Are veterans being replaced more often?

There’s a bunch of talk going around that MLB is in a youth movement, that older players are being brushed aside. Collusion, a desire to gain distance from the PED era, and quality of/increased trust in prospects are all theories to explain this youth movement with generic doses of plausibility. Of course, without context, it’s rather difficult to prove any of those theories. This article is an attempt to bring substantive context to the claim that veterans have been pushed aside.

As is the case with most statistical endeavors, I don’t pretend that this provides full answers, but I hope that it will make enough sense that either it satisfies your curiosities or lets you dig deeper. My tentative conclusion is that baseball has never changed its stance on veteran retention without something either external or premeditated, and that the lack of an obvious external motivation means we’re more likely seeing a concerted effort to go young. But don’t take my word for it…

I started by classifying the following players as veterans:

{exp:list_maker}All position players with 600+ games;
All pitchers with 300+ games; and
All starting pitchers (defined as at least 50 percent of appearances being starts) with 150+ games. {/exp:list_maker}

Coincidentally, as of June 8, 2009 precisely 3900 players met those criteria, stretching back to 1876. Now you know what Jeff Francoeur is for; by my count, he’s number 3900. (Joe Mauer will be 3901 by the end of the week.) For those 3900, I looked at the years they debuted in the majors and the years they last played to see if any of these debut or end years clump, so to speak. This means that players from 2005 are just now getting into the sample; Robinson Cano and Francoeur are the only two so far.

Now that we have our subset of relevant players, let’s go to a rarely-used feature (for me, anyway)…charts! (As a wise television program once told me: when in doubt, chart it out.) All three of these charts run from 1901 to 2000, since the 19th century had way too many ups and downs to be included and the 21st century is still adding players.

First, a look at the percentage of players used who were veterans in each season. This should give a decent read as to how many players of reasonable ability and service were in the majors at a given time.

image

So, for example, in 2000 teams used 749 players from our sample of veterans; that’s 24.97 per team, or the blue line. Teams used a total of 1230 players that season, or exactly 41 per team (or the maroon line). Veterans comprised 60.89 percent of all players playing in 2000, which was the lowest such percentage since 1970. The peaks and valleys of the chart are by and large easily explained: There’s the Federal League bump in the mid-teens, with the subsequent contraction leaving an unusual amount of veterans available to fill bench and fringe slots for the next few years (until they petered out; hence the downward slide afterward); there’s the subset swell of wartime (since any player who returned after the war counts in those intervening years), the which led to 1946, a somewhat crazy year where the returning vets and their substitutes had to get sorted out on the fly; that year’s 39.56 players per team would not be eclipsed until 1990.

Most importantly, there’s the series of peaks and valleys that are related to expansion. When expansion happens, young players suddenly have more employers, so they get more opportunities to stick around for awhile, but the youngsters come up behind them who may or may not last long, creating a sudden drop of veterans that doesn’t stabilize for a few years. The strike years also show up easily. Most of the highest percentages of veteran retention rate come in years right before expansion: three of the 10 highest percentages belong to 1968, 1976, and 1992 (1968 featured 69 percent of players who either had made the subset or would later; this is the modern record.)

image

image

These two charts are largely identical; the second chart incorporates the data of the first into an overall picture of player turnover. The black line is the combined percentages of the blue and green lines that measure veteran players debuting and exiting, respectively. The first chart measures this by number of players coming and going, while the second measures turnover as a percentage of total players. For example, 36 veterans debuted in 2000, while 40 left after the 1999 offseason; this makes for a 10.15 percent turnover rate, achieved with 4.81 percent new players and 5.34 percent old ones. As you can see, that sudden drop in 2000 is rare but consistent with other post-expansion years, such as the valleys of 1966 and 1979.

For the black, blue, and green lines, the peaks that are not Federal League, wartime, or expansion-related are the ones we’re looking for: the ones that indicate an unusual occurrence unrelated to circumstances. What surprised me was how few of them there are. There’s the Depression, where veterans theoretically would have been more willing to take pay cuts just to stick around, lowering the exit rate; there’s 1995 with its record amount of veteran debuts (83 players) in large part because there were no September callups the previous year … and there’s the peak turnover of 1985-87 caused primarily by collusion.

