I started this analysis prior to the announcement of the three year deal between the Astros and Wandy Rodriguez. At the time, we had some discussion in the comments of the possibility that Wandy might be traded. I brought up Wandy's substantially better performance at home as a possible issue which might worry other teams. That aroused my curiosity; the mystery is why the substantial home / road split exists.
The trading implications of this issue are no longer significant, given the new contract. To the extent that Wandy's substantial advantage at home is a given, the existence of the split isn't important to the Astros. The total performance, both home and away, makes Wandy a good pitcher for the Astros. But suppose the home / road split isn't immutable. Maybe a greater understanding of the cause of the split would tell us that there is something Wandy can do to improve his road performance. Given that Wandy's three year average home ERA is 2.68, an improvement in his road ERA to the level of an average split would make Wandy an elite pitcher like Halliday or Lee. I won't pretend that I have the answers, but take this investigation in the spirit of trying to see if there is a "why."
Wandy Rodriguez's Home / Road Split
Most pitchers have better performance at home then on the road; that's why they call it a home advantage. The average home / road split for NL pitchers in recent years is about 0.4 runs difference in ERA. But Wandy's home / road splits have been consistently and substantially higher than that over the last four years. Wandy's career home / road ERA differential is 1.5 runs. During the period 2007 - 2010, Wandy's home ERA is 2.68 and his road ERA is 4.80, for a 2.12 run differential. Wandy's OPS-against at home is .689 vs. .811 on the road, over his career. Given the strength and consistency of this split over a number of years, it becomes more difficult to discard this tendency as a fluke.
The history of Wandy's home / road ERA differential may hold some hints as to the "why" question, but it would take more detailed information on how Wandy specifically improved the direction of his career. At the beginning of his big league career, 2005 and 2006, Wandy had a below average split (.35 run advantage at home), which is another way of saying that his home and road ERAs were equally bad (in the 5.5 range). In 2007, the large split surfaced: his home ERA declined to 2.99, but his road ERA was over 6. In the subsequent 3 years, Wandy's home ERA declined further, and his road ERA settled into an ERA in the mid-4's. While Wandy's road performance is certainly better than it was at the beginning of his career, the most significant factor for Wandy's improvement was a big leap forward in home ERA.
Comparable Home / Road Splits
How common is a split of this magnitude? I examined home / road differentials for starting pitchers who had steady employment over the 2007 - 2010 period. The answer is that the magnitude of Wandy's differential is unusual. But there is one pitcher with a slightly greater split over that period, and three pitchers with relatively large home / road differentials.Kyle Lohse has a similar, even larger, ERA differential. Joe Blanton and Joel Piniero had large differentials, though less than Wandy's over that time frame.
Home / Road ERA 2007 - 2010
Lohse 3.59/5.98/ 2.39
Wandy 2.68/4.80/ 2.12
Piniero 3.34/ 5.03/ 1.68
Blanton 3.76/ 5.06/ 1.30
Really, I don't see any remarkable similarities among the types of pitcher on that list. Wandy was a significantly better pitcher than the other three starters, with better ERAs on both the home and away side. Lohse, Piniero, and Blanton have made a living as mid- to lower end rotation starters. On average over their careers, all four pitchers have FIPs below their ERAs. But the margin is relatively small for all of them. Interestingly, two (Lohse, Piniero) of the four pitchers were Cardinals during this period. Add in Jeff Suppan, also a former Cardinal for part of the time period, who had the fifth largest home/road differential in 2007 - 2010, and the Cardinals connection seems curious.
Is the H / R Differential Team-Specific or Player-Specific?
The fact that Lohse, Piniero, and Blanton have changed teams frequently may provide some information. If the split is caused by the player's personality or general comfort with pitching before a friendly crowd, I would expect the player to maintain a large split when he changes teams. If the player's home / road ERA differential is much higher with one team, then that could point to a ballpark-related factor.
I compared the average H / R ERA differeential for Lohse (Cards vs. Twins), Blanton (Philadelphia vs. A's), and Piniero (Cards vs. Mariners). For Blanton and Lohse, the differential changed significantly (doubled) with a change in teams. For Piniero, the change in differential was insubstantial.
