Notice: Trying to get property 'display_name' of non-object in /var/www/html/wp-content/plugins/wordpress-seo/src/generators/schema/article.php on line 52
keyboard_arrow_uptop

“Men do not quit playing because they grow old; they grow old because they quit playing.”

Oliver Wendell Holmes

Before we get to the topic at hand, I want to offer an additional insight on last week’s column, related to the influx of players in the early twentieth century.

In that column, I observed that the average weight of players who debuted during this time (roughly 1910-1919) dropped fairly sharply, and that the average age of those players also dropped. From this I concluded that if it were the case that weight were recorded early in each player’s career, the drop in mean weight for new players could be explained by younger-and therefore, on average, lighter-players entering the population pool. That would be relative to players who had debuted at other times, who were older and therefore generally heavier.

Reader Guy Molyneux challenged that line of thinking by drawing my attention to two important points. First, as shown in one of the graphs in that piece, mean age at debut fell fairly precipitously in the late 1950s and 1960s as well, yet the graph showing average weight at debut does not indicate the same kind of downward trend as seen in the early part of the century. Second, he noted (as I also mentioned) that average weight fell during World War II despite the fact that the average age of players who debuted during that time increased greatly.

Guy then offered what I now feel is probably a major reason (and simpler with a nod to Occam’s razor) for the decline in weight during the 1910-1919 period. Simply put, as the number of players used drastically expanded, teams were forced (in general) to add inferior players, and those inferior players tended to be smaller. (Nate Silver did a good study on the interrrelationship on size and performance that’s worth checking out.)

In regards to the first point above, in looking at the data more closely it is apparent that weight flat lines from 1961-1965 and then spikes in 1966 before falling back to the line in 1967. In other words, while the decreasing age of rookies may have kept the average weight from increasing (in other words, decreasing age may still have been a contributing factor), it did not cause it to fall, indicating that some other factor is probably at work in the 1910-1919 period. Secondly and more convincingly, Molyneux’s theory fits very nicely with the trend we see in 1943-1946, where weight decreased despite a skyrocketing average age. If World War II meant anything, it meant an influx of inferior players, and those inferior players were smaller despite being older.

Now back to our regularly scheduled topic…

In his 2005 book, The Singularity is Near: When Humans Transcend Biology, inventor and futurist Ray Kurzweil predicts that big changes are ahead for the human race in the next 20 to 30 years. In short, “The Singularity” is the point at which the three already exponentially growing and overlapping revolutions of genetics, nanotechnology, and robotics converge to usher in a fusion of biology and technology that will ultimately result in saturating the universe with a combination of our biological and non-biological intelligence. In essence, the law of accelerating returns will create “a rupture in the fabric of human history.”

Along the way “minor” problems such as disease, pollution, and the topic of this column, aging, will be overcome by the applications of that convergence. For example, Kurzweil notes seven key aging processes ranging from mitochondrial mutations to cell loss and atrophy that he expects will be stopped or reversed through techniques such as therapeutic cloning and the introduction of nanobots that perform a kind of maintenance into the bloodstream. Using some of these techniques “radical life extension has already been achieved in simpler animals,” and Kurzweil expects that once successfully applied to higher animals like mice, there will be no stopping their application in humans twenty to twenty-five years in the future.

Interestingly, until the technology is available, Kurzweil takes 250 supplements a day, a half-dozen intravenous therapies a week, drinks eight to ten glasses of alkaline water and ten cups of green tea per day, a regimen which he reports has lowered his biological age to forty. Not bad for a fifty-six year-old.

In the face of those kinds of advances, I’m afraid Nate Silver is going to have to recode his PECOTA spreadsheets if he wants to remain “deadlier than ninja accurate.” But I’ll leave Nate and Will Carroll (since nanobots coursing through the body will surely be able to repair shoulder and elbow injuries on the fly as well) to think through the implications of those kinds of advances on baseball.

Last week’s peripheral discussion on aging and Clay Davenport’s fascinating report of organizational ages did, however, get me to thinking about how the player population has aged over time and what that means especially for teams. So this week, while it’s still relevant anyway, we’ll explore aging in a little more depth.

Tipping the Scales

First though, we need a consistent and relative way to measure the age of teams. To that end I calculated two measures.

