Park Factors Explained: Why Coors Field Is Not Fenway Park

By 4Baseball Staff · November 5, 2026 · 12 min read

How park factors are calculated, why one season is noisy and three are better, what the BPF and PPF fields in the historical record mean, and how to use them to compare players fairly across ballparks.

A park factor is a single number that says how much more or less scoring happens in a ballpark than in the average ballpark. A value of 100 means neutral. A value of 112 means that about 12 percent more runs are scored in that park than the league norm; a value of 89 means about 11 percent fewer. The historical record carries two of them for nearly every major-league team-season back to 1871: BPF, the three-year park factor for batters, and PPF, the three-year park factor for pitchers. In 2025, Colorado's Coors Field shows a BPF of 112 and a PPF of 115, while Texas's Globe Life Field shows 89 and 89. That is a gap of 23 points between the most hitter-friendly park and the most pitcher-friendly one in the table.

The reason the number matters is that home runs, batting averages and earned run averages do not mean the same thing everywhere. A batter who hits .300 for a club that plays half its games in a hitters' park has not necessarily done the same thing as a batter who hits .300 for a club that plays in a pitchers' park. Park factors are the tool analysts use to put them back on a common scale.

This article explains the arithmetic, the reasons single-season figures are unreliable, what the two historical fields mean and do not mean, and how to use a factor to adjust a statistic. It also lays out what a park factor cannot tell you, which is a longer list than most fans assume.

What a park factor measures

The intuition is simple. If you want to know whether a ballpark helps hitters, compare how many runs are scored in games played there with how many are scored in games played elsewhere, with the same teams involved. The standard method does this by looking at one team at a time. For a given club, add up the runs the club scored and the runs it allowed in its home games, divide by the number of home games, and do the same for road games. The ratio of the two is the raw park factor.

Why count the runs of both teams? Because the park affects everyone who plays in it. A tiny right-field fence helps the visiting left-handed hitter as much as the home one, so the measure uses total runs in each game rather than only the home team's offense. Why use the same club's road games as the comparison? Because the club's own quality is in both halves. A powerful lineup scores a lot at home and on the road, and a poor pitching staff allows a lot in both places. By dividing home by road, the team's quality largely cancels and what remains is the effect of the park.

A worked example, with invented numbers for illustration, shows the mechanics. Suppose a club scores 405 runs and allows 441 in its 81 home games, and scores 360 and allows 342 in its 81 road games. At home, 846 runs in 81 games is 10.44 runs per game. On the road, 702 runs in 81 games is 8.67 runs per game. The ratio is 1.205, so the raw park factor is about 120. Games at this ballpark produce about 20 percent more runs than games the same team plays elsewhere.

Why the raw number needs adjusting

The raw ratio is a starting point, not a finished answer. Three issues make a refined version necessary.

The first is that road games are played in other parks, not in an average park. The road environment is made up of the other 29 ballparks (or however many there are), and the park in question is not among them. That makes the denominator slightly different from the league as a whole, and adjustments exist for it. The second is schedule imbalance: in an era when teams played mostly within their own division, a club's road games came disproportionately from a few parks. The third is sampling noise, which is the subject of the next section.

Different publishers handle these refinements in different ways. Some regress the raw figure part of the way toward 100, on the principle that most ballparks are closer to neutral than a single season suggests. Some use multiple years. Some separate out home runs, singles, doubles and strikeouts to build component factors. The historical record is not a method document: its readme defines the two fields in a single line each, as three-year park factors for batters and pitchers. Anyone using them should treat them as published estimates rather than as the product of a formula they can reproduce from the database's other columns.

One year versus three: the noise problem

A season has about 81 home games. That sounds like a large sample, but the number of runs scored in a baseball game varies a lot from night to night. A typical game total has a standard deviation of roughly four to four and a half runs around a mean of about nine. If you average 81 such games, the standard error is about half a run, or roughly five or six percent of the mean. The park factor compares two such averages, one for home and one for road, so the combined error is larger, around seven or eight points.

