keyboard_arrow_uptop

When Wandy Rodriguez takes the mound in Philadelphia tonight, what kind of performance should the Astros expect? That seems like a fairly simple question, but in Wandy’s case, it’s considerably more complicated than it might be for the typical pitcher. Wandy’s weighted-mean PECOTA forecast prior to this season called for a 3.66 ERA; to date, he sports a 4.00 ERA, supported by a 3.77 SIERA. His ERA actually stood over 6.00 as late as June 23, but he’s been one of the hottest pitchers in baseball over the last two months, posting a 1.84 ERA in his last 11 starts and 73 1/3 innings. Still, given what we know about the predictive power (or lack therof) of hot and cold streaks, it seems safe to say that the Astros would be wise to expect a rate of performance in that 3.50-4.00 ERA range, with the caveat, of course, that performance can fluctuate wildly in any given start.

However, there’s more to Wandy than meets the eye. Check out his home/road splits over the course of his career (italics means better in the given location, while bold means worse):

Year

Statistic

Home

Away

2005

HRr

3.4

3.3

 

UBBr

10.6

7.9

 

SOr

12.1

17.2

 

ERA

5.57

5.47

Year

Statistic

Home

Away

2006

HRr

2.5

3.2

 

UBBr

10.7

9.8

 

SOr

14.1

18.2

 

ERA

5.48

5.82

Year

Statistic

Home

Away

2007

HRr

2.1

3.5

 

UBBr

7.7

8.2

 

SOr

23.2

17.4

 

ERA

2.94

6.37

Year

Statistic

Home

Away

2008

HRr

2.1

2.8

 

UBBr

6.5

8.9

 

SOr

24.1

19.8

 

ERA

2.99

4.34

Year

Statistic

Home

Away

2009

HRr

2.3

2.6

 

UBBr

5.6

9.3

 

SOr

23.0

22.4

 

ERA

2.08

4.05

Year

Statistic

Home

Away

2010

HRr

1.5

2.3

 

UBBr

9.0

6.4

 

SOr

24.5

15.4

 

ERA

2.85

5.30

Year

Statistic

Home

Away

05-10

HRr

2.3

2.9

 

UBBr

8.2

8.5

 

SOr

20.4

18.6

 

ERA

3.49

5.21

Year

Statistic

Home

Away

07-10

HRr

2.0

2.8

 

UBBr

7.1

8.3

 

SOr

23.6

18.9

 

ERA

2.68

5.04

Wandy has had vastly improved results since the National League got its first looks at him in 2005 and 2006, but most of that improvement has come at home. Since 2007, Wandy has embodied the nebulous concept of an “ace” in Houston, but on the road, he’s been well below average, bordering on replacement level. What should we make of this disparity? Should Brad Mills expect a subpar performance when he hands the ball to Rodriguez tonight, or can he be confident that he’s getting the same Wandy that he would at Minute Maid Park?

Sabermetricians have long been conditioned to downplay the capacity of home/away splits to provide insight; in most cases, when single-year splits are cited in support of a point, we rightfully scoff at the sight of small-sample flukes masquerading as analysis. However, in Wandy’s case, we’re talking about a sample of 508 1/3 innings at home, and 430 1/3 on the road. Can we really explain that away as a product of chance?

We can get part of the way there by accounting for the fact that pitchers generally do pitch better at home (or perhaps it’s more accurate to say that hitters generally hit worse on the road). The following table shows the typical home/road pitcher performance split over the last five seasons:

Years

Home IP

Home ERA

Road IP

Road ERA

Delta

2005-2009

111375.3

4.17

105170.3

4.61

0.44

Although Matt Swartz found that Houston had among the highest home-field advantages in recent years, he also found almost no correlation between HFA from one year to the next, and concluded that no team except the mile-high Rockies has an inherently higher benefit conferred by its home park. Wandy certainly hasn’t been enjoying the effects of a noted pitcher’s park; here are Minute Maid’s three-year (well, really 2 1/2-year) component park factors:

Runs

Hits

2B

3B

HR

BB

HBP

SO

1.014

1.002

.925

1.016

1.155

.946

1.179

1.075

Houston’s home stadium is fairly neutral; if anything, it favors offense very slightly, so Wandy’s isn’t an easily explainable split of the sort that we see with pitchers who call Petco Park home.