  • Weighted Age (WA): This measure is calculated by first creating separate ages for position players and pitchers weighted by plate appearances and innings pitched respectively. The fractional age of players is calculated as of July 1st of the year in question. Then, the weighted age of position players is multiplied by .6 and that of pitchers multiplied by .4, and then summed to calculate the WA for the entire team. Obviously there are other ways to weight playing time (by games played to weight relief pitchers and pinch hitters more heavily, or innings in the field) and there may be quibbles with the weighting (keep in mind the 60% also takes defensive playing time into account), but overall this approach provides a reasonable indication of team age.
  • Normalized Weighted Age (NWA): As you can imagine, the average weighted age for teams has changed over the course of time, and differs by league, so we can normalize the WA by dividing it by the mean weighted age for the year and league. To illustrate why this is necessary, consider the graph below, that tracks average mean WA for teams in the American and National Leagues. While absolute team age has generally increased since the early 1970s (with a large increase with the advent of free agency as teams invested in older players for the longer term), team age was also high from the middle of the 1920s until the 1950s. The graph also indicates that the AL went younger in the 1910-1919 period, and that the NL was younger in the 1950s and 1980s, although more recently the AL has literally been the junior circuit. By normalizing for the average age of teams in the league, we’ll later be able to examine how being relatively old or young correlates with winning on the field.

image 1

Using WA and NWA, we can now determine the oldest and youngest teams in history based on these measures. First, the absolute oldest and youngest teams since 1901 using Weighted Age are shown in the table below, where WA is the Weighted Age of the Team, NWA is the Normalized Weighted Age, an LgWA is the average WA for teams in that league.

Oldest and Youngest Teams Since 1901 Using Weighted Age
Year    Team    Lg  HitterAge PitcherAge     WA     NWA    LgWA
2005    NYA     AL      32.81     34.68   33.55   1.137   29.51
2004    NYA     AL      32.76     33.49   33.06   1.119   29.55
2005    BOS     AL      31.98     34.19   32.86   1.114   29.51
1982    CAL     AL      32.94     32.39   32.72   1.123   29.15
2002    ARI     NL      32.38     33.01   32.63   1.102   29.60
1998    BAL     AL      33.88     30.71   32.61   1.096   29.76
1983    CAL     AL      32.01     33.37   32.55   1.113   29.26
2001    ARI     NL      32.82     32.11   32.53   1.102   29.53
2006    SFN     NL      34.54     28.98   32.32   1.094   29.53
2000    NYA     AL      31.85     32.81   32.23   1.092   29.51
1988    DET     AL      32.61     31.61   32.21   1.110   29.02
2003    NYA     AL      30.82     34.23   32.18   1.104   29.14
1945    WAS     AL      31.77     32.76   32.17   1.046   30.76
1983    PHI     NL      32.66     31.24   32.09   1.107   28.98
2006    NYA     AL      31.42     32.97   32.04   1.093   29.32
1945    CHA     AL      32.39     31.51   32.03   1.041   30.76
1984    CAL     AL      32.36     31.42   31.98   1.097   29.16
1982    PHI     NL      31.58     32.46   31.93   1.097   29.10
1999    BAL     AL      33.18     29.81   31.83   1.083   29.40
2005    SFN     NL      32.89     30.24   31.83   1.071   29.73
---------------------------------------------------------------
1967    KC      AL      25.28     23.84   24.70   0.897   27.53
1915    CLE     AL      25.03     24.45   24.80   0.942   26.32
1968    OAK     AL      25.19     24.30   24.83   0.901   27.57
1915    PHA     AL      26.99     22.38   25.15   0.955   26.32
1921    PHA     AL      25.58     24.98   25.34   0.894   28.36
1901    CLE     AL      28.33     20.96   25.38   0.921   27.56
1966    KC      AL      26.03     24.48   25.41   0.923   27.54
1916    PHA     AL      26.65     23.62   25.44   0.944   26.95
1920    PHA     AL      25.33     25.62   25.45   0.901   28.24
1911    BOS     AL      25.27     25.72   25.45   0.928   27.42
1914    BOS     AL      25.84     25.14   25.56   0.974   26.24
1998    FLO     NL      26.01     24.89   25.56   0.885   28.88
1972    SDN     NL      25.43     25.81   25.58   0.926   27.64
1917    PHA     AL      26.00     24.99   25.60   0.942   27.16
1973    SDN     NL      25.57     25.73   25.63   0.922   27.80
1914    WAS     AL      27.38     23.02   25.64   0.977   26.24
1974    SFN     NL      25.67     25.66   25.67   0.928   27.66
1982    MIN     AL      25.73     25.62   25.69   0.881   29.15
2006    FLO     NL      25.82     25.62   25.74   0.872   29.53
1910    BOS     AL      25.92     25.53   25.76   0.932   27.65

