Seats Still Available: Trends in Ballpark Capacity

Ballparks are different today. The architectural revolution in baseball that began with the opening of Oriole Park at Camden Yards in 1992 has run almost a full course through the game. Camden Yards is now the eighth-oldest ballpark in the majors; Tropicana Field in St. Petersburg, Florida opened two years earlier, but was not used by a major league team until 1998. Only one of the older parks, Oakland’s Coliseum, looks like it could be replaced anytime soon. The face of baseball ballparks is a thoroughly modern one now.

That revolution changed more than the look of ballparks. It carried a new perspective on how many fans a ballpark should accommodate, and that new philosophy appears to be that less is more. A decline in average ballpark capacity, a quarter-century in duration, began almost at the very time Camden Yards was opened.

But why? Why did a trend that had persisted since the dawn of professional baseball reverse itself?

Hardball Times writer John LaRue gave some thought to ballpark seating a few months ago in his piece on the homogenization of major league ballparks. But matters of capacity took up only a few paragraphs of a much broader work. The issue deserves a closer look, and not only by the numbers.

Finding the Numbers

I decided to study the patterns of ballpark capacity in the major leagues, from 1901 and the foundation of the American League to last year. I used two sources to gather the necessary figures. The first was Philip J. Lowry’s Green Cathedrals, the 2006 edition. When its data ran out, I went to the ballparks database maintained at the Seamheads website. I used my authorial judgment where the two sources were in conflict.

When there was more than one capacity listed for a field during a season, I used the final capacity. When a team used more than one home ballpark during a season, again I used the final one, except where a team was forced to play elsewhere by mischance. One example is the deadly grandstand collapse at Huntingdon Grounds on August 8, 1903, which compelled the Philadelphia Phillies to move to the Athletics’ Columbia Park for the rest of the season.

One special case occurs in Cleveland, where from 1937 to 1946 the Indians split their home games between cozy League Park and cavernous Municipal Stadium. I chose to count as the home field the location where Cleveland played more games in each separate season. This makes League Park the home field during 1937-39 and 1941, Municipal Stadium in 1940 and 1942-’46. (And afterward, as the team abandoned League Park entirely in 1947.)

A different challenge arises with the Baltimore Orioles AL franchise of 1901 and ’02. Neither source offers any ballpark capacity for Oriole Park (no relation to Camden Yards). Not an estimate; not even a guess. To have some reasonable figure to work with, I assigned Oriole Park the same capacity as the smallest park in the majors of the day: the 6,500 of the Pirates’ Exposition Park. That guesstimate will suffice for the two most obscure years we’ll be observing.

Running the Numbers

What do historical ballpark capacities look like? LaRue presented a chart, but I will make my own here. The central line is the average (mean) capacity over the last century and a sixth. Above and below it are the maximum and minimum ballpark sizes over those same years. I will do some work with median capacity later, but for now the mean will suffice.

The rise in average capacity has had a number of surges. One began in the late 1900’s and lasted to around the First World War. A second surge began in the early ‘20’s, and was smothered by the Great Depression and later the Second World War. The third surge was more gradual, beginning around 1950 and enduring, with hiccups, until around 1990, when the downturn began.

The league maximums are more idiosyncratic, though they shadow the first two surges and the recent decline fairly closely. The minimums have followed their own beat. Their floor has almost always been rising, excepting the patch between 1941 and 1961 and a notable dip only in the last few years. There are explanations for both, lying outside the general trend, that I will talk about in due course.

The gap between the lines has a nearly symmetrical pattern of widening for several decades, then narrowing for several decades. A more precise measure is the standard deviation of capacities. One can show it in raw numbers:

This tells us that today’s variance has fallen almost, but not quite, to the level of the early 20th century—but that’s misleading. Today’s stadia are several times larger than those early parks that were, by our standards, very rickety and very flammable. Measuring the standard deviation as compared to the mean capacity gives us this:

Here we see an early trend toward sameness, interrupted in the early 1920’s and only resuming 40 years later. The relative standard deviation bottomed out in 2009, the year that Citi Field and the new Yankee Stadium came into the league. After a “dead-cat bounce,” it has fallen back almost to that level, much lower than in previous eras. This is the better representation of the variance.

John LaRue’s assertion is confirmed by the deeper numbers. Ballparks are much more homogeneous in capacity today than ever before, which is but one element of the general homogenization he saw.