We’ve seen the difference in Wandy’s results, but let’s take a quick look at his process. Does the lefty’s stuff show any improvement at home? Here are his cumulative PITCHf/x splits for 2007-2010, courtesy of Eric Seidman:

Pitch

Setting

Thrown

Velocity

PFX

PFZ

SWSTR%

Fastball

Home

2237

89.8

3.93

10.77

5.4

Fastball

Away

1866

89.5

4.10

10.7

6.2

Curveball

Home

1397

76.1

5.34

7.99

13.1

Curveball

Away

1141

76.0

4.51

7.41

9.7

Changeup

Home

234

83.9

7.56

6.07

12.8

Changeup

Away

283

83.7

6.83

6.27

7.8

Rodriguez does seem to have gotten a touch more movement on his off-speed stuff at home, and a few more swinging strikes to go with it, but he’s also generated fewer whiffs with his fastball in Houston. Given these numbers, it’s safe to say that his repertoire doesn’t transform into that of a different pitcher in Texas; he’s thrown the same pitches, approximately the same percentage of the time, with roughly the same velocity and movement, but gotten much better results when doing so in his pinstriped home uniform rather than his road grays.

According to the Astros’ dynamic beat-writing duo of the Houston Chronicle’s Zachary Levine and MLB.com’s Brian McTaggart, both of whom graciously responded to my inquiries, Rodriguez has been asked about his struggles on the road many times, and has yet to offer a reason for his subpar performance. Maybe that’s because there simply isn’t one—at least not one that’s under his control. Rodriguez’s career BABIP at home is .289, but his BABIP on the road is an inflated .330.

That BABIP boost on the road hasn’t been accompanied by a vastly different batted-ball profile; Rodriguez has allowed only slightly more line drives away from Minute Maid, and given Colin Wyersresearch into scorer bias, it’s difficult to say whether the minimal home/road batted-ball splits that do show up in Wandy’s numbers accurately reflect what transpired on the field. As Matt Swartz observes, it’s difficult to envision a scenario in which a pitcher would induce weaker contact without commanding the strike zone to a dramatically greater extent, so we may have to pin Wandy’s road struggles on poor luck after all.

SIERA tells a dramatically different tale than Rodriguez’s ERA splits, backing up our BABIP-based explanation. Rodriguez’s career 4.13 road SIERA is higher than his 3.90 home SIERA, but not by a whole lot; in fact, that sort of differential is perfectly in line with what we’d expect, given the league-average splits quoted earlier. Of course, that doesn’t erase the fact that Rodriguez’s ERA split from 2005 to 2009 was roughly three standard deviations away from the mean. We’d expect only 0.3% of the population to fall outside of that range, but while it may not be an exciting answer, given our other evidence, it’s quite possible, if not downright probable, that Wandy belongs to (or is) that 0.3%.

It may be difficult to believe, but as dramatic as Wandy’s home/road splits have been, we probably shouldn’t expect him to show splits any more pronounced than the typical pitcher’s in the future. For his first two seasons, his pattern of performance at home and away looked much like any other pitcher’s; in his very last start, he shut down the Marlins in Florida for six innings, striking out 10 and walking only one. Rather than believe that Wandy’s mental or physical block away from Houston has lifted and descended selectively depending on his mood or the time of day, we should probably assume that both the pitcher and the process are the same, regardless of the mound on which they appear. When Wandy takes the ball tonight, the Phillies won’t be facing a weaker foe than they would in Minute Maid.