Five of the last seven Yankee teams make the top 20, with the 2005 version taking the top spot. Some other notably old teams include the California Angels led by the aged trio of Tommy John, Rod Carew, and Reggie Jackson, and the Philadelphia Phillies (Pete Rose, Tony Perez, Joe Morgan) of the early 1980s. The 1998 Orioles (Jesse Orosco, Harold Baines, Joe Carter), the 2001-2002 Diamondbacks (Mike Morgan, Randy Johnson, Mark Grace), the 1988 Tigers (Darrell Evans, Doyle Alexander, Fred Lynn), and the Giants of the past two seasons also are prominent. (Ed. Note: The Giants’ collective age is a major subject in the team essay in this year’s edition of the annual.)

You’ll also notice that in addition to a strong representation of recent teams, two teams from 1945 make the list, since both periods feature historically high average team weighted ages.

On the other side of the spectrum, the 1967 A’s (Rick Monday, Catfish Hunter, Bert Campaneris) take the distinction as the youngest team since 1900, with the franchise’s 1966 and 1968 squads also making the list. It should also come as no surprise that the post fire-sale Marlins of 1998 (Mark Kotsay, Brian Meadows, Derrek Lee) and 2006 make the list, as do five versions of Connie Mack‘s Philadelphia A’s. After Mack released or traded most of the stars from his 1914 pennant winner, the A’s teams on the list (1915, 1921, 1916, 1920, 1917) never won more than 55 games, and it wouldn’t be until 1925, with an older team (WA of 28.06), that they would again break .500.

In order to account for the age of the league let’s next take a peek at the oldest and youngest teams using Normalized Weighted Age.

Oldest and Youngest Teams Since 1901 Using Normalized Weighted Age
Year    Team    Lg  HitterAge PitcherAge     WA     NWA    LgWA
2005    NYA     AL      32.81     34.68   33.55   1.137   29.51
1982    CAL     AL      32.94     32.39   32.72   1.123   29.15
2004    NYA     AL      32.76     33.49   33.06   1.119   29.55
1973    DET     AL      32.26     30.78   31.67   1.116   28.38
2005    BOS     AL      31.98     34.19   32.86   1.114   29.51
1983    CAL     AL      32.01     33.37   32.55   1.113   29.26
1988    DET     AL      32.61     31.61   32.21   1.110   29.02
1983    PHI     NL      32.66     31.24   32.09   1.107   28.98
1979    NYA     AL      30.85     31.71   31.20   1.107   28.18
2003    NYA     AL      30.82     34.23   32.18   1.104   29.14
1960    CHA     AL      31.16     32.51   31.70   1.103   28.75
2002    ARI     NL      32.38     33.01   32.63   1.102   29.60
2001    ARI     NL      32.82     32.11   32.53   1.102   29.53
1905    BOS     AL      31.94     30.77   31.47   1.099   28.64
1998    SFN     NL      31.23     32.42   31.71   1.098   28.88
1982    PHI     NL      31.58     32.46   31.93   1.097   29.10
1984    CAL     AL      32.36     31.42   31.98   1.097   29.16
1988    NYA     AL      31.10     32.91   31.82   1.096   29.02
1998    BAL     AL      33.88     30.71   32.61   1.096   29.76
2006    SFN     NL      34.54     28.98   32.32   1.094   29.53
---------------------------------------------------------------
2006    FLO     NL      25.82     25.62   25.74   0.872   29.53
1982    MIN     AL      25.73     25.62   25.69   0.881   29.15
1998    FLO     NL      26.01     24.89   25.56   0.885   28.88
1999    FLO     NL      25.71     26.59   26.06   0.893   29.19
1921    PHA     AL      25.58     24.98   25.34   0.894   28.36
1999    MON     NL      26.18     26.06   26.13   0.895   29.19
1967    KC      AL      25.28     23.84   24.70   0.897   27.53
1968    OAK     AL      25.19     24.30   24.83   0.901   27.57
1920    PHA     AL      25.33     25.62   25.45   0.901   28.24
2000    MON     NL      26.55     26.51   26.53   0.902   29.41
1999    CHA     AL      26.28     26.96   26.55   0.903   29.40
1998    MON     NL      26.15     26.12   26.14   0.905   28.88
1936    PHA     AL      26.24     26.45   26.33   0.906   29.04
1956    PIT     NL      25.45     27.08   26.10   0.907   28.79
2000    FLO     NL      26.52     26.91   26.68   0.907   29.41
1983    MIN     AL      26.48     26.76   26.59   0.909   29.26
2000    MIN     AL      26.92     26.75   26.85   0.910   29.51
1950    SLA     AL      25.96     27.50   26.58   0.910   29.20
1914    CHF     FL      26.29     25.82   26.10   0.911   28.66
1981    TOR     AL      26.46     26.31   26.40   0.912   28.94