That figure is a back-of-envelope estimate rather than a published number, but it illustrates the point. Even if a ballpark were perfectly neutral, a single season could easily produce a published factor of 93 or 107 from luck alone. Combining three seasons divides the noise by about 1.7, and five seasons by about 2.2. That is why the historical fields are labeled three-year factors, and why serious sources prefer multi-year figures when a park has not changed.

The trade-off is that the longer the window, the more likely it is to include years when the park or the ball was different. A ballpark that moved its fences, a league that switched to a livelier ball, or a club that began storing baseballs in a humidor changes the true factor mid-window. A three-year figure then reports a blend of the old and new environment. The sample-size problem in park factors is a case of the broader issue discussed in sample size and regression to the mean: the more data you pool, the steadier the estimate, and the more it may lag the truth.

Coors Field shows both effects. Its historical BPF was 129 in 1995, 125 in 2000 and 116 in 2002, then 112 in 2005, 109 in 2008, 120 in 2012, 118 in 2019, 113 in 2022 and 110 in 2024. The Rockies began storing game baseballs in a humidor in 2002, and the long decline from the 1990s is consistent with that change, but a three-year smoothing, plus the year-to-year noise just described, means the database cannot cleanly separate the effect of any single change.

What the BPF and PPF fields in the historical mean

In the Teams table, each team-season has a BPF and a PPF. The readme defines them as the three-year park factors for batters and for pitchers. Both use 100 as the neutral value. A BPF above 100 means the park favored hitters over the window, and a PPF above 100 means the same thing from the pitching side: the club's pitchers gave up more runs at home than the league norm would predict. In other words, a high number on either field means a hitters' park.

In principle the two numbers describe the same ballpark, and in practice they usually land within a couple of points of each other. Coors Field in 2022 shows 113 and 115. Oakland's Coliseum in 2023 shows 92 and 95. Cincinnati's Great American Ballpark in 2023 shows 103 and 104. Where the gap is larger, treat it as noise, not as a deliberate statement that the park helped one side more.

The values reach back to 1871, but early seasons were short and unbalanced, with few games and few parks, so the older numbers are far less stable than modern ones. They are also keyed to the team and season, not to the ballpark as a permanent object, which means a club that moves parks gets a new series of factors, and a ballpark that is shared or temporarily used shows up in the data under whichever club played there. In the 2025 table, for example, the Athletics appear at Sutter Health Park with a BPF of 102 and the Tampa Bay Rays at Steinbrenner Field with 100, both temporary homes.

The table shows how compressed the range is. Across the 30 clubs the average BPF in 2025 is almost exactly 100, with most parks between 92 and 105. Only two parks reach 108 or higher, and only one falls below 90.

Coors Field and Fenway Park

The two parks in the title are the standard contrast, and the historical data shows how the textbook picture needs updating. Coors Field's reputation is deserved: its BPF has been 109 or higher in every year shown and above 120 in the 1990s. The reasons are physical. Denver sits about a mile above sea level, where the air is thinner, so a batted ball meets less drag and a pitched ball breaks less. The outfield is also large, especially to center, which turns fly balls that would be outs elsewhere into hits.

Fenway Park has a different reputation, built on the Green Monster in left field, a very tall wall that stands close to home plate. The historical figures put Fenway modestly above neutral: BPF 104 in 2000, 106 in 2010, 105 in 2019, 106 in 2023 and 101 in 2025. That is a hitters' park, but a mild one. The trick is that the wall turns many fly balls into doubles rather than home runs, which boosts batting average and doubles more than total runs, and the park also has a fence that sits farther out in center and the gaps than along the lines, which takes some of the edge off. A single run-scoring number can hide a park with very different parts.

This is the main limitation of a total-run factor: it does not tell you how the park helps. Two parks with the same BPF can have opposite profiles, with one inflating home runs and another inflating doubles and singles. For analysis that requires that distinction, analysts use component factors, which calculate separate ratios for home runs, hits, doubles and other events. The historical Teams table does not include them. You would need event-level data to build them, and a good starting point for the physical side of the argument is exit velocity, launch angle and barrels.