Such homogeneity wasn’t pre-ordained. Baseball parks went through a lot of changes before we reached the approach to equilibrium, and the retreat from peak seating capacities, that we have today. So how did we get here?

Behind the Numbers

It began with the greatest year of pennant races in baseball history: 1908. Both leagues had breathless, tight multi-team chases, both decided on the leaders’ final day of play. A number of critical late games overflowed the modest stands of the day, including the final and deciding game in the National League. The replay of the “Merkle’s Boner” game between the Cubs and the Giants took place at the Polo Grounds, whose official capacity was 16,000. Over 40,000 crammed into the ballpark, filling foul ground and the depths of the outfield. The number of people who tried to get in is estimated at a quarter million.

Baseball had apparently reached a tipping point of popularity. It made sense to make the stands bigger, as accommodating a few huge crowds a year could pay for the expense. It also made sense to invest in more robust parks, as the surge of national interest had swept away any lingering fears that baseball might not endure as a business.

The plans for the Philadelphia Athletics’ Shibe Park were made before the 1908 season, but the attendance spike convinced several other owners to follow that example. Shibe Park opened at the start of the 1909 season. Forbes Field would open in Pittsburgh three months later. Those two fields led a wave of large steel-and-concrete parks that lasted through 1914, trailing off for some years afterward.

That’s where the first surge in capacity, and the initial drop in standard deviation ratio, came from. As it turned out, 1908 was an outlier in public obsession with baseball. Fans did not flock to the now-enlarged ballparks as much as owners had anticipated, not until Babe Ruth brought a new flavor of obsession to the sport.

With Ruth—indeed, for Ruth—came something new to baseball parks. It’s easy to decry the wave of multi-purpose stadiums that came into baseball in the 1960’s and ‘70’s, and I gladly join in that chorus. But they weren’t really new. The first one arrived in baseball in 1923, and it was Yankee Stadium.

The stadium wasn’t built to house only huge crowds watching the Bambino. It was also built with football in mind, which in that era meant college football. The NFL was in its infancy, and was looked on as faintly disreputable, much as baseball had been a generation before. The stadium was also intended to host track and field events. The original “warning track” was made of red cinders, to be used as a running track, and was set some distance from the outfield walls, except in left field. The long loop of the track also produced the extremely deep left-center field that came to be known as “Death Valley.”

A similar case arose in Ohio not quite a decade later. The city of Cleveland built its huge Municipal Stadium in hopes of drawing the Olympic Games in 1932—which went to Los Angeles and its huger Memorial Coliseum. “The Mistake by the Lake” was soon adapted to baseball use, settling after a while into its time-share operation with the much smaller League Park. Municipal Stadium got the dates when attendance was expected to be high, such as Sundays, holidays and, later, night games.

These two new, large stadiums helped drive the second surge of ballpark capacity toward the plateau it reached around 1940. It also pushed up the variance of capacities from the mean, their sizes of 60,000 to 80,000 counterpoised against small ones like Baker Bowl and Ebbets Field. Popularity of the new Babe-inflected homer-happy game, though, drove bandboxes like Ebbets and Wrigley Field to expand their seating, and the variance did not hold at its peak.

A new wave of ballparks came starting in the 1950’s, when teams began relocating. Interestingly, this initially didn’t move the needle on mean capacity too far. The A’s, in leaving Philadelphia for Kansas City, actually moved into a smaller ballpark. Braves Field in Boston had been expanded the year before the team decamped for Milwaukee, so their capacity ended up roughly the same at County Stadium. Only the Browns improved their seating significantly, going from Sportsman’s Park to Memorial Stadium in Baltimore.

The traumatic move of the Dodgers and Giants to California wasn’t a great overall change. The Dodgers did temporarily end up in L.A. Memorial Coliseum, with a capacity of over 90,000 and a field that could accommodate baseball after a fashion. This was balanced by San Francisco’s Seals Stadium, the dinky park that preceded Candlestick. The Giants’ ballpark capacity plunged by over 30,000 from this move, and even Candlestick left them over 10,000 short of the old Polo Grounds.

