Did the Mets miss something with Collin McHugh?

On June 18, 2013, the Mets made a seemingly inconsequential trade, swapping the recently-designated-for-assignment Collin McHugh for the recently-designated-for-assignment Eric Young Jr. At the time, the trade was nothing really notable; two mediocre players getting swapped by teams with different areas of depth. When Young went on to steal 30 bases in the second half and provide arguably the most electric performance at the top of the lineup since the days of Jose Reyes, the trade looked like a definite win for the Mets.

However, this year Young has come back to Earth. The speed is still there, but he’s put up a paltry 78 wRC+ in 253 plate appearances. Young is at best a fourth outfielder on a good team, given that he offers basically no other skills aside from the ability to steal a bag. Meanwhile, McHugh, who pitched all of 19 innings with Colorado in 2013 before moving on to the Astros, has struck out more than 10 batters per nine innings en route to a 3.32 ERA in about a half-season’s worth of innings.

Without opening the process-versus-results can of worms, is there something we could point to that would show the Mets made a mistake back in 2013? Or has McHugh simply improved as a pitcher, as young players often do? PITCHf/x data offer a good tool for answering this question, providing a method for analyzing the effectiveness of each of McHugh’s pitches as well as any changes in his pitch mix. Before jumping into more obscure data, though, we should look at McHugh’s minor league stats with the Mets, presented below, in order to gain some context:

McHugh actually put up decent numbers at every stop in the minors. Aside from his two brief stints in rookie ball (41 innings in 2008, 27.1 in 2011), McHugh’s highest FIP mark comes in at 3.56. Though always old for his level by prospect standards, McHugh put up strong strikeout numbers and did a decent job of limiting walks. Scouting purely based on the stat line is never a good idea, but McHugh seemed to have a possible future in a major league rotation based on these data. Indeed, this agrees with scouting reports published here back in 2012 (link below); not the best stuff, but good results and upside about that of Dillon Gee.

Upon his arrival to the majors, McHugh had an awesome debut, but he was then mediocre-to-awful for the rest of his time on the major league roster and was eventually designated for assignment. Here is the PITCHf/x data for McHugh’s major league innings with the Mets:

The sample size isn’t fantastic, but it’s what we have to work with. I’ve also included benchmark swinging strike percentages and an approximate expected batting average on balls in play (xBABIP). As a quick note, take the xBABIP data with a grain of salt; these values are not readily available for overall performance, let alone by pitch type, nor are up-to-date calculators (which rely on historical regression figures) readily available. Treat these numbers only as a rough estimate.

Right off the bat, McHugh’s pitch type values support his poor results. His four-seam fastball, which was his most common pitch, was well below average in terms of run values, though an unsustainable high batting average on balls in play and slightly above average swinging strike rate as a reason for optimism here. His slider, which was supposed to be his premium out pitch, delivered a below average swinging strike rate and got hammered to the tune of a .729 wOBA. His curveball was actually his best pitch in this limited sample, but even then it was buoyed by a 70 percent infield fly ball rate and was only marginally above average.

Ultimately, it doesn’t look like there are too many positive conclusions to draw from McHugh’s brief time in the majors. It’s worth noting that there is too little data for many of these statistics to stabilize, but it’s not as if there are multiple glaring examples of McHugh getting unlucky that would cause one to throw out his awful results by traditional metrics.

So what has changed to turn McHugh from a pitcher who didn't belong in the majors to one of the better starters on an up-and-coming Astros team? Below is his PITCHf/x data from this season:

McHugh’s biggest improvement is in his slider. After the pitch posted an awful –11.19 Val/C for 2012, McHugh is actually using his slider more to greater effect. He’s held batters to a .276 wOBA even with some apparent BABIP misfortune, and while the swinging-strike numbers are below the benchmark, they’re not awful. Further, he’s vastly improved his curveball in terms of swinging strikes, inducing them at more than double the rate of his 2012 season. The pitch value is similar, but the additional strikeouts are a sign of some significant improvement.

Below are the movement numbers for McHugh’s slider and his curveball:

x-Mov indicates horizontal movement from the perspective of the catcher, with positive being to the right, while z-Mov indicates vertical movement when removing the effects of gravity. With these definitions, it’s easy to see that McHugh’s pitches are substantially different. His slider is much more a flat, sweeping pitch now, and his curve added significant tilt at the cost of some vertical movement. There are no set marks for how a certain pitch should move in order to be considered good, but the improvements to the movement numbers on both pitches provide a reasonable explanation for their improved performance.

Lastly, the zone data for McHugh’s slider and curveball:

The overall numbers for the slider look very similar. However, there are some differences when one digs deeper. The pitch is inducing fewer swings outside the zone but more in the zone, and is much better at inducing swings and misses in the zone than in 2012. In a likely related result, McHugh is throwing the pitch more often in the zone. Though the contact and swinging-strike rates didn’t move much, McHugh has been able to better utilize his slider within the strike zone. The change for the curveball is similar, inducing many more swings at pitches within the zone and many more swings and misses at pitches out of the zone. These data points explain the higher swinging-strike rate.

McHugh is a fundamentally different pitcher this season. His pitches have better movement and he’s changed his usage in terms of where he throws them, showing abilities he never displayed in his time with the Mets. From a statistical standpoint, the Mets' decision to trade McHugh was sound, particularly when one considers the wealth of pitching the team has. It’s unfortunate that Young will likely never be more than a useful bench piece, but this is not a trade that should have Mets fans calling for the head of the front office’s analytics division.