Once again, the Yankees of 2005 come out on top, almost 14% older than the average team in the American League that season, while many of the same teams from the first list find themselves on this one as well. Looking a bit out of place, the 1905 Boston Americans make an appearance with a team that featured 38-year-old Cy Young (who would throw 24% of the team’s innings), 36-year-old outfielder Jesse Burkett, and 35-year-old third baseman Jimmy Collins.

The 2006 Marlins make their way to the top as the relatively youngest team in history, with the 1982 Twins led by Tom Brunansky (21), Kent Hrbek (22), Brad Havens (22), and Gary Gaetti (23) taking the runner-up spot.

Because position hitters’ and pitchers’ ages were calculated separately, we can take a quick look at the youngest and oldest teams in each category.

Oldest and Youngest Hitters by Weighted Age
Year    Team    Lg  HitterAge PitcherAge     WA     NWA    LgWA
2006    SFN     NL      34.54     28.98   32.32   1.094   29.53
1998    BAL     AL      33.88     30.71   32.61   1.096   29.76
1999    BAL     AL      33.18     29.81   31.83   1.083   29.40
1945    DET     AL      33.01     28.53   31.22   1.015   30.76
1985    CAL     AL      32.98     28.51   31.19   1.064   29.30
---------------------------------------------------------------
1915    CLE     AL      25.03     24.45   24.80   0.942   26.32
1975    MON     NL      25.10     27.22   25.95   0.939   27.65
1973    CLE     AL      25.11     29.36   26.81   0.945   28.38
1968    OAK     AL      25.19     24.30   24.83   0.901   27.57
1911    BOS     AL      25.27     25.72   25.45   0.928   27.42

It’s not surprising that the 2006 Giants and their collection of obsenely geriatric position players took the top spot. What is somewhat confusing is that in their stated effort to “get younger and healthier” in 2007, they’ve added a 35-year-old center fielder, a 32-year-old catcher, and a 29-year-old starting pitcher who they signed for seven years to go along with their 42-year-old left fielder and 40-year-old shortstop.

By contrast, the 1915 Indians did not have a starting position player who was 30 years old and were led by 24-year-old shortstop Ray Chapman, while the 1975 Expos featured a pair of 21-year-olds, Gary Carter and Larry Parish.

Oldest and Youngest Pitchers by Weighted Age
Year    Team    Lg  HitterAge PitcherAge     WA     NWA    LgWA
2005    NYA     AL      32.81     34.68   33.55   1.137   29.51
2003    NYA     AL      30.82     34.23   32.18   1.104   29.14
2005    BOS     AL      31.98     34.19   32.86   1.114   29.51
2002    NYA     AL      30.48     33.63   31.74   1.085   29.24
1935    BSN     NL      29.63     33.58   31.21   1.081   28.88
---------------------------------------------------------------
1915    PHA     AL      26.99     22.38   25.15   0.955   26.32
1914    WAS     AL      27.38     23.02   25.64   0.977   26.24
1916    PHA     AL      26.65     23.62   25.44   0.944   26.95
1967    KC      AL      25.28     23.84   24.70   0.897   27.53
1901    NY      NL      27.83     23.95   26.82   0.948   27.72

Three recent Yankee teams make the top five, along with the Red Sox of 2005, while the 1935 Boston Braves, losers of 115 games, sneak into the fifth slot by virtue of 40-year-old Bob Smith‘s 203 1/3 innings pitched and 38-year-old Huck Betts‘ 159 2/3 innings.

The 1915 A’s had the youngest pitching staff as 23-year-old Weldon Wyckoff threw 276 innings, and 20 year-old Rube Bressler threw 178 1/3.

