Ballpark Geography

How 20th century building trends shaped the modern ball field

by Roger Weber

To build a major league ballpark generally requires more pages of construction documents than the number of innings that will ever be played there. No doubt they are some of the most complex creations ever built. To their designers, their architecture is a science of design fittingly parallel to the craftsmanship put into the statistical measure of the game they house.

Thus, it is worth a quick look at how that design has changed over the last 100 years since the introduction of the steel and concrete ballpark, specifically with regard to those elements that affect how the game is played. In this article, I consider several statistics to explain how the geography, or physical characteristics, of ballparks has changed since 1900 and influenced retro parks today.

Ballpark trends

Through 2008, 78 ballparks have opened for major league play in the contiguous United States and Canada since 1900. However, unlike, for example, the gradual increase in home runs over time, alterations to the ballpark landscape have come in drastic bursts. Of course, drawing lines to differentiate those bursts is no objective tasks, but consider, for example, the 12 ballparks built between 1909 and 1915. Other than these 12, zero ballparks opened between 1904 and 1922. Of the 12, eight were designed by the same firm – Osborn Engineering. Similarly, while just three parks opened between 1978 and 1991, 22 have opened since 1992, including 17 new retro-style ballparks, a renovated Anaheim Stadium, and four stadia that were converted for baseball having previously existed for other uses. For the purpose of this study, I decided it was worth considering construction in terms of trends, to eliminate outliers affecting the overall definable shifts over time. Under this assumption, we can define two clear groupings right away: the 1908-1915 parks and the retro parks, which, including Anaheim Stadium, which was renovated to fit the retro mold and renamed Edison International Field of Anaheim, comprise 18 of the parks built since 1992. Not coincidentally, 13 of them were designed, at least in part, by the same firm, HOK Sport (1). 

Of course, in addition to these two clear bursts of ballpark construction, a look Graph A (below) suggests another boom during the 1960s and ‘70s. However, it is more sensible to separate parks that opened during that period into two groupings. A number of new parks that were baseball-only and built away from downtowns opened in cities getting an expansion or relocated franchise. And several more, including two domed stadia, opened as multi-sport enclosed round venues. While it is not necessarily clear whether to include some of the parks that opened on either list, for the purpose of this study I used the following groupings for these two trends.

1950s/60s parks
County Stadium
Memorial Stadium
Candlestick Park
Dodger Stadium
Shea Stadium
Anaheim Stadium
McAfee Coliseum

1960s/70s round stadia
Astrodome
Fulton County Stadium
Busch Stadium
Jack Murphy Stadium
Riverfront Stadium
Three Rivers Stadium
Veterans Stadium
Kingdome

Graph A: New major league ballparks opened per year

Thus, it is at least reasonable to differentiate 45 of the 78 ballparks that opened between 1900 and 2008 into these four trend groupings. While the organization of these groupings is not an exact science, using them is helpful for understanding the reasons behind a number of the changes that have occurred to ballparks over the last century.

The size of ballparks

The 12 parks that opened between 1908 and 1915 were markedly different from today’s retro parks, despite largely being their inspiration.

Most visibly, the retro parks are significantly larger than the earlier breed. On average, the parks built between 1908 and 1915 had about 27,000 seats, compared to an average opening seating capacity of 44,720 among the retro parks. And while the average park in the Deadball era occupied about 5.9 acres (2), Jacobs Field, for instance, occupies about 12 acres (3). In fact, “the Jake” occupies about the same space as the former Cleveland Ballpark, Municipal Stadium, despite containing 86% fewer seats. That is not an anomaly, however. The parks built between 1908 and 1915 contained roughly over 20% more seats per acre of land they occupied.

This is not explained by a lack of foul territory or smaller average field size among the older parks. In fact, both of those notions are incorrect. The distance from home plate to backstop at Forbes Field exceeded 100 feet, and was originally 60 at intimate Fenway Park, far greater than the 48 feet from home to backstop at AT&T Park, which opened in 2001, or 49 feet at Minute Maid Park, which opened in 2000. Later in this article, I will address field and outfield size.