No definitive conclusion here, but this small comparison doesn't support the notion that the pitchers take their home / road splits with them.
Does the Ballpark Affect H / R Split?
When the Astros first moved to their new ballpark, amid the concern over the short LF dimensions, manager Larry Dierker said that pitchers would learn how to pitch to the park. Over the last few years, Minute Maid Park's ball park factor has reflected a mild advantage for pitchers. The recent ballpark factor for NL Central parks doesn't indicate that divisional play would cause an abnormal ballpark factor for the road.
Over the period 2007 - 2010, the Astros' pitchers as a group exhibited almost 1 run higher ERA on the road compared to the road. (Of course, this includes Wandy's contribution to the home / road split.) This is more than twice the average differential for starting pitchers. This suggests a possible team specific H / R split. However, the MMP park factor wouldn't appear to be large enough to account for the differential by itself. Maybe it's possible that Astros' pitchers have learned to pitch to the nooks and crannies of MMP, and perhaps Wandy's repertoire is particularly well-suited to that task? Notice that I put a question mark there.
Fielding Independent Pitching
So far, I have been utilizing ERA to compare home / road splits. Fielding Independent Pitching (FIP) is a better pitching stat for isolating the performance factors which are within the pitcher's control. Will this stat tell a different story?
We have to realize that the range of FIP is somewhat more compact than the range for ERA. For example, the difference between highest and lowest teams is 70% higher for ERA than FIP. The median difference between home and road splits is around 0.3 runs on the FIP scale.
Wandy's career H / R differential for FIP is 0.49 runs, which is still larger than the average FIP split. We expect that Wandy's differential would decline, relative to ERA, due to the more compact range for FIP. But a reduction from 1.2 runs (ERA) to 0.49 runs (FIP) is larger than expected due just to that factor. Interestingly, Kyle Lohse's career H / R differential shrinks in a similar manner if FIP is used instead of ERA.
The fact that Wandy's home / road differential is more significant based on ERA instead of FIP suggests a couple of possibilities:
- Perhaps Wandy has had worse defense behind him on the road compared to at home. The defensive efficiency ratio (DER) for Wandy at home is .714, but only .672 on the road. DER has its flaws as a measure of team defense, but the differential is consistent with this possibility. We know that ballparks influence defense, but the relationship is murky. Maybe the Astros' defenders are more familiar with the nuances of MMP and field better there. This begs the question as to why the H / R impact would be greater for Wandy than other Astros' pitchers.
- Maybe Wandy has just been unlucky on balls in play on the road (the fluke answer). This follows from the first point, since Batting Average on Balls in Play (BABIP) is the inverse of DER. Wandy has an above average BABIP on the road (.328) and below average BABIP at home (.286) over his career. We don't know how much to attribute this to defensive differences, luck, or some other factor. Since this is a career differential, I lean against the idea that its purely a fluke.
The difference between Wandy's x-FIP and FIP at home and on the road may provide some additional ideas. x-FIP normalizes a pitcher's FIP for an average rate of HRs per fly ball. For his career, Wandy's home FIP is somewhat below his x-FIP, and on the road, Wandy's road x-FIP is somewhat below his FIP. (Based on FIP minus x-FIP, the difference between the two metrics as a H/R split is .35 runs.) In other words, over his career, Wandy has been more effective at controlling HRs in MMP than on the road. What makes this interesting is that MMP's park factor for allowing HRs is above average and LHPs are most at risk for allowing HRs in MMP, given the short Crawford Box porch. Perhaps Wandy has found a way to minimize HRs to LF; this helps a lot more at MMP than other parks. I should point out that this FIP minus x-FIP differential isn't completely consistent on a year to year basis, which means that it probably isn't a sole or even primary explanation for Wandy's H/R splits.
At the end of the day (and this article), I don't have an answer to the "why" question. But I feel like I have made some inroads to understanding the issue. As far as strategic considerations go, I would suggest maximizing Wandy's home starts, to the extent that can be accomplished through rotation scheduling. It's also worth considering if improvements could be made to the defensive alignment when Wandy pitches on the road. In parks with large LFs, perhaps Carlos Lee should play 1st base and a better fielder should be stationed in LF.