clock menu more-arrow no yes mobile

Filed under:

Dominguez's Second Half

Talking Astros' Sabermetrics: What does Matt Dominguez's improvement in the second half of the season mean? Part I of 2.

Bob Levey

Matt Dominguez has been one of the more paradoxical Astros' hitters this year. Dominguez's offense struggled early in the season, and his offensive game was bad enough to raise questions about his ability to stick as the Astros' third baseman of the future. However, Dominguez improved his offense progressively each month of the second half. The question I will address is "what does this mean?"

Put Dominguez's first and second half together and the result is the following slash line for 2013:

.245, .290, .413, .703 OPS+ 93

Despite hitting 21 HRs, Dominguez's offensive line would seem mildly disappointing if you didn't know anything about his first and second half splits. Dominguez's .703 OPS is substantially below the .787 OPS he exhibited in 113 plate appearances last year. Dominguez's 93 OPS+ is well below the average AL third baseman OPS+ of 102. Dominguez's .290 OBP is poor by any standard, and falls well below the AL third baseman average of .323. Dominguez's 5.1% walk rate is weak and could lead to skepticism that he will have a decent offensive ceiling.

But the first and second half splits paint a picture of two different seasons for Dominguez:

First Half .227, .257, .383, .640 OPS+ 71

Second Half .271, .335, .454, .789 OPS+ 121

Even more telling is the progress each month of the second half. Here are the monthly wRC+ for June through September: 61, 100, 107, 121. If Dominguez's second half is indicative of the future, he could be one of the best third basemen in the league. Which performance is a better representation of Dominguez's future: the overall 2013 stats or the second half stats?

As a starting point, let's analyze the first and second half splits through the lens of advanced metrics.

Matt Dominguez

w/RC+ BB% K% ISO
1st half 72 3.60% 15.50% 0.156
2d half 119 7.00% 17.40% 0.183

1st half 0.235 17.9% 20.2% 10.2%
2d half 0.292 20.2% 9.2% 15.4%

Again, we see rather stark contrasts between the two halves of the season. Some observations:

  • His second half wRC+ would be the 7th best wRC+ among ML third basemen, behind Evan Longoria. Dominguez's full season wRC+ is ranked 18th.
  • Dominguez's second half walk rate is still below average, but appears serviceable at 7%. A 7% walk rate for a 23 year old in the major leagues is acceptable.
  • Dominguez appears to have exchanged a slightly higher strike out rate in the second half for more power and a higher walk rate. Given that Dominguez's strike out rate is relatively low, this is a good trade off.
  • The second half improvement has been driven to a large extent by a higher Batting Average on Balls in Play (BABIP). Normally, this would be a bad sign for the sustainability of the second half results. However, Dominguez's second half BABIP is not abnormally high and, for that reason, appears to be sustainable. On June 27, I calculated an x-BABIP for Dominguez of .274 - .295. At the time I suggested that the x-BABIP "target" could allow "him to push his OPS over the .700 threshold." That conclusion seems prophetic, and it's worth noting that his second half BABIP is only three points off the x-BABIP calculated with the Hardball Times formula and his full season OPS exceeded the .700 mark by 3 points.
  • Dominguez's infield pop up rate has declined significantly in the second half; his very high infield fly rate in the first half was a major factor in suppressing his BABIP.
  • Dominguez's line drive rate jumped from below average to above average. Line drive rates can be erratic, and this is a component which could see some regression from the second half to next season.
  • Dominguez's HR/Fly rate moved from below average in the first half to well above average in the second half. HR/Fly rates for batters can reflect a number of influences, including ballpark factors and the hitter's inherent power. At this point, I don't have enough information to predict the sustainability of his HR/Fly rate.

Dominguez's ISO appears to be lower than one would expect from a hitter with 21 HRs and a high HR/fly rate. Dominguez hits no triples (and his slow speed ensures that a triple would be a fluke in the future). He seemingly hits relatively few doubles, considering his strength and power. His lack of speed may have some effect on the lack of doubles too. But it's also possible that Dominguez's HR/Fly improvement in the second half reflects some doubles turning into HRs. It's useful to calculate Dominguez's HRs and doubles as a percentage of his total bases to analyze this possibility.

Percentage of Total Bases

Doubles HRs
1st half 25.4% 37.3%
2d half 20.2% 40.4%

The comparison shows that the rate of doubles declined from the second to first half, and the rate of home runs increased over that period. Taken in combination with the change in HR/Fly rates for the two periods, it's possible that Dominguez was more fortunate in the second half, with fly balls that fell for doubles in the first half going out of the park in the second half. This could suggest that Dominguez will face some regression from his second half HR rate in the future. However, even if Dominguez's full year HR rate is more reflective of his future HR power, it still results in a 21 HR run season and a HR/fly rate of 12%, which is above average.

How good is second half performance as a predictor of next year performance? A 2010 ESPN fantasy baseball article by Tristan Cockroft tackled this question. The article notes that many fantasy baseball players rely heavily on players' second half performance to predict break outs for the next season. Examining data for 2004 - 2008, the article found that 60% of players failed to sustain second half performance into the next season. As a general indicator, second half performance is not as good a predictor as full season performance. This isn't surprising that the larger sample of a full season is a better indicator of future performance.

But, hold on before you throw out Dominguez's second half stats. The ESPN article also found a strong age impact on the results. The study showed that players 23 years old or younger have a higher liklihood of sustaining or improving on the second half surge. 53% of players in that age group improved their performance in the next season and only 26% returned to past form or regressed. This pattern did not continue in the 24-26 year old group, and the predictiveness of second half performance deteriorated rapidly at higher age ranges. The 23 and younger age effect could be a selection issue (i.e., prospects who make it to the majors at that young age are inherently more talented), as well as a reflection of rapid development in that age range.

Guess what? Dominguez is 23 years old. This provides some hope that his second half performance surge is predictive of future performance.

In a second part to this article, I will examine the second half performance results for 23 and younger ML hitters in recent seasons to see if the data supports the conclusion of the ESPN study.