What did rise markedly was the standard deviation ratio. The moves to the huge Coliseum and the petite Seals Stadium raised that variance to its highest level ever. By 1961, it crept higher still thanks to expansion. The Los Angeles Angels played their first season at Wrigley Field in Los Angeles, a bandbox seating 20,457 that was the smallest park in the majors (and would have been as far back as 1938). The biggest ballpark held 362 percent more fans than the smallest that year—and both parks were in the same city!

The capping irony is that both teams moved in 1962 into the same facility, known as Chavez Ravine when the Angels were playing and Dodger Stadium when the Dodgers were. Standard deviation ratio plunged that year, starting a trend lasting half a century, as a fresh movement began pulling capacities closer together.

This was the multi-use stadium movement, though really it meant dual-use. New sports facilities were built with the needs of both baseball and football in mind. The main driver, for our purposes, was football’s need for bigger stands (logical when you consider that NFL teams had seven home games a year at the time, as opposed to 81 for baseball teams). This movement distorted both the field and the seating away from the ideal for baseball, but it was the price of getting the stadiums built in the first place.

It got a lot of them built. Some were of the despised, circular “Ashtray” style, but it went beyond that shape. From 1961 to 1977, I count 11 teams taking up residence in facilities built with both baseball and football in mind. This does not count the Montreal Expos’ move into Olympic Stadium, or the Toronto Blue Jays taking up residence in Exhibition Stadium, both in 1977. Baseball (and football in Montreal) was an afterthought at both locales.

This wave of parks ushered in the third great increase of seating capacity, a climb that would last for 30  years. It also brought a sea change in the distribution of capacities. For the previous half-century, early teens to early ‘60’s, the biggest parks had been the statistical outliers. Within a decade, the multi-use stadiums turned that around, and the smaller parks became the odd ones out.

I will demonstrate this with a chart showing mean capacity minus the median capacity (the point with half the stadiums above and half below).

When mean minus median is positive, a handful of bigger stadiums are pulling up the mean. (This is the more normal state. The sky’s the limit for higher capacity, but you cannot go below zero, so upward pressure is liable to be greater.) When the number’s negative, it means there’s a bulk of stadiums on the higher end, with a few small venues tugging the mean down.

The difference rose jaggedly from around 1910 to its peak between 1940 (when Cleveland Municipal Stadium became the Indians’ primary home) and 1960 (when the Dodgers used L.A. Memorial Coliseum). Then came the tumble. The difference was negative by 1970, as little parks like Wrigley, Fenway, and Parc Jarry in Montreal became notable deviations from the larger multi-use norm. Effectively, Ashtray Stadium was now the rule, not the exception.

In 1992, with the opening of Camden Yards, the most recent era of baseball parks commenced. It represented a decisive turn away from multi-use stadiums, which for our purposes means a withdrawal from the high end of capacities. I will give a fresh mean capacity chart here, for convenience and to spare your mouse’s scroll wheel.

The peak mean capacity came in 1990, at 53,880, but in 1993 was almost as high at 53,566. The downturn began in earnest the next year, when the Indians departed Municipal Stadium in favor of cozier Jacobs (now Progressive) Field. The fall-off was fairly steady for over a decade afterward, though it is now apparently bottoming out. Last year the figure was 43,142, up a couple hundred year-to-year even though the Atlanta Braves’ move to SunTrust Park took nearly 8,500 seats out of the major league total.

That move from Turner Field, originally built to host the 1996 Olympic Games, is part of the tail end of the rejection of multi-purpose facilities by baseball. It’s not the very end, though. Teams in Oakland and St. Petersburg are still playing in multi-use parks, and probably not coincidentally they are generally thought the two worst parks in the majors.

Both the A’s and Rays have recently taken whole sections of seats out of commission for their games. The A’s notably did this between 2005-6, tarping over 9,655 seats during their games so the empty expanses wouldn’t look too bad on television. The Rays removed around 3,000 seats from circulation for 2012 and another 3,000 for 2014, entrenching Tropicana Field as having the smallest capacity in the league. When Dodger Stadium, steady at 56,000 since 1962, emulated them and removed 1,500 of its seats for the 2015 season, it was an upheaval of baseball’s stability akin to the Yankees abandoning pinstripes or the Cubs winning the World Series.

Maybe I should rephrase that. Or maybe it’s right on target.