The discrepancy in seating is explained in part by smaller seats at the older parks, but more by the way the seating structures were configured. At all 18 retro parks, the top row of the upper deck is farther back horizontally from the field than the top row of the lower deck. Yet this was only the case at one of the nine ballparks constructed between 1909 and 1915 that had an upper deck when they opened. Four had an upper deck whose top row sat exactly atop the top row of the bottom deck. And at the remaining four, the upper deck sat atop the roof covering of the lower deck in such a way that the top row of the upper deck was actually closer to the field than the top row of the lower deck.

This type of construction required posts extending into the lower seating bowl to support the large amounts of weight the upper deck provided. As many Boston Red Sox fans are familiar, these posts commonly obstructed views in the lower deck. While the post phenomenon persisted at County Stadium, which opened in 1953, it largely disappeared in the 1950s and 1960s, when the second ballpark building boom occurred. The architects of Dodger Stadium, Anaheim Stadium, and the Oakland Coliseum, for example, recessed the upper deck so that supports for those upper levels could be placed behind the lower seating bowl, rather than obstructing views from seats. Unlike at the earlier parks which were jammed into already-crowded cities, it was not necessary to build seating levels directly atop one another because all seven of the parks I have classified as part of this trend were built significantly away from their cities’ respective business districts, centered among parking lots and naturally bordered by few obstacles that would have restricted the design to a set area.

The third major building trend -- the round “cookie-cutter” stadia -- brought the least-utilitarian utilitarian concepts out of the previous two trends in its seating structures. While these parks were large, with capacities averaging not quite 52,000, they were actually about as close to their cities’ downtowns as were the 1909-1915 parks, which were known for being shoehorned into tight city blocks. According to Google mapping, the sites of these round stadia were on average about 3.5 miles from their cities’ central business hubs, while the 1909-15 parks were slightly farther away, about 3.9 miles.

Despite their near-urban locations, though, the cookie-cutters employed the post-less technique from the ‘50s and ‘60s parks. As a result, these round ballparks covered large areas, and thus were largely responsible for the dramatic increase in ballpark building costs between the century’s earliest parks and those built today. Definitive cost data was available for eight parks built between 1909 and 1915 and seven of the round stadia. The average cost of these 1909-1915 parks was about $11 million in 2008 dollars, while the cookie-cutter parks cost, on average, over $230 million by the same measure (4). The architects of the retro parks have followed suit by recessing the upper deck at each of the new designs. Partially as a result of forcing the parks to occupy a larger footprint in this way, the retro parks, excluding Anaheim Stadium’s renovation, have averaged a $315 million price tag at the time they were built, costs that have, but for new Busch Stadium, been paid entirely by taxpayers (5).

Quirkiness

The retro ballparks are a mixed group as far as parks built near downtowns and those built away from them. But even the new parks built near downtowns are more like the cookie cutter ballparks than they seem aesthetically. The quirky features they contain are rarely a result of necessity, but of visual appeal. Consider, for example, the shape of the playing field at these parks. To do this, I looked at two variables: the first row of seats bordering the field in foul territory, and the outfield fence.

Excluding the oddly-shaped Polo Grounds, 11 other parks were constructed between 1908 and 1915. Of them, six contained first rows of seating down the baselines that, somewhere down the line, changed direction by an angle of less than 30 degrees.

These angle changes in the direction of the first row facilitated better sightlines for fans seated far down the lines. However, as parks became more spacious in the 1950s, ‘60s and ‘70s, those angles disappeared. Of the seven parks mentioned earlier as part of the ‘50s and ‘60s building trend and the eight cookie-cutter stadia, not a single one contained such a change in orientation of the first row. Their first rows, by contrast, from the dugout to near the outfield corner, followed straight lines angled toward the plate. Thus, when the retro ballparks began opening, they didn't include this phenomenon either. The first rows of seating at Camden Yards and Jacobs Field follow a contiguous angle all the way down the baselines. However, ten of the retro parks built since have employed those angles.

