This article is a friendly rejoinder to a very stimulating post by Matt Gross over at Purple Row. So first things first. If you haven’t read that piece yet, you should do so.
In brief, Gross looks at the difference between the Rockies wRC+ at home and on the road since 2002. He identifies a significant discrepancy, to use his terminology, between the Rockies offensive performance at home and on the road. Notably, the difference is more extreme for the Rockies than any other National League club, and Gross maintains that the discrepancy supplies strong evidence that playing at elevation leads to a competitive disadvantage because of the difficult adjustment Rockies hitters repeatedly have to make to pitches below 5,000 feet of elevation. Gross, as well as a variety of the commenters in the thread to his post, point to a possible “reverse Coors effect.” While playing at elevation inflates offensive numbers, the ancillary result of playing 81 games in Denver is that it actually deflates the Rockies offensive numbers while playing on the road. Coors Field is the candy, and everywhere else is the toothache.
One of the advantages of using wRC+, according to Gross, is that it is a park adjusted metric. League average wRC+ is 100, and the park adjustment is supposed to mean that we can look at the Rockies wRC+ and be able to compare it with any other team’s wRC+. For example, if the Rockies have a wRC+ of 98, the team’s collective offensive performance was two percent worse than league average. If another team has a wRC+ of 102, they played two percent better than league average. Ostensibly, no home-ballpark caveats apply. Gross, as well as Richard Bergstrom, does consider the possibility that there is something uneven about the park-effect adjustment in wRC+, which might account for the Rockies’ severe home/road split. But if wRC+ reveals what it’s supposed to, then the data suggests that “park adjustment” doesn’t just apply to statistics, but is a necessity for the Rockies offense that has historically led to a disadvantage on the road.
I’d like to approach the same problem from a slightly different angle. To get a more complete picture of the Rockies past performance on the road, it is necessary to examine pitching as well as hitting. Instead of wRC+, I looked at the National League West’s differential between their offensive weighted On Base Average (wOBA) and wOBA against since 2009 (as far back as these statistics go on ESPN Stats & Info). I did this for each team’s total performance, as well as their home/road splits. I then identified the gap between each team’s home and road wOBA differential. The fact that wOBA is not park adjusted might mean that it’s more revealing, especially when considering home road splits. There is no need to adjust for park effects because both teams are subject to them. This exercise should provide a clear picture of team performance in relation to the rest of the division. As Dave Cameron notes, looking at wOBA differential is similar to looking at run differential to get at a team’s actual talent level. While run differential is a better representation of a team’s talent because it circumvents the randomness of win sequencing, wOBA differential is better yet because it avoids the randomness of the sequence of runs. Additionally, as Chris Chrisman writes at his excellent blog Checkswing Roller, while it’s important to take team hitting metrics into account, it’s critical not to ignore other, more basic, data points: namely, wins and losses. Chrisman examines the pitching side of things vis-à-vis the Rockies road record by exploring the “gap” between the Rockies home/road xFIP. The analysis below attempts to take in the team’s total performance.
The data dump for the 2009-2013 seasons is below. It includes the National League West teams as well as league average. A plus wOBA differential suggests that the team is performing well in terms of hitting and pitching—a minus differential points to a poor performance. Remember that wOBA is scaled to look like on base percentage, so a differential of .020 is equivalent to the difference between a .350 and .330 OBP. For example, in 2013, the Rockies’ offense had a wOBA of .319, while the opposition had a wOBA of .329 against Rockies pitching, resulting in a differential of -.010. The home/road gap, which is bolded below, refers to the difference between each team’s home and road wOBA differential. This number represents the degree to which each team’s overall performance differed while playing at home compared to on the road on the road.