Surprising as that move was—Dodger Stadium had had an unchanged capacity for longer than any stadium in major league history—it left the place as still the biggest stadium in baseball. It’s been that since 2009, when New York traded Yankee Stadiums. (Champions before that? The pre-PETCO Padres. I’m amazed too.) If the Dodgers made that move, the momentum of smaller ballparks would appear inexorable.

It is not. Old, small ballparks that aren’t cursed by being inaccessible eyesores (sad but true, Rays fans) have been pushing upward. Fenway Park has gained around 4,000 seats since 2003, the Green Monster seats being a mere fraction of that. Wrigley Field has added almost 2,000 seats since that same date. And in a shocking about-face, Oakland opened up 12,000 upper-deck seats in 2017, giving the A’s their highest capacity since 1995.

The trend by the Cubs, Red Sox and A’s seems opposed to that of the Braves and Dodgers, but from another perspective they are the same trend. Teams with parks on the large side changed to smaller ones, while those with smaller capacity made their parks bigger. All ended up closer to the average, both mean and median.

Whether by changing locations or by adjusting the seating at their current ones, teams are creeping toward something of a consensus ideal stadium size. This consensus is shifted by market sizes, and sometimes restrained by sticking to older parks that aren’t easily changed, but the core number’s attraction is strong. That the Rays swam against the current toward this central figure speaks toward the distress of their market and stadium location.

(The Braves did have similar complaints about a bad location, yet built a smaller park in their preferred spot. However, they went not from small to smaller, but from large to around median.)

Still, that consensus number is lower than it once would have been. Much of the 21st-century capacity decline comes from a recoil from multi-use stadiums, but not all. Part of the remainder may be aesthetic, as John LaRue posited in his article. A stadium with fewer empty seats looks better, sounds better, feels better, both in person and on television. Adding to marginal enjoyment is considered worth trimming away marginal ticket sales.

Another part likely comes from owners reading the attendance numbers. They reached a peak just before the 1994-’95 strike, partly due to record-smashing attendance figures for the expansion Rockies, playing at immense Mile High Stadium. The strike brought a big drop in average attendance, one that took more than a decade for baseball to recoup. Economic calamity erased those gains with a fresh attendance plunge in 2009. The leagues rebounded partway over the next several years, then started dropping again. 2017 saw average attendance fall below the 2009 nadir.

Two sudden and sharp drop-offs in attendance definitely could convince owners they’re on the right track with retrenching on seating. One does wonder, however, how much the current fade in attendance represents a self-fulfilling prophecy. Lower the single-game ceiling, and you’ve necessarily lowered your season attendance if you have any sellouts. That number is thus likelier to drop, but when it does, did you anticipate it or cause it? Or some of both?

In the short term, the ebb in attendance isn’t hurting major league baseball. Teams still get their gate revenue, from luxury boxes and other super-premium seating, not to mention “dynamic pricing” that hikes prices for more desirable games. (There’s an echo of the old Municipal Stadium/League Park system there: instead of more seating for the better dates, it’s higher ticket prices.) In the long term, squeezing out fans from a day at the ballpark, an “initiating” experience that can deepen dedication to the game, carries quiet risks. One hopes this does not drive down the “ideal” stadium size in years to come.

Summing Up

There have been several discrete eras in ballpark capacity, largely paralleling the eras of ballpark design. It is perhaps an irony that ballpark sizes have become more standardized than ever in a time when teams have striven (with success) to make their parks’ appearances, especially from the inside, far more distinct. It’s less ironic that the average capacity has fallen in an era when baseball has gotten away from the distorting effects of oversized dual-use football-baseball stadiums.

An equilibrium seems to have arrived in recent years, perhaps to be expected as the post-Camden Yards surge in construction has wound down. Such relative stasis is not normal in the history of ballpark capacity. There are two possibilities. One is that teams really have found an ideal ballpark size, a steady point around which to hover. The second is that this is a short breathing spell before things go into motion again.

Which way would they go? To predict that might be to predict the future health or infirmity of baseball itself. So watch those upper-deck seats, and the sizes of the proposed new ballparks in Arlington and the Tampa area. They may have a lot to tell us.

References and Resources

• John LaRue, The Hardball Times, “The Homogenization of Ballparks”
• Philip J. Lowry, Green Cathedrals, 1992 and 2006 editions
• Seamheads’s Ballpark Database
• Baseball Reference
• Pro Football Reference
• Cait Murphy, Crazy ‘08