The (Double) Play’s the Thing

On August 13, Matt Carpenter stepped to the plate against the Nationals in a crucial situation. In a 2-2 game with Kolten Wong on first and one out, he hit a soft line drive to third base. Wong was running on the play, and Anthony Rendon easily caught the liner and threw to first to complete the double play. It was the third double play Carpenter had hit into this year. Amazingly enough, however, none of the three has been a ground ball double play. This is nothing new for Carpenter, who has long been one of the best in the game at avoiding double plays. In 4,232 plate appearances across eight major league seasons, Carpenter has grounded into only 31 double plays. Those numbers don’t sound like much out of context, but consider this: Fangraphs has a stat called wGDP, which measures the run value produced by avoiding grounding into double plays with a runner on first and less than two outs. Carpenter has been above average every year of his career by wGDP, adding nearly 11 runs of value over those years. That’s the fifth-highest total in the majors over that time frame. Matt Carpenter is certifiably elite at avoiding ground ball double plays.

The following day, Ryan Zimmerman came to bat in a similarly high-leverage situation. Trailing 4-0, he stood in against Daniel Poncedeleon with men on first and second and one out. On the first pitch, he hit a ground ball to shortstop, which Paul DeJong gamely converted into an inning-ending double play. Nationals fans won’t be surprised to hear this. When it comes to double plays, Zimmerman is the anti-Carpenter. Since coming into the league in 2005, he’s accumulated 6,884 plate appearances and grounded into 198 double plays. He has somehow been below average every single year of his career by wGDP, accumulating roughly -12 runs of value in that time.

In many ways, Matt Carpenter and Ryan Zimmerman are similar players. Carpenter is younger and retains more defensive versatility, but Zimmerman was an excellent third baseman earlier in his career, and both are best deployed at first base at this point. They’ve both struck out between 18 and 19% of the time in their major league careers. They’re both batting between .275 and .280 lifetime. Their slugging percentages are nearly identical; .475 for Carpenter, .477 for Zimmerman. The similarities continue. Matt Carpenter has maxed out at 26.5 feet per second on the basepaths this year per Statcast’s Sprint Speed metric. Zimmerman? He tops out at 26.6. Pretty much across the board, they’re similar offensive players. What, then explains the dramatic difference in double plays?

The first thing I thought was that batting leadoff would do a good job of limiting the opportunities Carpenter had to get into trouble. Surprisingly, however, there wasn’t all that much difference in opportunity. I looked to see how many times each player had come to bat in double play situations since 2012, Carpenter’s first full year in the league. In that span, Carpenter had 581 such opportunities while Zimmerman had 659. As I mentioned above, Carpenter has hit into 31 double plays. In 13% more chances, Zimmerman has hit into 86 double plays, a 177% increase. Opportunities don’t seem to be the culprit.

The next place my mind went was ground balls. Hitting the ball on the ground is just not a Matt Carpenter thing, you know? Take a look at the launch angles he has produced in his opportunities to ground into a double play in the last three years:

The peak of his launch angle chart is at 20%, a line-drive or hard fly ball angle, and there’s almost nothing below horizontal. Zimmerman, by contrast, is practically slamming the ball into the ground:

That is, indeed, a peak angle of 0%. Zimmerman has a great swing to produce all-fields contact. He pulls only 36% of his balls in play, below Carpenter’s 40% and significantly below the 50%-ish rate at which Carpenter is pulling the ball this year. This doesn’t keep him from topping a lot of balls, though. His swing might be pretty, but it incontrovertibly produces a lot of double play opportunities.

As beautiful and on-narrative as those graphs are, however, they don’t tell the whole story. In all, Zimmerman has produced 213 grounders out of his 659 plate appearances, while Carpenter has produced only 140 in his 581 PA’s. That still doesn’t come close to explaining the difference. If they hit into double plays at the same rate on each ground ball, Zimmerman would have 47 double plays, not the 86 he’s actually produced. There must be something else there.

I’ve left out one thing so far, one thing that I’m sure you’re screaming at me to address. Metaphorically screaming, hopefully. Zimmerman is right-handed while Carpenter is left-handed. Those extra few steps coming out of the box could be all the difference. Here’s the thing- I don’t have a good way to measure that. I tried for a proxy, though. I looked at the rate at which the two batters have turned ground balls into infield singles, reasoning that this would be a good proxy for how quickly they get from home to first. Still, though, no love. Matt Carpenter’s career infield hit percentage (infield hits as a percentage of ground balls fielded by infielders) is an uninspiring 5.7%. Over the same time frame, Ryan Zimmerman clocks in at 6.9%. As far as I can tell from the data, Zimmerman is better at turning grounders into singles. It would logically follow that he’s a little quicker getting to first base on average, but the double play numbers just don’t seem to line up. What gives?