|Home wOBA Diff.||.037||.028||.005||.036||.011|
|Road wOBA Diff.||.005||-.025||-.028||.028||-.019|
|Home wOBA Diff.||.050||.024||.018||.020||.011|
|Road wOBA Diff.||.-.021||-.004||-.009||-.005||-.044|
|Home wOBA Diff.||-.002||.007||-.006||.006||.024|
|Road wOBA Diff.||-.006||.-.002||.020||.006||-.017|
|Home wOBA Diff.||.000||.022||.006||.024||.004|
|Road wOBA Diff.||-.047||.-.006||-.016||-.010||.008|
|Home wOBA Diff.||.017||.003||-.002||.003||.010|
|Road wOBA Diff.||-.040||-.007||-.031||.028||-.018|
The data suggests that the Rockies wOBA split at home compared to the road is less due to extreme park factors and more due to fielding either good or bad teams. There is a lot of information here, so I’ll focus on a few things that stand out to me (by all means, chime-in in the comments below). Let’s look at 2009. In that year, two teams in the NL West had a positive wOBA differential both at home and on the road, the Rockies and the Dodgers. Both teams’ home differential was essentially the same, .037 for the Rockies and .036 for the Dodgers. The Dodgers home/road gap, however, was smaller (.008) because they played almost as well on the road as they did at home. The Rockies gap was larger, .032, which suggests a more pronounced home/road split in terms of performance. Notably, the team with the most severe gap between home and road wOBA differential was the San Francisco Giants, at .053. This was almost entirely because the Giants were awful away from AT&T Park. In fact, the Giants finished the year with the best home record in the division, at 52-29. They were just 36-45 on the road, however, which is reflected in their poor differential. In the end, the Giants finished four games behind the National League Wild Card, the Rockies, and seven games behind the division winning Dodgers.
The following season told a similar story. The Rockies led the division in wOBA differential, with a very high .050 differential at home. They also finished with the most home wins, 52, which was the same number of games the Giants won at home in 2009. But, put kindly, the Rockies road performance was substandard. The team’s road record, 31-50, was worse than the Giants’ record in 2009, but the Rockies road wOBA differential was marginally better: -.021 compared to -.025. In 2010, the Giants ended up winning 43 games on the road despite a -.004 wOBA differential while playing away from home. It’s essentially an even mark, but it was enough to get them into the playoffs—a necessary condition to winning the World Series, which is what they went on to do.
The Rockies were good in 2009 and 2010—what about bad seasons? There isn’t one worse than 2012, so it should serve as a fine example. In 2012, the Rockies were bad both at home and on the road. In terms of wOBA differential, they were even in 2012, which was a slight improvement from the negative home differential the team experienced in 2011. The team’s road wOBA differential was a very bad -.047, which resulted in a home/road gap of exactly .047. These performances schemed to provide 98 losses for the Rockies. While bad, the gap isn’t really an aberration, even in this limited sample. The 2010 Diamondbacks, for instance, finished with a -.055 wOBA differential. They managed to lose one fewer game than the 2012 Rockies. Last season was also shoddy, but it wasn’t all-around bad like 2012 was. The Rockies actually led the division in wOBA differential at home and ended up going 45-36 at Coors Field, but their road mark of 29-52 neutralized any home success the team might have. Again, these performances weren’t anomalous even last season. The Padres finished with the same home record despite a roughly neutral home wOBA differential. They weren’t quite as bad as the Rockies on the road, as they finished with a -.031 wOBA differential and a paltry two more wins.
The only consistent story in the data above seems to be that good teams tend to be on the plus side of wOBA differential while bad teams are in the negative. Most notably, the Rockies home/road gap from year to year has fluctuated wildly. It was as large as .071 in 2010. While that is an extreme gap, it’s almost totally due to inflated hitting at Coors Field. The team’s .050 home differential was the greatest of any NL West team since 2009, but the Rockies’ -.021 road differential was just in the normal range of bad. Moreover, the Rockies extreme .071 gap in 2010 starkly contrasts to the gap the very next year, which was .004. It then jumped up to .047 the following year. The other NL West teams also experienced ranges. San Francisco’s gap was as small as .009 and as large as .058, the Dodgers ranged from .008 to .034, and the Diamondbacks from .004 to .055. The San Diego Padres had the most consistent gap which might have suggested something systemic…. except in 2010 they were better at home, in 2011 better on the road while having a similar “gap”. Then in 2012, they were better again at home. In other words, there was nothing consistent that pointed to the Padres perpetually doing better at home or on the road.
If the Rockies as a team are subject to a systemic competitive disadvantage due to playing at elevation, then there should be more consistency in measures of performance. I see the fluctuations of good and bad teams. Competitive disadvantages abound in Major League Baseball—small markets, relatively limited finances, and poor ownership are just three variables that can lead to competitive disadvantages. I don’t think playing at elevation is one of them, but the question certainly deserves attention.
All Statistics courtesy of ESPN Stats & Info