Monday, February 02, 2009

The Ten Years War

For once, I’m going to attempt to be early to address a topic instead of late. I am usually pretty reactionary; something will happen, everyone in the world will weigh in on it, and then I come in a month later when no one cares anymore (not that my take on it would have been particularly interesting if it were timely). This time, I hope to beat the rush by six or seven months.

Since we are now in 2009, I believe it is safe to assume that we will be treated to a number of pieces on “X of the decade”. The team of the decade, player of the decade, trade of the decade, left-handed sidearming reliever of the decade, etc. This is all well and good, and harmless enough; there’s nothing wrong with frivolous baseball discussions and there’s certainly nothing wrong with putting things in (short) historical perspective.

The key word is “frivolous”. Once the question is allowed to take on greater significance (like being used as a Hall of Fame argument, with Jack Morris as “Pitcher of the Eighties” (*) the prime example), then I have problems with it.

Being the best over a ten-year period is certainly a noteworthy accomplishment. The problem is that the decade approach to issue emphasizes one particular period over all others. Take for instance the fact that Mark Grace led MLB in hits for the 1990s (he had 1,754, seven more than Rafael Palmeiro). In fairness, if anyone has used this to proclaim Grace’s greatness, it has not gained much traction, but still, it sounds a lot more impressive than it actually is. The next six men on the list (Palmeiro, Biggio, Gwynn, Alomar, Griffey, and Ripken) all had more career hits than Grace.

Yes, the list is for hits in the 1990s, not career hits, but this is not a case of getting Grace plus a bunch of guys near the end of their careers or just getting started. With the exception of Ripken and possibly Gwynn, the 1990s were the prime decade for all of them (and Ripken and Gwynn were both productive throughout the decade). It just so happens that for those ten seasons, Grace had a few more hits.

If you look at the ten years from 1989-1998, Grace is third with 1,731 hits, 46 behind Gwynn. The ten years from 1991-2000? Grace (1,715) and Palmeiro (1,719) trade places. So what is it exactly that makes the ten year period from 1990-1999 more important than periods offset by just one year on either side? “The ‘90s” sounds a lot better than “the ten years beginning with ‘89”, but there is no real distinction. By picking out “the ‘90s”, we get the one ten year period in which Mark Grace happened to lead the league in hits. It would be much more impressive is if you could find a player who led the league in hits over multiple ten-year periods. By not examining any other time period, Grace is held up as a singular player, when he should be viewed in a group with other ten-year hit leaders around the same time (for the ten year period beginning with 1988, the leader is Molitor; 1989--Gwynn, 1990--Grace, 1991--Palmeiro, 1992--Alomar).

Organizing history into workable periods can be helpful, but doing it based on endpoints that happen to make a nice numerical cohort (namely, the same digit in the tens place) is a crude way of doing so. If you intend to put any stock in the exercise, why not go through the trouble of dividing time on the basis of important changes in the game? It might be helpful to start a new “era” of baseball history sometime around the middle of the 90s, with events like the strike, the offensive explosion, the influx of Asian talent, the two expansions, and the purported steroids era all as possible markers of a historical shift. I’m not advocating placing too much stock in this sort of classification (for the purposes of establishing statistical leaders, not constructing a historical narrative), either, but if you must break up time into rigid segments, I believe it’s a better way to go.

I’ve been assuming throughout that the “decade” is a natural way to divide time, but who can forget the fascinating (**) arguments that erupted over what constitutes the decade? This issue popped up a lot in 1999 and 2000 as people debated when the third millennium actually began. Since there is no year zero AD (CE if you must, but if you must, please leave), if you actually start counting off ten years to define your decades, the first would be years 1-10, which would eventually lead you to 2001-2010 as the current decade. Furthermore, there’s really nothing to stop you from re-centering the entire calendar of Western Civilization and declaring that the year we call 34 BC should be year 1, and starting from there.

Don’t get me wrong; I have no qualms about calling this the year 2009. The preceding digression is just intended to drive home the point that no specific ten year period is inherently significant solely because of the digits we use to represent it.

In that vein, George Brett’s Hall of Fame plaque includes the tidbit that he was the “first player to win batting titles in three decades (1976, ’80, ’90). There have been objections to this statement on the grounds I described above--a non-negligible portion of observers claim that ’76 and ’80 are in the same decade. Regardless, it’s a fairly impressive feat, but what impresses me about it is that he had batting titles separated by fourteen years, not that the three each fell in a different decade. That’s a long time to be a candidate to lead the league in anything.

However, Ted Williams won batting titles separated by seventeen years (’41 and ’58), and Stan Musial matched Brett at fourteen (’43 and ’57). Brett still has one of the longest gaps between first and last titles, but it’s no longer a historical oddity. And Williams and Musial both won a bunch of them, not just three, the bare minimum needed to have any chance at three decades. The distinction I draw is that Brett’s feat is an ok trivia question, but a really stupid thing to put on someone’s Hall of Fame plaque.