Much of the reason the mid-century parks lacked this trait was that at higher levels of the seating bowl, intentionally-placed curves in the seating structure rerouted sightlines closer toward the action. Such a concept was foreign among the architects building parks between 1908 and 1915. In fact, at no less than four of the 1908-1915 parks, no curvature was even used to wrap the seating structure the 90 degrees it must bend around home plate. At those parks, however, curving seating structures would not have been terribly useful, anyway, because they were built on small and restrictive plots of land. At the multipurpose parks, especially, the curve was useful because it allowed architects to design innovative rotating semicircular sections that facilitated the frequent conversion of the field between baseball and football.

However, as we have shown, the retro ballparks are not tightly bordered. Slightly curving seating structures would be more sensible for ensuring the best views. However, these were, after all, parks designed to be retro – emulating the 1909-1915 parks, which included few curved sections. Furthermore, many architects were likely hesitant to use a technique that was so popular at the round, multipurpose stadia. After all, those parks were becoming exceedingly unpopular around the time retro parks began being built. But For Busch Stadium in St. Louis, which the St. Louis Post-Dispatch lauded as one of the majors’ finest parks as late as 1990 (6), others of the round stadium trend were harshly criticized quickly once modern parks such as Comiskey Park and Camden Yards opened in 1991 and ’92. For instance, in late, 1991, the Pittsburgh Post-Gazette blasted the design of Three Rivers Stadium for its “artificial turf, thousands of seats far from the action, and the where-am-I feel of the place.” (7)

Ultimately, the cookie-cutter shaped ballparks hosted major league baseball an average of just 31.6 years, paling in comparison to the 63.0 years that the 1909-1915 parks survived (through 2008, when both Wrigley Field and Fenway Park still house MLB teams).

As a result of these factors, except at Turner Field, a converted Olympic stadium, not a single major curve was employed to the horizontal borders of the seating structure facing the field at a retro park between 1992 and the opening of the Great American Ballpark in Cincinnati in 2003.

The changes to outfield fences can be summed up similarly. As of rules established in 1892, balls hit over a fence less than 235 feet from home plate in fair territory were to be scored as ground rule doubles (8). This rule was revised several times, but the mere fact that it existed at all points to a much wider a range of dimensions between home plate and the outfield fences at parks built early in the 20th century. Consider the greater variation in distance from home plate down the left and right field lines and to dead center that existed at the parks built between 1909 and 1915 than at later-constructed parks.

Standard Deviation of distance from home to outfield fence:

It is worth pointing out that the middle two trends included only symmetrically-designed ballparks. Because the infield was symmetrical, designers assumed, so should the outfield when there is no constraint to where the fence should be. Such reasoning changed among the retro-park designers. But still, while asymmetrical ballparks are the norm today, outfield fences are nearly as similar to one another from park to park as they were during the cookie-cutter era, when most parks were within a few feet of 330-400-330 dimensions.

Some of this gradual standardization is due to rules established in 1959 that finally set official limits on field size:

“Any playing field constructed by a professional club after June 1, 1958, shall provide a minimum distance of 325 feet from home base to the nearest fence, stand or other obstruction on the right and left field foul lines, and a minimum distance of 400 feet to the center field fence.” (9)

The limited variation that does exist in fence distances at the retro parks, though, statistically appears somewhat to be random, and logically can be attributed mostly to aesthetics. For example, consider the outfield walls. At the early parks, walls that were especially tall in one area of the ballpark existed mainly because the distance from home to the fence was very short, or because team owners wanted to prevent nearby residents from getting free entertainment from their rooftops. Six of the 1909-1915 ballparks stand out for having an especially high wall in either left or right field. Among them, the dimension down the line to that higher wall averaged 334 feet, while the dimension down the opposite line to a comparatively shorter wall averaged 382 feet.

