable, quality regular for a long time, but it also means that Billy Beane must continue to pull bargains out of his hat.