I had one last idea. Well, that’s not fair. I had plenty of ideas, but most of them were terrible. I had one last brilliant idea. As I mentioned above, Zimmerman has a classic flat swing. It’s designed to lace grounders and line drives, and so Zimmerman tends to hit the ball on the ground pretty hard. Carpenter, on the other hand, is going for loft. He’s consistently worked to hit line drives, and his swing plane intersects most cleanly with the ball at those launch angles. When he hits grounders, it’s often the result of a mishit, a mistake rather than a purposeful result. Indeed, since the introduction of Statcast in 2015, Zimmerman has hit his ground balls at 89mph on average, while Carpenter has averaged only 82.5mph.

There’s an old announcer trope that a hitter didn’t hit the ball well enough for the defense to convert a double play. It’s usually said with a chuckle, a wry dig at the hitter for his poor contact. What if it’s true, though? What if Carpenter’s poor quality of contact on grounders works to his advantage? To check, I ran one more study. In the 2018 season, ground ball double plays have an average exit velocity of 87.6mph. Every other ground ball out, by contrast, averaged only 83mph. This is no small-sample-size mirage, either. There have been more than 3000 ground ball double plays recorded in baseball this year. A test for statistical significance returned a P-value of zero. The samples are different.

Here, I think, we’ve found a culprit. Zimmerman smashes his ground balls. Some of them make it to the outfield. Some find their way to fielders. The ones that find their way to fielders, though, turn into double plays at a brisk clip. It makes a basic sort of sense. The extra speed gives fielders longer to set and throw, and every split second counts when it comes to turning a double play. Paradoxically enough, Matt Carpenter doesn’t hit grounders with enough authority to ground into double plays.

Is this anything more than a statistical curiosity? I mean, there are two ways to look at it. One way would be to say that the difference between Carpenter and Zimmerman has been about 2.5 wins over their entire careers. That’s not a lot of wins. It’s hard to ground into enough double plays to truly hurt your value, and it’s hard to avoid enough of them to meaningfully boost your value. That’s a pretty boring way to look at it, though. There are all kinds of things that aren’t worth a whole win a year, and it doesn’t mean they don’t have value. Matt Carpenter incontrovertibly has a skill at avoiding double plays. It can’t be explained by his plate discipline, by his avoidance of ground balls, or by his footspeed. It’s a unique ability that comes out of the intersection of all those things as well as his unique swing. Every year, he adds value to the Cardinals with this ability. It’s real and it’s countable. Ryan Zimmerman is draining runs away from the Nationals in a similarly subtle way.

This brings me to a really neat fact about baseball. These effects, as real as they are, only build up slowly over time. We’re talking about 50 double plays over 7 seasons, basically. Seven double plays a year, barely more than one a month, separate one of the best in baseball from one of the worst. Ask Nationals fans if Ryan Zimmerman hits into a lot of double plays, though, and they’ll instinctively know it. Ask Cardinals fans if Carpenter does, and they’ll know he almost never gets doubled up. They may not know why- heck, I don’t know why, and I just spent hours trying to figure out statistics to explain it. Still, though, the baseball fan’s brain is an incredible computer for spotting patterns. One double play a month might be a faint signal, but it’s enough to be instantly visible to the dedicated observer.

If you take anything away from this article, take away the above paragraph. You, the reader, are amazing. You can tell at a glance something that is subtle and difficult to discern in the statistics, even with all the data available in baseball these days. If you take two things, though, also take away the fact that the “little things” in baseball really do have small effects. Baseball announcers aren’t wrong- avoiding hitting into double plays is a skill. It’s the kind of skill that is measured in 10ths of wins per year, though. Advancing the runner, hitting the cutoff man, avoiding double plays- they all have value. Anyone who tells you they’re worthless hasn’t really thought it through. What is accurate, though, is that their value pales in comparison to getting on base or hitting for power. Don’t let anyone tell you that wRC+ or wOBA or whatever statistic they want to use can describe a player’s whole offensive existence; but don’t put too much stock in the little edges either.