Another issue raised by the decade family of questions is sort of related to the classic career v. peak value issue. Let’s say we were going to pick the player of the ‘00s, as of this moment (I’m writing this in January). Should our pick be the player who contributed the most value over the course of the decade? Should it be the player who was the most valuable in his decade prime, whether he played for the whole decade or not? Should it be the best player, over the course of a full career, for whom this was his most productive decade, even if somebody was more productive in this decade. If you use the first definition, then Jack Morris is a *potential* choice for pitcher of the 80s, and Roger Clemens is probably a poor choice. If you use the second definition, then Clemens or Dwight Gooden is probably your guy (please don’t argue with me on the specific names here; I don’t care and I didn’t check to see which I would pick if I had to). If you use the third definition, you could make an argument along the lines of “I think Hank Aaron was more valuable than Willie Mays in the 1960s, but I also think Mays was a more valuable player over the course of their entire careers, and therefore Mays is the player of the ‘60s.”

If you think it should be the player who added the most value in the decade, period, then what do you do about Albert Pujols (let alone Barry Bonds)? Ignoring fielding (but considering position), I have Alex Rodriguez at 70 WAR and Pujols at 71 WAR for 2000-2008 (those figures are a little higher than most, because I’m using a lower RPW than most people; the specific numbers aren’t really the point). That is too close to call, but consider that Pujols did not play at all in 2000. What if ARod was at 75 WAR and Pujols at 71? That’s still well within a margin of error, but it would make it makes ignoring ARod’s advantage a little tougher. A one win difference is easy to waive off with “He did that in eight seasons rather than nine, but one extra season of 4 WAR is a darn good season.

Another point which should be obvious but that I’ll mention anyway is that the used of a fixed decade period can be a boon to players based on their birthdays. ARod, born in 1975, went from 25 to 34 in the 00s. This is a good ten-year stretch to have considered, as it safely includes his expected peak seasons and avoids seasons in which he would be very young or very old. What about a player like Hanley Ramirez, though? Born in 1983, the 00s are his age 17-26 seasons. And the 10s will be his age 27-36 seasons. Even if Ramirez is the best something (shortstop, maybe even player) in the majors over a legitimate stretch, there is a non-negligible chance that he will never be the “shortstop of the decade” simply because of how his age corresponds to the definition of decade in use.

I chose to highlight ARod/Pujols precisely because it’s so close that it’s easy to justify Pujols if you’d like. But what happens when the difference in playing time is more profound? I previously mentioned Roger Clemens v. Jack Morris for “pitcher of the 80s”. Do Clemens’ brilliant seasons make him better, or do you go with Morris because it’s hard to take a 95-45 pitcher over a 162-119 pitcher? We can all agree that Clemens was a much greater pitcher than Morris, and over each of their best ten-year periods, but if we limit ourselves strictly to just 1980-89, we almost are forced to honor Morris and not Clemens.

The career v. peak issue is a legitimate one, one which is largely a matter of personal preference and on which many reasonable people arrive at very different preferences. Why muddy the waters even more by forcing comparisons into arbitrary timeframes?

My protestations aside, the end of the year figures to yield an explosion of “X of the decade” stories. Hopefully, they will be offered in the spirit of fun, not profundity, and everything will be kept in perspective. The good news is that it’s just baseball, not a war, and so the worst thing that can happen is that we get treated to a number of ridiculous and overly serious choices from the media and groups of fans. Like the Sporting News choosing Pete Rose as “Player of the ‘70s” over two superior teammates. Or the players selecting Ken Griffey, not Barry Bonds, as “Player of the ‘90s”. And who could forget the “All-Century Team”, in which a group of experts had to be impaneled to avoid leaving Lefty Grove, Christy Mathewson, Stan Musial, Warren Spahn, and Hans Wagner off the team in favor of Sandy Koufax, Nolan Ryan, Ken Griffey, Bob Gibson, and Ernie Banks? I’ll embrace these opportunities to chuckle, even as I shake my head at the thought-process that generates them.

(*) Tango wrote a piece for the Hardball Times about Jack Morris and the “Pitcher of the Eighties”, which addresses the arbitrary definition of a decade and the silly emphasis placed on it.

(**) Fascinating in the sense of arguing over the pronunciation of "tomato".

The National Pastime Almanac is a very handy tool for answering questions like “Who had the most hits between years X and Y”, and it was the source for the data in this piece. If you are not a database whiz and like to fiddle around with the numbers, I highly recommend it. And it’s free.

2 comments:

  1. (It's Honus Wagner, not Hans Wagner.)

    ReplyDelete
  2. Is that in the same sense that it’s Grover Cleveland Alexander, not Pete? Or Henry Aaron, not Hank? Or Charles Radbourn, not Hoss Or Old Hoss Radbourn, not Hoss?

    Don’t make me post a link to a whole bunch of Baseball Magazine articles from the 1910s that refer to him as “Hans”.

    I’m not claiming that Hans was more used than Honus or anything, but the man was certainly referred to as Hans in addition to Honus.

    ReplyDelete

I reserve the right to reject any comment for any reason.