What if Carpenter Went the Other Way?

Update: Carpenter talked more about his approach with MLB and addresses some of the issues I mention in this article below. You can watch the video here:

Original article:

It was the Cardinals’ first game of the spring. A completely meaningless game that still somehow served as a microcosm of fans’ hopes and dreams for the season. Matt Carpenter stepped to the plate with Cardinals faithful anxious for any reason to believe that the former All-Star and MVP candidate had rediscovered his form over the winter. Here’s what happened:

(A huge thank you to STLSportsCentral for providing me with the video – which can be hard to get during the spring!)

Carpenter’s struggles in 2019 are well documented. I wrote extensively about it here: Age & Exit Velocity: Making Sense of Matt Carpenter’s Decline. (I encourage you to go back and review the material I presented back in December as it’s highly relevant to our conversation today.) In light of Carpenter’s lack of strength and declining exit velocity in 2019, this kind of swing and result were exactly what many Cardinal fans hoped to see early this spring. A thicker version of Matt Carpenter took a fastball middle-in to the opposite field with some authority for a line-drive single. Fortunately, Statcast was active for the game and they confirm with scientific precision what fans saw with their own eyes. The exit velocity on the hit was 95.9 mph.

The response from some on social media was exactly what you would expect. More than a few fans proclaimed that this opposite field approach represents exactly what Carpenter needs to do to recover his old form, beat the persistent shifting, and raise his anemic batting average.

Is that true? What if Carpenter went the other way more often?

First, let’s acknowledge two obvious points:

1) If Carpenter continues to get first-start-of-Spring-Training-quality middle-in fastballs and drives them to left-field at 95.9 mph during the season only good things will happen.

2) The next time Carpenter sees a 91.2 mph meatball, he should deposit it in the seats instead of proving he can inside-out a liner to left for a single.

Singles the other way are a fine result in certain situations and against certain pitches, but home runs hit anywhere are better. That’s not a discussion we have to have is it? Because that’s the question I have with the “Carpenter needs to take the ball to left more often” argument. Can he intentionally take balls to left field without sacrificing production?

Let’s focus specifically on what I am going to call Matt Carpenter’s “HR period”. From 2011-2014, Carpenter’s high in homers was 11. In 2015, he hit 28. From there forward, Carpenter actively began to sacrifice batting average in favor of a higher ISO (isolated power) and more round-trippers. His pull percentage climbed, and his middle- and opposite-field contact percentages dropped. The results were some of Carpenter’s best seasons by wOBA. From ’15-’19, Carpenter’s wOBA was consistently in the .370’s, with a low of .361.

Baseball noticed the change in Carpenter’s hitting profile. Extreme shifting became more common throughout the game, and Carpenter became one of the shift’s most popular victims. Opponents shifted just 34% against him in ’16. That increased to 88% in ’19 – the 8^th highest rate in the league.

The result was a further drop in batting average and batting average on balls in play (BABIP), but no noticeable change in wOBA until his down season in ’19.

In other words, the infield shift routinely steals the singles that prop up Carpenter’s batting average, but it has almost no impact on his power and that is where the heart of his production lies.

The following chart outlines Carpenter’s production on balls hit in the “opposite” direction, in his case, left field.

In 2016, Carpenter saw his lowest shift percentage during this time frame and also his lowest opposite field percentage and production on balls hit to left – 19.1% of contacted balls for a .133 wOBA.

As the shift became more common, so did Carpenter’s desire to go the other way. His opposite field rate increased 5% over the next three seasons and he seemed to get better at it. His left-field wOBA rose to .225 in ’18 and .312 last season.

That’s great, right?

Not really. His production on balls hit the other way has only improved from horrendous (.133) to marginally acceptable (.312) and his best season going the other way coincides with his worst season of overall production.

So, why does Carpenter lose production when intentionally going to left field? Exit velocity helps explain it. While Carpenter can occasionally drive a ball to left with authority (like we see in the video above), he can not do it consistently. His average exit velocity on contact to left field is well below his average exit velocity (to all fields) and his pull velocity to right field.

Range of Carpenter’s exit velocity by contact location:
Exit velocity on all contact ‘15-’19: 89.7 - 87.2
Pull exit velocity ‘15-’19: 92.3 - 88.8.
Opposite field exit velocity ‘15-19: 85.9 - 81.5

Sometimes baseball is simple. Hitting the ball harder and faster results in higher production pretty much regardless of defensive alignment. The difference between his right and left field exit velocities are remarkably consistent over this time frame. Carpenter hits balls to left field on average 7 mph slower than he does to right. That’s a huge difference, and it shows up in Carpenter’s startlingly high wOBA/xwOBA on pulled contact since 2015:

Matt Carpenter wOBA/xwOBA by season on pulled contact:
2015 - .513/.508
2016 - .556/.539
2017 - .476/.496
2018 - .577/.554
2019 - .410/.446

A .577 wOBA on pulled contact in ‘18 is just amazing. Scroll up and compare these pull wOBA’s with the opposite field numbers over the same time frame (which range from .133 to .312) and decide for yourself if you really want Carpenter to actively try to hit the ball the other way.

Of course, there is nuance here. Some pitch locations and pitch types benefit from an opposite-field approach.

For example, this pitch is similar to his first at-bat this spring. It’s a two-seam fastball thrown a little down and in. Carpenter smokes it at 101 mph for an opposite-field homer. Hitting the ball hard regardless of field is always a good idea.

This next pitch was lower than it was away, and was just barely out of the zone. With a 2-1 count, Carpenter was sitting fastball low/away, and he peppered it for a double to the left-field corner at 92.7 mph. If he had tried to pull this ball, the likely result would have been a grounder into the shift. Taking this ball to left with some heat was an excellent result for this pitch location. (Note: this is exactly the pitch location Carpenter is referring to in the MLB video posted in the article update.)

This last example is the kind of plate appearance that fans have to learn to accept. Carpenter crushes a middle-middle fastball 101.6 mph for a ground ball right at Albert Pujols. The shift positions Pujols perfectly, and our old friend can still pick it at first with elite skill sometimes. Carpenter’s approach here is perfect - he turned on a meatball and crushed it. Unfortunatley, he did not barrel it, resulting in a lower-than-deserved launch angle and an out. It happens.

Of course, there are plenty of examples of Carpenter rolling over on balls for ground-outs into the shift. No one wants to see those. And that’s why it does not matter so much where Carpenter hits the ball. What matters is how hard he hits the ball. Hopefully, Carpenter can continue to show improved production on balls hit the other direction. But, he also can’t let that take away from the pull power that is the basis of his once elite production.

Credits: A special word of thanks to Zach Gifford (@zjgifford) and John LaRue (@tdylf) for their help in working through the search parameters for this data.
Videos from MLB and Baseball Savant.
Stats from Fangraphs and Baseball Savant.
Note: Fangraphs and Baseball Savant have slightly different wOBA’s on batted ball location. Opposite field percentage is from Fangraphs, but the wOBA’s used are Baseball Savant, so they will be consistent with Savant’s Statcast exit velocities.