Thus, such designs seemed sensible. Yet among the retro parks, this relationship between wall height and distance from home plate to the wall does not follow so smoothly. Among the 13 retro parks whose wall height is over four feet higher than the rest of the outfield wall in either left or right field, the average dimension down the line to the higher wall is 326 feet, while it is 330 to the shorter wall, a very small difference.

Ultimately, though, the distance from home plate to the outfield walls is not especially unique at any of the retro parks. The shortest distance down one of the foul lines is 315 feet to the left field wall at Minute Maid Park, and the longest distance to center is 435 feet, also in Houston. Even those distances, however, pale in comparison to the Polo Grounds, whose right field foul line stretched just 257 feet to the wall, or Connie Mack Stadium, which, when it opened, required a 515-foot shot to hit a home run to center field.

The retro design again fails to match the parks it emulates when one considers the change in the shape of the outfield over time. At the early parks, the outfield fences often included sharp corners, especially in center field, because the fields had to follow the contours of the streets that bordered them. Because so many were built into cities whose street layout was a grid pattern, outfields developed in many of the parks at first as relatively squared-off areas. Consider that at the parks built between 1909 and 1915 the average distance from home to the center field wall was nearly 1.3 times the distance from home to the foul poles. A perfectly square field would have a ratio of 1.414. Yet over time, that sharpness to the shape of the field has muted to more uniform distances from home to the walls all the way around the playing surface.

Center field distance / Average foul line distance:

As the table above shows, this gradual decline has not let up with the design of the retro parks. Despite having dimensions more similar to the cookie cutter parks than to the 1909-1915 parks, architects have intentionally re-incorporated angular designs to the outfield walls that hint of the 1909-1915 parks. In fact, among the 18 retro parks we have considered, the average outfield fence includes five corners in fair territory when mapped from above, whereas the cookie cutter ballparks essentially included none. In appearance only, the retro designs mimic the century’s earliest outfield barriers.

Ultimately, however, the size of the outfield has less to do with a ballpark’s effect on the game than do other factors – namely elevation above sea level. While a large field may yield fewer home runs, it may also allow for more singles, doubles and triples. Thus, there is only a very weak correlation between field size and run production at the park.

Additional findings

The fact that retro parks are far more like one another than many fans realize can be further emphasized by a brief return look at their capacities as of when they opened:

A quick glance reveals that the capacities of the retro parks are actually even more similar to one another than those of the cookie cutter ballparks, which were so often maligned for their uniform designs. This data is slightly confounded due to the fact that, for this study, I have considered the Astrodome to be a cookie cutter. I did so because, upon being expanded in 1989, its design closely resembled that of other cookie cutters. Its original capacity, though, was only about 42,000. However, even removing the Astrodome for this analysis, the standard deviation among the multipurpose parks remains well above that of the retro parks.

The notion that retro parks lack reasons for many of their quirks can be paralleled by the argument that they were constructed less out of necessity and more out of fan interest than previous trends like the cookie cutters and ‘50s and ‘60s stadia. The cookie cutters that were replacing an old stadium in the same city were replacing parks that were, on average, pushing 50 years old. Retro parks, by contrast, replaced parks that were, on average, not quite 40. And the ‘50s and ‘60s-trend parks were all opening as the first home of a franchise in a new city. Furthermore, more retro parks have been built. Surely eighteen teams were not actually in desperate need of a new park all at the same time.

