Saber-Slant #16: The Juan Pierre Question

Last week, I introduced a concept that most fans understand but few know how to properly apply: park factors. We all know that parks can definitely affect a players’ hitting, pitching, or even fielding performance (just ask left fielders who have to man the short space in front of the Green Monster in Fenway Park). To adjust for these problems, we can apply park factors that will neutralize these stats.

The park factors we talked about last week were based on runs, the most basic denominator in baseball. Run park factors are convenient because we can apply them to both offensive and pitching statistics pretty readily. However, park factors for individual stats can also be provided, and some people prefer such a method because of players like the punchless Juan Pierre.

The Juan Pierre question

Juan Pierre played his first two and a quarter seasons in Colorado, the most extreme hitters’ park in baseball. Because of the thin air in Denver, the ball tends to travel very far in Coors Field leading to a large number of home runs and increased power numbers by players in the park. However, here are Pierre’s numbers in his time in Colorado compared to his two seasons following his trade to the Florida Marlins.

When you factor in the fact that Pierre was a bit older in his first two seasons in Florida and likely aged into a slightly better player than in his Colorado years, you can see that the effect of switching from to an extreme hitters’ park like Coors Field to a more level (slightly pitchers’) park like Dolphin Stadium, you can see that there was little to no change in the performance of Pierre. If Pierre fairs no differently in Coors than he does in other stadiums, then should we even park adjust his stats at all?

The short answer is, as always, “it depends.”

Define the question

The first thing we should do when considering this is to define the question. When asking questions about baseball or in general, we should always know what type of answer we are seeking. In this case, why would we want to tackle the Juan Pierre question of park adjusting his statistics? There are two reasons why anyone would want to strip the effects of the park in a player’s performance: performance valuation and performance projection.

Evaluating value

The first task, valuation, has the goal of assigning value to a player’s performance. The value of a player’s contributions is heavily tied to the context in which those contributions occur. In the case of run contribution, the context involved is the run environment in which the player plays. Essentially, the run environment is a measure of how many runs are scored by teams in a certain game state. In this case, the game state we want measured is the league average versus games played in Coors Field. How valuable are Juan Pierre’s contributions in the overalll league versus in Coors?

Let’s assume Pierre has a skillset that produces the same numbers regardless of stadium. The following table compares the runs/out contributions of Juan Pierre* in 2002 compared to the runs/out (used as a measure of run environment) for Coors Field and the National League in that same season.

*Note: Runs for Pierre given by FanGraphs’ wRC metric, which is derived from wOBA and linear weights

With a quick glance of that chart, you can tell immediately that, even if Pierre’s statistics were exaclty the same regardless of where you put him, the value of his stats to the Rockies would be diminished because the team plays half of its games in Coors Field, a park with a much higher run environment. Pierre was .08 runs worse per out than the average hitter in Coors Field in 2002, while his performance was only about .01 runs worse than the league average per out. But since he played half of his games in Coors, half of his numbers are significantly diminished because of the higher run environment, thus requiring a park adjustment.

Eye towards the future

The second aspect mentioned, performance projection, involves not evaluating the past but predicting the future. If we were to predict the future for Pierre, the fact that his stats were further below average in Coors Field has less importance than if we were evaluating the past. On the other hand, knowing that he would produce similar numbers regardless of the stadium would definitely factor in how we can predict his future performance if he were to move to, say, Petco Park in San Diego. Even though we should use overall run park factors for past performance because of the effect of the park’s individual run environments, information about how a player’s skillset plays in different environments can play a much more valuable role when considering future performance in different environments.

There are plenty of examples of hitters and home parks interacting in odd ways because of the combination of skillset and dimensions. For example, Adrian Beltre played his home games in Safeco Field in Seattle the last six seasons. Safeco is well known as a bane to right handed power, and Beltre’s game played right into that aspect. Even though Safeco rated as a pitchers’ park overall, it probably affected Beltre’s game more than the average hitter. However, when evaluating his performance in Seattle, using the normal run park factor is more than appropriate. When Beltre moved to Fenway Park this season, many fans expected an offensive rebirth because of the move from Safeco’s death grasp on right-handed power to Fenway’s right-hander haven. According to park factors published by Patriot (linked last week), the move should affect his contributions by four percent, but Beltre may yet improve more than that because of those factors surrounding his skillset.

Conclusion

The information about how a player interacts with his park is important for establishing future performance, particularly when discussing moving to a different home park with a different run environment. However, even if the park seems to have little effect on a player’s stats, a run park factor should still be used to account for how that player’s performance compares to the run environment of the park. So, as always with questions of sabermetrics, the answer depends on what question you are really asking.

Follow @fscalltothepen

Tags: Adrian Beltre, Boston Red Sox, Colorado Rockies, Florida Marlins, Juan Pierre, park factors, Seattle Mariners