The Baseball Economist: The Real Game Exposed (14 page)

The teams at the bottom are a sad bunch. The Mets, Mariners, and Dodgers did not get much out of their players beyond what they paid for them, yet they still could not manage to put very good players on the field. The Dodgers did make the playoffs in 2004, but they were not very good in 2003 and 2005. These teams are full of expensive mistakes, and it’s unsurprising that all three of them recently experienced some turnover in front office leadership.

It may be tempting to view this as an evaluation of general managers— the person typically in charge of the franchise’s baseball operations. Unlike players, general managers don’t always play on the same field. The GM must navigate many unique obstacles besides putting talent on the team. Some organizations have intrusive owners that demand immediate solutions to intractable problems, keep relatives on staff, sign extravagantly expensive free agents, or refuse to sign free agents who are worth the price, to mention just a few of the venalities owners often commit. Other GMs may be saddled with the past mistakes of previous inept management. I’m not sure Billy Beane would act much differently than the Yankees general manager, Brian Cashman, if he had to answer to George Steinbrenner. The job has various responsibilities and political battles, so it doesn’t seem fair to pin the entire credit and blame of a club on this one person. Moreover, GMs and their staffs come and go. The best organizations are those that can continue to function well even after a few key people leave. A good organization is built to get the right people who can make the right decisions.

The recent managerial excellence of the Marlins, Indians, and A’s leads me to believe that these teams possess the ingenuity needed to put teams on the field better than the rest of the league. Billy Beane and the Oakland A’s were a great subject for Michael Lewis to make the point that good management is the key ingredient to success. But let’s not forget that the A’s aren’t the only well-managed team out there. The Marlins and Indians don’t have that California flare, but they deserve some credit, if not more, than the A’s. Now here come the pennants!

8

The Evolution of Baseball Talent

—AL BRIDWELL, CIRCA 1965 (MAJOR-LEAGUE CAREER: 1905–1925)
⁴⁷

WHEN WAS the “Golden Era” of baseball? Was it the era of Ruth and Gehrig, DiMaggio and Williams, Mantle and Mays, Aaron and Robinson, Boggs and Schmidt, or how about A-Rod and Bonds? I am biased; I like baseball today the best. For some reason, more people (mostly my elders) feel the best baseball was played prior to my birth in 1973. How can I argue with these people? I didn’t watch much baseball until the 1980s. Was I deprived? Certainly, I never got to watch Babe Ruth swat sixty home runs in a season; however, I also avoided the pain of watching the 1968 season, the worst offensive season in baseball’s modern era. A lot of the bias in favor of the past is plain nostalgia. It is the same thing that causes my parents to tune in to 1950s radio stations and has me listening to music from the 1980s now. I am not sure if I really like eighties music—or if my parents really like the “oldies”—but it seems to attract my attention out of some longing to remember the past. That same feeling is sure to have some effect on our judgment of baseball history.

What if we could identify an objective way to compare baseball eras? To make this comparison, I developed a metric that can measure the quality of competition. Judging the quality of today’s game with this tool takes nostalgia out of the process.

Furthermore, the analysis reveals some important information regarding the power surge in today’s game. Many pundits claim that the modern era is tainted by steroids, which they see as the only explanation for the hitting achievements of a few great sluggers. These players may have used performance-enhancing drugs—only a handful of people know if they did—but it’s not the case that their home run prowess could only come from “the juice.” In fact, their great achievements in home run hitting are exactly what we would expect given the current distribution of talent in the league. The statistics don’t convict them.

It is tempting to judge baseball quality according to absolute statistics, and to admire seasons with record statistics in different hitting and pitching categories. The problem with this analysis is that baseball players do not play games across generations, so such comparisons provide very little information about the quality of the game. Individuals generate these statistics relative to their peers. It was certainly a great feat that Roger Maris hit sixty-one home runs in 1961, but this in no way means that if we used a time machine to bring the 1961 Maris to face the pitchers in today’s game he would perform at the same level. In fact, the idea is laughable. Athletic performance has improved drastically in all sports where we can observe absolute quality. For example, the world record marathon time has fallen by about 8 percent since Maris’s glorious season. I expect that baseball players have improved similarly. Now, this does not say anything bad about Maris—he certainly cannot be held accountable for playing baseball when he did— but it does show the problem of using absolute statistics to judge the quality of a sport based on relative competition. As Figure 3 shows, the average offensive output has fluctuated quite a bit over baseball history. The runs-scored-per-game statistic reflects talent on both sides of the ball, because worse pitching must beget better hitting and vise versa.

If absolute statistics are not an appropriate metric to judge baseball
performance, what metric is appropriate? Well, what makes the game of baseball fun to watch? The drama, of course. Men at the highest end of the baseball talent spectrum bring forth their best in order to defeat one another in accordance with the rules of the game. These rules provide a roller coaster of action, from meetings on the mound to collisions at the plate. We tolerate the lows to see the highs, until one team emerges the victor. It is the drama along the way that makes the game interesting. But can we develop a statistic to measure drama? I think so.