Regardless, retro parks have changed the landscape of baseball. The average major league ballpark is today seven years younger than it was in 1991. The range of ballpark ages has also increased slightly over that time period. However, such influence on the field of major league parks is not unique to this trend. Consider how the numbers changed during the time periods in which the other bursts of construction occurred:

A final point might be made on a non-physical characteristic of the parks - attendance. The parks that opened between 1909 and 1915 marked the most sudden burst of new construction, but it was reasonable considering the possibilities steel, a newly evolving technology at the time, provided ballpark construction. They replaced parks that were on average only 14 years old, and attendance at the first year of the new parks rose 21 percent from the last year at the old parks.

However, that is less than the 34 percent by which attendances rose on average the first year in retro parks. And even that figure is significantly less than the more than doubling of attendance on average at the first year of the cookie-cutter ballparks as opposed to the final year of the parks that preceded them. Even Veterans Stadium, the Philadelphia Inquirer noted, was once considered a “crown jewel.” (10)

Additional relevant points

• Among those 12 parks that opened between 1909 and 1915, the average seating capacity at the time the park opened was roughly 27,000, but they gained an average of 11,000 seats over the course of their existence from various renovations. At Tiger Stadium, the double decking of the stands around the outfield raised its capacity by over 30,000.
• The last park to open with artificial turf was Tropicana Field in 1998, and the last park to include sliding pits around the bases rather than an entirely dirt infield was Skydome, now Rogers Centre, which opened in 1989. Among current parks, the only one which maintains a field without a pattern mowed in the grass is AT&T Park in San Francisco, which opened in 2000.
• At several of the cookie cutter ballparks, the chalked linings of the batter’s box were reshaped not to include complete boxes outlining where a right and left handed hitter should stand, but instead one incomplete box surrounding where the catcher should squat as well as, within the same outline, where a batter on either side of the plate should stand.
• According to the baseball rule book, “it is desirable that the line from home base through the pitcher’s plate to second base shall run east-northeast.” In fact, this is the case at just nine of the 18 retro parks. 
• Post-1900 field-related rule changes (source: baseballlibrary.com):

1900 – home plate changed from 12-inch square to 17-inch pentagon
1904 – pitcher’s mound may not be more than 15 inches above home plate
1950 – pitcher’s mound at 15 inches above the base lines.
1968 – pitcher’s mound at 10 inches
1931 – fair ball to bounce over or through a fence is a ground rule double
1926 – ground rule double rather than a home run if ball is hit over a fence less than 250 feet from the plate

• The distance from home plate to second base is, by rule, about 127 feet, not 121 as a doubling of the distance from home plate to the pitcher’s mound might suggest.

Sources

Capacities, dates of ballpark opening and closing, dimensions, and cost figures used in this analysis are taken from Paul Munsey’s ballparks.com.

For several figures, I studied ballpark plan and section maps provided by Andrew G. Clem on his website, andrewclem.com/Baseball.

For much of my analysis, I referred to photographic evidence from the books, Blue Skies, Green Fields by Ira Rosen, The Sporting News’ The Ballpark Book, and Josh Leventhal’s Take Me Out to the Ballpark.

Attendance figures used for calculations in this study are borrowed from baseball-reference.com.

Photo of Nationals Park at the top of this article is by Joe Mock. All rights reserved.

References

(2)        Selter, Ron. Early Wrigley Field (Weeghman Park) 1914-23. The Baseball Research Journal, No. 37. Published 2007.
(3)        Brattain, John. My Dinner With Bess. TOTKSports.com. Published 6/8/01.
(4)        The Consumer Price Index: Inflation Calculator
(5)        Baltimore’s Experience, The St. Louis Post Dispatch, May 28, 2000
(6)        New St. Louis Ballpark Just Part Of Downtown Revival, Dow Jones News Service, April 10, 2006
(7)        New Ballpark is Coming, but then again, so is Skybus, Pittsburgh Post-Gazette, September 6, 1991
(8)        baseballlibrary.com
(9)        The 2004 Major League Baseball Rule Book
(10)      Phillies open 21st season at the Vet: A Concrete ‘Jewel’ shows its age, The Philadelphia Inquirer, April 13, 1991