The 1985-1986 offseason saw a record number of veterans (56) not come back. The first half of the decade had seen totals in the 40s; historically high, but with 26 teams that was normal. In 1986 a record number of players who later become veterans (64) debuted. The first half of the decade had seen totals in the 40s and 50s; this was not close to a record, as 1974 had set the high with 60. 1986-87′s offseason saw another 56 leave, only to go back down into the 40s after that. The 1993 season, due to expansion, broke the influx record with 65; the 1995 season for reasons already mentioned shattered that with 83. Without the strike, 1993 would likely remain the record, as no other year has cleared 63 debuts.

The exits, on the other hand, are a bit troublesome. In 2001, the amount of leaving veterans hit 74, leaving the collusory winters in the dust. This is in part expansion-related, but it still is a high number. After 45, 55, and 53 exits in the next three years, the total jumped up to 68 in 2005, dipped to 63 in 2006, and currently stands at 75 for 2007 and 86 for 2008. The ’07-’08 figures are slightly misleading, however; Kelvim Escobar was on 2007′s exit list until very recently, and several injured parties from 2008 have yet to make their 2009 appearances.

Based on the players who are currently veterans, my estimate is that 2007 and 2008 will both wind up in the low 70s, a substantial uptick from 2005 and 2006 which were themselves banner years for exits. There’s been no expansion for a decade, no wars, and no competing leagues. What gives? It’s possible that, as we only have three periods in history where there’s been at least continuous decade of no expansion (1916-60, 1977-92, and 1999-present), that we’re still sorting out from expansion but have too small a sample size to know for sure. 1926 had an unusually high rate of exit, as did 1986-87; perhaps collusion and expansion hangover worked together in the latter instance.

It’s possible that the awful rookie class of 2000, when only 36 current veterans debuted (lowest non-strike total since 1979, and that was with four fewer teams), let some veterans stay around that would not have otherwise and they’re only now getting the axe. Clearly, due to the high exits of the 2005-2007 offseasons, the current economic state is not the answer, especially when teams seem perfectly content to increase the amount of players they use in a given year (which adds expense in several ways).

If these external forces are not the answer, then the two internal changes seem the most likely explanations: teams are increasingly trusting youth/distrusting veterans, or collusion on some scale. I lean toward the latter for now, for no other reason than that the former apparently has never happened in the history of baseball while the latter has. You go where the precedent is absent anything else. Does this mean it’s Ueberrothian in scope? No; it doesn’t have to be to serve the purpose.

The dead giveaway of the mid-80s was that household names in their prime were getting turned down by an amazingly high number of teams. No one with a straight face could say that a 27-year-old Tim Raines couldn’t help their team, and by the time the lawsuits were settled it was clear there was nothing straight about it. Collusion doesn’t have to be that bald-faced or extend to that caliber of player to keep salaries down across the board. In this era, you can get a lot done by worrying excessively about older players, or—and this is where I suspect most of the problem is—by refusing to sign middle-tier free agents. The 2006 collective bargaining agreement and its reclassification of free agents may fit somewhere in here, but its first offseason was 2007, and we’d still have to explain 2006.

Still, as esoteric as Elias’s rankings can be, it’s not hard to argue that the guys just outside Type B may by having no compensation attached to them be worth more than the Type B guys. You can offer the no compensation guy a little bit more than he thinks he’s worth but less than the Type B free agent thinks he’s worth and still come out ahead. It’s like upping your offer on a house but having them pay closing; it’s the same deal for you but it feels better to the seller. To use a different realm of analogy, just as the bulk of chess is not won by the queen and rooks but your knights, bishops, and occasionally pawns holding the center, so the center of free agency is a better indicator of the market than the top, and if the owners can hold that by casting off the occasional medium-leverage veteran and going with youth, then they’ve won something. It’s much easier PR work for your team to go cold on Juan Uribe and Gregg Zaun than CC Sabathia and Mark Teixeira, and the first are far more fungible than the second. It’s not collusion in its strictest form, but it still can have a substantial effect on who’s playing and for how much.

So, to answer the question of the title, it does appear that more veterans and a higher percentage of them are leaving the game than before, and aside from ’80s collusion the reasons aren’t explained by external factors. Make of it what you will, but hopefully the charts above will help get to the bottom of it.

References & Resources
B-ref, and a few listings of free agents and compensation explanations to get my facts straight.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Abolish the minor leagues
Next: Top 50 Fantasy Prospects in the 2009 MLB Draft with Comments »

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *