Bandbox Revisited: A Quantitative Examination of Camden Yards
Never Let the Facts Get in the Way of a Good Stereotype

It's been two seasons, due to lack of data, but for several years I've annually updated an article I originally wrote in 1999 about Camden Yards and its offensive proclivities...or lack thereof.

In the original version of the article, I went back over all the numbers from Camden Yards' first eight seasons to see if anything had changed over time.  Eight seasons is a relatively short slice of life over a ballpark's history...now we have 12.

What I've done is update the data through the 2003 season, the twelfth in the park's history, and freshened the accompanying text a bit.  So let's get on with it.  We'll start by covering the method by which traditional park effects numbers are calculated and then jump right into the OPCY-specific stuff. 

Oriole Park at Camden Yards...well-known bandbox, acknowledged hitter's paradise, recognized pitcher's nightmare.  Right?  Well....

CALCULATING PARK EFFECTS AND DISCERNING MEANING

Let's take a moment to discuss how ballpark effects are determined and what the resulting indices mean.

The general concept that underlies all of the measures of park effects compares aggregate performance in the ballpark in question with aggregate performance in that ballpark's team's road games.  For instance, Camden Yards park effects are determined by comparing what happened in Orioles home games with what happened in Orioles road games.  It's important to understand that this means that the analysis includes what the Orioles and their opponents did in Orioles home and road games.  The logic behind this is that any impact of a team being "built around the ballpark"--that is to say, any bias introduced by only looking at players on the Orioles--is mitigated by considering (at least theoretically) all of the other players in the league.  It should be further noted that interleague games are not considered because teams don't play all of the teams in the other league and only play home-and-home with one interleague opponent.  Rather than accepting the potential bias thus introduced, these results are tossed out of the analysis.

The results are adjusted for differing opportunities--typically determined by the number of at-bats or plate appearances--in home and road games.  (In the ensuing analysis, at-bats are used for adjustment purposes.)

Let's look at an example:

In determining the Camden Yards park effect for home runs in 1993, we compare the number of homers hit in Orioles home games (168) with the number hit in Orioles road games (142).  We divide the home HR number by the road HR number and come up with 1.18.  In this analysis, it is traditional to use 100 as a base (meaning, no effect at all--the same number at home or on the road), so we multiply 1.18 by 100 and get a "Home Run Index" of 118.  But we haven't made an adjustment for opportunities.  In 1993, there were 5553 at bats in Orioles home games and 5427 at bats in Orioles road games.  That's an "At Bat Index" of 102.   In other words, there were 2% more ABs in home games than road games.  So, we must adjust the tentative HR Index based on the fact that there were more opportunities to homer in home games...we divide 118 by 102 (result:  1.16) and multiply by 100 again to reinstate the base-100 standard and come up with a total of 116.  That is the 1993 Camden Yards Home Run Index and it means that Camden Yards yielded 16% more home runs than could have been expected in a "neutral" American League ballpark that year.

CAMDEN YARDS OVER THE YEARS

The table below shows how Camden Yards breaks down statistically for each year since it opened in 1992.  The final row in the table shows the composite numbers for the ballpark--a cumulative look covering all of the park's nine years in one summary measure.  The yellow columns show the indices for the six statistical categories presented--from left to right, at bats, runs scored, batting average, home runs, doubles and triples.   The remaining columns show the raw totals for each category.  A quick perusal of the various indices will show that most (with the obvious exception of at bats) are susceptible to a fair amount of fluctuation over time; batting average is a notable exception.

There are some clear conclusions that can be drawn.   First--and this should come as no surprise--Camden Yards is a dreadful double/triple park (particularly the latter).  Composite indices--compiled over the park's twelve seasons of operations--show a triple index of 68 and a double index of 84, making the park a notably difficult place to double and triple.  The reasons for this are fairly obvious--a relative dearth of outfield space to cover between the lines and the foul lines, the high wall in right field which not infrequently produces some very long singles and retaining walls down both lines which do a fairly good job of funneling fair balls that roll into foul territory back toward pursuing outfielders.  Camden Yards' composite double index (87) from 2002-2003, for instance, represents the third lowest mark in the American League during that time frame.  (Only Safeco Field and Comerica Park had lower doubles indices and that 100 index in 2003 looks like a massive aberration.)  The composite triple index (48) over the same time period was the lowest in the American League.

Second, despite Camden Yards' reputation as a bandbox, its composite mark of 106 shows it as a pretty good homer park, but nothing like that which the self-proclaimed baseball pundits would have you believe.  It's 2002-2003 mark of 113--generally high during the park's history--placed it fifth highest in the American League over that period.  I would guess (I don't have the data in such a form to be sure) that it might actually rank a slot or two lower over the entire history of the park.  It has been a decent--but hardly magnificent--home run park.  (In fact, the home run analysis surrounding Camden Yards is more interesting than this when broken down further; we'll address that below.)  This wasn't always the case, however.   It's rather interesting to note the change that appears to have taken place at OPCY since the park opened.  After its first four years--1992-95--the park appeared poised to rival Tiger Stadium as the AL's most prolific homer facility.  But from 1996-2001 it resembled a truly good home park only once (1997) and slightly above average once (2000) while actually falling below the median line four times and has looked, on balance, very average.  In 2002-03, of course, the index has picked up again, but still hasn't reached the apex of 1994-95.  

Third, in terms of batting average, the park plays out just a shade below neutral.  Exactly why this is true is unclear, particularly given the tiny amount of foul ground at Camden Yards.  Check out the following table:

In trying to determine why Camden Yards played out as an average-at-best batting average park despite the lack of foul ground, I rejected the idea that long grass was keeping more ground balls in the infield--the grass just isn't that high.  (High grass definitely was at least part of the explanation for the relatively low batting averages at Tiger Stadium in what is otherwise a very good hitter's park.)  I speculated that it might be a problem with the lighting or the visibility generally.   If visibility was poor, I surmised, this would help explain the unimpressive batting averages.  I further concluded that, if visibility was a problem, the park should show a pretty substantial strikeout bias.  Now, the fact that I'd never heard anyone complain about the visibility at Camden Yards made me dubious of my own hypothesis and an investigation into the strikeout totals did little to mitigate that doubt.   There has been literally no park strikeout effect at all over the 12-year history of the facility so exactly why the park shows such mediocre batting average scores remains something of a mystery.

Fourth, and probably most importantly, the park doesn't score well on the run index at all.  With a mediocre twelve-year composite run index of 96--slightly below average--Camden Yards confounds the frequently heard attempts to label it as a "hitter's paradise."  It's not.  Not only is the ballpark not a great run-yielder, it's actually been very pitcher friendly of late.  Camden Yards has fallen below the AL mean in run index for eight consecutive seasons and has been below 95 in five of the last six years.  During the time frame 1999-2001 it sported a composite run index of 88, the lowest mark in the American League.  In 2002-03, the composite index of 93 was fourth lowest in the AL.  The poor run index is plainly a function of the mediocre batting average index and the very poor double and triple indices.  The only kind of hits, in other words, that the park yields fairly liberally are home runs.  Non-home run extra base hits (i.e. doubles and triples) are so low that, coupled with the unimpressive batting average index, the positive effect of the extra home runs is not only negated, it's overcome.

The conclusion is unmistakable.  Despite all the blathering to the contrary, Camden Yards has not been a great hitter's park over its relatively short existence.  It started out that way, but after the first three years, the park has been an average hitter's locale at best, and a relatively good pitcher's park the majority of the time.  Its batting average yield has been consistently middle-of-the-road or worse, its double and triple yield has been consistently poor (to the point where it has become the league's worst triple park in recent years and one of the very worst for doubles) and the home run yield--while still better than average--has largely been unremarkable.   If all of this surprises you, you're not alone.  Once upon a time, I too was sucked in by the park's cozy appearance, the unceasing "bandbox" punditry and the first few years where the park did, in fact, yield good offensive numbers.  But the composite numbers don't lie.  As good a hitter's park as it appeared early on, Camden Yards has actually been a pretty good place to pitch in recent years; certainly better than the average American League ballpark.

ADDENDUM:  THE LEFTY-RIGHTY HOME RUN DICHOTOMY

As promised above, a closer breakdown of the OPCY home run numbers reveals something very interesting.  Peruse the following table.

HR-L indicates home runs hit by left-handed batters; HR-R indicates home runs hit by right-handed batters.

Contrary to what seems to be the conventional view, Camden Yards is a better home run park for right-handed hitters than lefties.  While the number of opposite field home runs appears to have exploded over the past few seasons, there are still far more home runs that are pulled than hit to the opposite field.  When Camden Yards first opened I routinely heard people describe the park as a lefty home run hitter's paradise because "it's only 319 down the line in right."  For my own part, I took a look at the seven-foot high walls and the 364-foot sign posted at the left field power alley and figured that it was a great home run park for right-handed hitters.  

If you ever get a chance to see a scatterplot of balls put into play by big league hitters you'll see that a very small percentage of fair balls that travel at least 250 feet in the air land within 30 feet of the foul lines.  Most balls are hit closer to the middle of the field.  That's why the power alley distance is more important in determining a park's home run friendliness than the distance to the foul poles.  (Old Memorial Stadium played just 309-feet to the lines, but for most of its existence was not a good home run park.  Fenway Park plays only 302 feet to the right field foul pole but has traditionally been a terrible home run park for left-handed hitters; it should be noted that the Fenway right field power alley is 380 feet.)   Additionally, while the fence in left field is low all the way around, the scoreboard area of right field has a high wall.  While towering fly balls can yield cheap home runs, low line drives that might leave parks with deeper (but lower) fences stay in play at Camden Yards. 

Over the past few seasons, Camden Yards has settled into an essentially neutral home run park for left-handed hitters while displaying a clear and fairly consistent advantage to right-handed hitters.

Back to the Belfry