Astros Analysis: Where did the offense differ between 2017 and 2018?

Thanks in part to a historic offense the season before, the Astros were able to win the franchise’s first World Series. The question is how the offense differed in 2018.

To expect for the Astros‘ 2017 historic offense to repeat itself this year was always a bit presumptuous. Dependent on player regression, defense, luck, and other various factors, year-to-year performance can vary by small-to-large margins. Here are some questions to ponder before we see the data: Was the lineup more “unlucky”? Were notable injuries involved? What about the overall competition? Did the ball exhibit the same “juice” as last year? Park factors? The shift? Launch angles and exit velocity?

Let’s take a look the team’s offensive performance of the team as a whole during the past two years.

2017

• 238 home runs
• 896 runs
• 8.1% BB%
• 17.3% K%
• .196 ISO
• .309 BABIP
• .282/.346/.478
• .349 wOBA
• 122 wRC+

2018

• 205 home runs
• 797 runs
• 9.2% BB%
• 19.5% K%
• .170 ISO
• .289 BABIP
• .255/.329/.425
• .327 wOBA
• 110 wRC+

On the surface, it is plain to see that the offense was less potent than it was the season before. All the numbers trended in the wrong way except for walks, which rose by 1.2%. But the Astros’ offense was obviously not quite as dominant as the year before.

The eye-catching difference to me was the decrease in power as Houston hitters had 33 less home runs than the previous. The lineup’s isolated power (ISO) and slugging percentage also dropped by notable amounts. Honestly, it did feel like the Astros were missing one power bat all season.

Now, to answer the questions I presented earlier.

Question No. 1: Was the lineup more “unlucky”?

In short, yes, Houston’s lineup was more “unlucky” in 2018 when compared to 2017. For example, the club’s BABIP dropped by .026. Only four teams had a larger drop in BABIP than the Astros: the Diamondbacks, Marlins, Reds, and Tigers. Only Arizona finished with a winning record out of those four clubs, yet suffered a September collapse to prevent a return trip to the postseason. A roughly .300 BABIP is considered league average, and Houston was above the mark last year at .309. In 2018, the lineup’s BABIP dropped to .289. Keep in mind that “unlucky” may include a multitude of factors that we can’t properly quantify today. In general, though, “unlucky” is the appropriate term in this example.

Question No. 2: Were notable injuries involved?

Duh, they were involved. Notably to Carlos Correa and Jose Altuve, who both missed significant chunks of playing time. In fact, one could blindly argue that their lost playing time partially contributed to the deflated power numbers. Brian McCann and George Springer also missed some time as well. Throw in Yuli Gurriel‘s hamate bone surgery that robbed him of some power, and it is apparent to see a major reason behind the power decline.

Question No. 3: What about the overall competition?

Based on a simple review of ERA and FIP by opposing teams, specifically in the AL West, the opposing pitching staff’s did improve to a degree. Only the Rangers finished with a worst ERA in 2018 than the season before in the AL West. Both the A’s and Mariners exceeded expectations in this year while the Angels were near-.500 club. The fact that the Astros won 103 games speaks volumes about their club and how efficient the pitching staff was all season long.

Question No. 4: Did the ball exhibit the same “juice” as last year?

Last year, there were 6,105 major league home runs. This year? Home runs decreased by 520 down to 5,585. The comparison between the two numbers doesn’t confirm nor deny any changes with the makeup of baseballs, but it does start the conversation. There are plenty of studies and opinions out there to explore on the subject. Over at FiveThirtyEight, Rob Arthur and Tim Dix took an extremely close look at the baseballs used in recent years (prior to the 2018 season) to see if the physical properties of the baseball correlate to the home run spike. In short, it appears that the baseball did change. I’d highly recommend that you read their work like I am currently. The question to answer this offseason is whether the baseball was “de-juiced” in 2018, or did the adjustments made by hitters in the home run frenzy of 2017 didn’t fully translate over into this season?

Question No. 5: How about park factors?

I’m not going to pretend that I know everything about park factors. There is a lot that I’m still learning about with this information. That said, there were run scoring environment was actually better at Minute Maid Park this year when compared to last year.

2018

HR: 1.040
Runs: 0.987
Hits: 0.948
2B: 0.858
3B: 0.857
BB: 0.992

2017

HR: 1.009
Runs: 0.826
Hits: 0.900
2B: 0.810
3B: 0.654
BB: 0.927

The strange part of this data? Houston hitters actually performed worse at home in 2018 than the season before. For example, the Astros had a 123 wRC+ in 2017 at Minute Maid Park with 115 home runs. This year? A 108 wRC+ with 92 home runs at Minute Maid Park. For reference, Houston had a 111 wRC+ with 113 home runs away from Minute Maid Park. I’m curious to see how the park factors have ranged in the preceding ten or so years, but the drastic change of offensive performance at Minute Maid for the home team is something else.

Question No. 6: The shift?

In 2017, the Astros’ lineup had 1,209 plate appearances with a shift (traditional and non-traditional) was in play. Houston finished with a 86 wRC+ in those situations with 366 hits. For context, the best team wRC+ against the shift last season was the Marlins at 98. Fast forward to the 2018 season, and the Astros finished with a 88 wRC+ and 355 hits in shift situations through 1,190 plate appearances. Those are FanGraphs’ shift numbers, by the way.

By Statcast data, McCann was the most shifted hitter in the Astros’ lineup in the last two seasons. That’s not surprising. But numerous Houston hitters saw an increase in how many times an opposing team would shift against them. For example, Alex Bregman had a 17.1% increase in the number of shifts this year. Marwin Gonzalez saw his jump by 11.1% as a left-handed hitter. Josh Reddick‘s went up by 10.9%. Kyle Tucker saw a shift 36.1% of the time.

The offensive results were slightly better for the Astros when facing a shift, but it is noteworthy to see such an increase in the number of shifts from one season into the next. This is something to watch next season. The league average shift rate did jump from 12.1% in 2017 to 17.4% in 2018.

Question No. 7: Launch angles and exit velocity?

In 2017, the Astros’ lineup had an average launch angle of 12.0 degrees with an 87.7 MPH average exit velocity. Fast forward through 2018, and those numbers are now 13.2 degrees and 87.6 MPH. The difference in the average distance of a batted ball is only three feet less in 2018. Without pouring into even more data, there isn’t enough here to exclusively explain Houston’s decreased offensive output. But it is clear that much didn’t change between the seasons on this front outside of launch angle.

Overall, the Astros saw a decrease in offensive production from a number of categories. Personally, I believe regression, injuries, and a change in the baseball are the primary reasons why we saw a less potent Houston offense this year. The shift and a change in average launch angle may explain a small portion of the change. The improvement in the division is noteworthy, but it doesn’t explain the sudden difficulty to hit at Minute Maid Park. I probably didn’t provide a satisfactory answer, but I do believe it is important to start the conversation.