The Wisdom of Age

Having these measures in hand gives us the tools for answering several interesting questions. I’m sure most BP readers are familiar with the normal career trajectory (or, if you prefer by position, the trajectories for different types of players). This can be graphed to illustrate that offensive production increases rapidly from ages 20 through 25, peaks between the ages of 26 and 28, and then declines more gently through the mid-30s. Given that the 26 through 28 age range registers the peak of individual performance (on average), the question is whether or not that’s the case for teams as well. In other words, do teams with WAs between 26 and 28 outperform younger and older teams?

To examine this, we can simply calculate the aggregate winning percentage for all teams of a certain age. The graph below does this on the blue line, while the red bars represent the number of teams that fell into each age range (i.e. the bar representing 26 includes all teams whose WA rounded to 26).

image 1

What is clear from this is that a minority of teams (42%) fall into the 26-28 age range, and those teams in the aggregate do not win half of their games. Teams who are 29 and older are the ones who perform better (which the attentive reader probably noticed from the tables presented above). Interestingly enough, the trend continues with each successive year, even up to the single team, the 2005 Yankees, who round to a collective 34 years old while posting a .586 winning percentage. The correlation coefficient between winning percentage and team age for all 2,136 teams was also calculated at .227, which also illustrates the relationship.

This result is confirmed by a study done by Chip McNamara in 2003 titled “Does Team Age and Success Correlate in Major League Baseball?” Although the methodology was slightly different, McNamara produced a table that shows regular season winning percentage increasing steadily from a team age of 25 through 31, and then slightly decreasing at age 32. Contrary to the popular notion that younger is always better, when it comes to winning games on the field it appears that age trumps youth.

But perhaps these results are skewed by the underlying differences in league age? To correct for that we can produce the same graph, this time using Normalized WA, and the results look very familiar:

image 1

Once again, teams that are relatively older than other teams in their league perform better. The correlation coefficient in this case is slightly higher at .272, and goes up to .373 when considering only teams from 1977-2006.

So, we’re left asking, why is it that old or older teams perform better? I believe there are two primary reasons.

First, as is evident in the aging curve discussed earlier, the slope of the curve at younger ages is steeper than it is for later ages. A team that is heavily populated with young players (in the 21-25 age range for example) will therefore likely have more players who are still developing, and therefore generally don’t perform quite as well as established veterans. In large part this is probably due to the twofold effect of younger players getting acclimated to the league while honing their skills, the greater variability in the performance of younger players both as they adjust, and to some extent reflects the promotions of young players to The Show who simply don’t succeed at the major league level for whatever reason.

Secondly, old teams retain old players in large part because those old players are productive players and help produce winning records on the field. In other words, the teams that continue to employ older players have selected the best older players (perhaps the only older players, as the others are forced into retirement), so there is less space in the universe of teams for those who are old and yet not very good (the 1935 Boston Braves are a stunning exception to the rule). This effect has intensified in the free agent era, and even more so in the recent past, as wealthier teams are able to retain the services of the best veteran players. Once those older and productive players reach the end of their careers the team is often forced to retool to some degree, getting younger and typically worse in the process. The current Yankees being the exception, having remained a relatively old team since 1994 while putting up good records.

An illustration of these trends can be seen in the recent history of the Detroit Tigers:

image 1

The Tigers reached their peak winning percentage in 1984 with a team that had aged steadily since 1980, and was at that time slightly above the mean age. That group continued to age, and while the team performed well through 1988, they went over the cliff in 1989 as veterans like Darrell Evans moved on and Chet Lemon became unproductive. They retooled with the likes of Cecil Fielder and Travis Fryman in the early 1990s, and remained competitive until the decline and retirement of first Lou Whitaker and then Alan Trammell following the 1996 season. They rebuilt again with Bobby Higginson, Jeff Weaver, and Tony Clark before turning wholesale to the youth movement as Dave Dombrowski took the reigns in 2002. Since then the team has gotten older with aggressive free agent signings prior to the 2004 season and progressively better as the youngsters jelled with the veterans in 2006, winning 95 games with a team whos age was slightly higher than average.

Maturing

Is there a lesson here? If anything, these results should simply serve as a reminder that youth (at the team level anyway) is not necessarily all it’s cracked up to be. While young teams can win (the 1928 New York Giants, 1948 Brooklyn Dodgers, and 2001 Twins are three examples), they’re often battling uphill against teams with a substantial veteran presence. Maturing that young talent and supplementing it with veterans at the appropriate time, like the recent Tigers, is what can make for a formidable combination.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now

Dan Fox

 

You need to be logged in to comment. Login or Subscribe