How to use a park factor to adjust a statistic

The most common use is to adjust a hitter's or pitcher's raw numbers. The key idea is that a player plays only about half of his games at home, so the park factor applies to only half of his plate appearances. If a ballpark has a factor of 112, a hitter on that club faces a run environment 12 percent higher at home and a neutral one on the road, so his season-long environment is about 6 percent higher.

The simple adjustment is the average of the home factor and 100. For a park factor of 112, the average is 106, or 1.06. A hitter whose raw run production is 100 units should be credited with 100 divided by 1.06, about 94 units. For a pitchers' park with a factor of 89, the average is 94.5, or 0.945, and the same raw production becomes 100 divided by 0.945, about 106 units. The two hitters are producing the same raw numbers, but the second one is doing more.

Several public statistics build this step in. Ratios that scale a player's performance to the league average, such as the adjusted OPS and adjusted ERA families, include a park term of this kind. If you are reading a statistic with a plus sign or a 100-based index, check whether it is park adjusted before comparing players on different clubs. The same logic matters for team-level work like the Pythagorean expectation, which estimates a record from runs scored and allowed: a club whose park inflates both numbers will look different on the raw runs than on adjusted ones, although the ratio of runs scored to runs allowed is not affected much.

The next step is to remember that adjusted does not mean exact. The adjustment assumes the park helps every hitter equally, but a ballpark with a short right-field porch helps left-handed pull hitters far more than right-handed ones. It also assumes the park's effect is the same on the road as at home, when it is not: a club built around its park may be much better or worse at home than a generic team would be.

What a park factor cannot tell you

Several limits are worth keeping in mind before leaning on one number.

The first is selection. Clubs build rosters to fit their parks. A club that plays in a park that rewards fly-ball hitters may collect them, and a club in a pitchers' park may collect pitchers who allow few home runs. The park factor then partly measures the club's choices rather than only the ballpark, and it can overstate or understate the park's true effect.

The second is weather and conditions. Temperature, wind and humidity all change how far a ball carries, so the same park produces different results in April and July, and in a dome or under a closed roof. A single annual figure averages across all of those.

The third is that the factor is a statement about run scoring, not about skill. It does not say which players are good. It says how much the environment inflated or deflated their numbers on average, and the best use is as a correction, not a ranking.

The fourth is the quality of the data. The historical record provides the published fields, but not the underlying game-by-game splits, so you cannot recompute a factor from it or test an alternative method. For historical comparisons, the numbers are best read as a rough guide to which parks leaned which way, not as a precise measurement. The Athletics are a good case: the franchise has played in many parks across its history, and the figures for its recent Oakland years (BPF 95 in 2019 and 92 in 2023) fit with a pitchers' park, but a full picture of the franchise's greats, which you can read in the best players in Athletics history, has to be built park by park.

Common questions

What does a park factor of 100 mean? A neutral park, where the same number of runs are scored as in the league on average. Above 100 favors hitters, below 100 favors pitchers.

Which is better, BPF or PPF? Both describe the same ballpark and usually differ by only a point or two. BPF is the version from the batter's side and PPF from the pitcher's side. If you are adjusting a hitter, use BPF; for a pitcher, use PPF.

Why do park factors change from year to year? Part of the change is sampling noise, part is real: fences move, balls change, rooftops and humidors are installed, and the weather varies. Three-year windows smooth the noise but lag behind real changes.

Does a hitters' park make every hitter better? No. It raises the average run environment, but players with certain profiles gain more than others. A short fence helps pull hitters, and a large outfield helps fast hitters who hit line drives into gaps.

Can I compare a player's raw stats across eras and parks? Only roughly. For fair comparisons, use park-adjusted and league-adjusted numbers, and keep in mind what the database does and does not provide.

Where to go next

To see how scoring is distributed in the modern game, look at the league leaders and the stat calculator, and explore individual parks in the ballparks section. For the idea of valuing events by how much they change the odds of scoring, read run expectancy and the 24 base-out states. The same habit of comparing a result with what a situation would normally produce runs through aging curves in baseball.