The predictability of the outcome affects the drama of the game. With too much predictability the game is no longer very interesting. Baseball is a game of relative strengths. To win the game you do not have to be the best team in the league, just better than your opponent, plus or minus a little luck. A very good team playing a bad team is more likely to lead to a blowout than a game between closely matched talents. As the teams become more even, the uncertainty of the outcome will grow. But it is possible for the teams to be too even; predictability can result from extreme similarity as well as extreme differences. That “plus or minus a little luck” can put a damper on the enjoyment of the game. Luck will certainly play a role in many games, but I do not want it to be the main factor. As the teams become too equal, the game can become predictably boring as games will turn on wind direction, shadows, and rocks in the infield.

In terms of the wins and losses, the equality of teams is measured by competitive balance—the smaller the difference between the best and worst team, the greater the competitive balance. But the main game also contains a series of sub-contests between players, which can affect the predictability within the game. When Cy Young winners and home run champions face off, the excitement is intense, while having these star players face borderline Triple-A talent is less exciting. But this does not mean that lineups composed entirely of stars are necessarily more exciting than lineups with some non-stars. The game might be equally dull if the best batters only faced the best pitchers. Imagine a world where Barry Bonds faced Roger Clemens at every at-bat. In this world neither Clemens nor Bonds would seem all that special. There would be few automatic outs and few offensive outbursts. There would be strikeouts and home runs, but Clemens would never strike out twenty in a game, and Bonds would never hit seventy-three home runs in a season. In contrast, if Bonds faced pitchers in Single-A ball, the results would be spectacular, so spectacular you would hardly notice. Bonds would mash homers three and four a game. If Clemens faced only Single-A hitters, twenty strikeouts might be a bad game. In these worlds there would be no challenge, no competition, and little excitement. Clearly, the most desirable world is something in between. The best players in the game ought to be able to play well, but not to consistently dominate. Falling too far from this middle ground can make the game less interesting.

Evolutionary biologist Stephen Jay Gould first proposed the notion that altering the distribution of talent across the league can affect the level of competition between players.
⁴⁸
He was attempting to explain why no player has hit .400 since Ted Williams in 1941. As the overall quality of players improves, extreme achievements should decline. Adding and subtracting players affects the distribution of talent in the league by including or excluding marginal major-league players. Over the years baseball has altered its talent pool on many occasions—mostly
increasing the size—but this seems to have had little effect on the average quality of play because the outcomes derive from
relative
competition. Theoretically, when MLB expands its size the league becomes worse in absolute terms, because those who were previously not deemed worthy to play in MLB become members of new MLB teams.

The histogram in Figure 4 represents the distribution of baseball talent across the entire population from which MLB can select its players. Most of the population has skills close to the average of the entire population. The best and worst players are much less numerous and are represented in the right and left tails of the distribution of baseball talent. MLB draws all of its players from the far right tail of the distribution. Line M, on the far right tail of ability, represents the cutoff for being a major-league player before an expansion. When the league expands, the cutoff shifts to the left to point E, thereby degrading the talent pool by filling the new roster spots with players who were previously unqualified to play in the majors. However, this decline will not show up in the average statistics that measure performance. The talent pool includes both offensive and defensive players; therefore, both offense and defense will suffer dilution. If the population distribution of quality for hitters and pitchers is the same, the effect of adding players should not affect the overall average performance of teams. Fringe hitters will hit the ball less, but fringe pitchers will give up hits more frequently; thus, there ought not be any change in the average measures of performance. The relative quality of play stays the same, as long as it is distributed equally across all clubs.
⁴⁹

However, the variability of performance will change, and this is where Gould’s contribution begins. The good players will have very good years while bad players will be even worse, because good players will have more opportunities against bad players. Though the relative quality of the game will stay the same, its absolute quality will decline. Fans will endure a lower absolute quality of play as the difference between the player abilities widens. There is no doubt that a wider difference in abilities can create some exciting moments, but we certainly do not want the game to go too far in allowing the very best in the league to dominate the worst. The good news is that we are not anywhere near the extremes of excessive talent compression or dispersion—where the players are extremely similar or different in ability—but how close are we to some desired optimum?

Many things have changed in baseball since the birth of MLB at the turn of the century. With the emergence of the American League in 1901, the size of the major leagues remained relatively unchanged for its first half decade. However, in 1917 MLB did expand team rosters from twenty to twenty-five players, which was equivalent to adding four new teams to the overall talent pool. From 1901 to 1960, MLB consisted of sixteen teams, until the league expanded in 1961–1962 (four teams), 1969 (four teams), 1977 (two teams), 1993 (two teams), and 1998 (two teams). But, while the league has nearly doubled in size since its founding, the population of the United States has nearly quadrupled. In addition, MLB has since integrated, utilizing African-American talent that it shamefully neglected for a half century, and many teams now fill their rosters with players from Latin America and Asia. If the population grows at a rate faster than baseball uses players, the number of high-quality players available to teams will rise. It stands to reason that if MLB draws its talent from an increased population base that the distribution of talent today must be smaller than in the past. This means that the absolute quality of play must be increasing.