Posted by: Peter | May 9, 2008

Has Carlos Gomez turned a corner?

This blog exists because I learned how to extract pitchf/x data from the Major League Baseball website. So I’m starting with some pitch analysis!

After he hit for the cycle the other day, everyone started talking about how Carlos Gomez had turned his season around after a one-game benching on April 23rd. And he certainly looks better! Dave Studeman notes that “Gomez was hitting .230/.247/.310 when he was benched for a game on April 23, and has gone 13-for-30 (.433) with two homers and four steals in eight games since.”

But has Gomez really changed his approach, or is he just getting lucky? The problem he had early in the season was his tendency to swing at anything, especially breaking balls way off the plate. Aaron Gleeman showed recently that Gomez leads the Twins in swinging at pitches out of the strike zone (although the rest of the team ain’t exactly showing a Ted Williams eye at the plate either.) So I figured it would make sense to look at the pitch-location data from before and after April 23rd, to see whether Gomez has gotten any savvier about laying off junk pitches.

I downloaded the data for every pitch Gomez has seen, up through the May 8th loss to the White Sox. Then I divided the data into the games before April 23rd, and the games after. Gomez saw 307 pitches before his benching, and 119 after it, so the sample size, while not huge, is big enough for some exploratory analysis.

Below are two graphs showing the before-and-after versions of Gomez at the plate. Each pitch is shown where it crossed the plate (according to pitchf/x). The view is from behind home plate, looking at the pitcher–so the points on the right are away from the right-handed Gomez. The points are color-coded according the the outcome of the pitch: green points were called balls, red points were strikes (either called, swinging, or fouled off), and black points were put into play (for either a hit or an out).

So what do we see in these plots? Well, the plot above confirms that Gomez was swinging at a lot of bad pitches. All those red dots way out in the lower right are unhittable balls that Go-Go went hacking at. If I were a pitching coach looking at this graph, I would tell my pitchers to pitch Gomez low and away, and nothing but low and away.

OK, so what about the games after April 23rd? Well, it looks a heck of a lot better! In the graph above, there are are hardly any of those red dots low and away out of the strike zone. Gomez is still no Jason Giambi, but he seems to have learned to recognize pitches a little bit better. From this plot, his weakness now looks to be balls up and out of the strike zone, rather than low and away.

Now, the graphs above don’t tell us what kind of pitches those were. And the consensus seems to be that like a lot of young hitters, Gomez has particular problems with breaking balls and changeups. Some people have even suggested that he may be permanently cursed with poor pitch recognition. So let’s break down the analysis a little to see if this true.

Below, I have once again graphed pitch locations, before and after April 23rd. This time, I have included only those pitches that were either swinging strikes or fouls. These are the pitches that Gomez shouldn’t be swinging at, because he can’t make solid contact with them. I have color-coded the pitches according to the type of pitch. This classification is according to Major League Baseball’s system, and it is included in the pitchf/x data. These classifications aren’t perfect, and others have suggested better classification systems. But for my purposes, the MLB classifications are close enough, because they do a pretty good job of distinguishing fastballs from changeups and curveballs.

OK, let’s look at the graphs again. (A few pitches are missing in the first graph because they didn’t get classified by pitchf/x–see below for more on those.)

Again, the pattern for pre-April 23rd Gomez is striking. Those low-and-away pitches he was hacking at were almost all offspeed pitches, not surprisingly. It should be noted that that wasn’t his only problem, though–he was missing or fouling off a lot of fastballs, too, including some that were down the middle of the plate.

And after the 23rd? Again, it looks like there’s been a big change. Gomez is getting fooled on outside breaking balls far less often.

To get a better sense of what’s going on, let’s calculate some simple percentages. There are a lot of numbers below, but the conclusion is interesting–so bear with me!

OK, first off, is Gomez making better contact than he did before? Here’s how often he was missing or fouling off pitches in each period:

As a percentage of total pitches seen:
                       Swinging strikes         Fouls
  before 4/23              13.7%                22.5%
  after 4/23               13.4%                23.5%

OK, looks like that’s not what’s up. But what about the improved discipline we saw in the graph? Running the numbers does seem to suggest that Gomez has really improved:

Offspeed pitches only (Changeups, curveballs, sliders):
                      Swinging strikes         Fouls
  before 4/23              16.4%               24.6%
  after 4/23               18.9%               13.5%

It’s a mixed picure–he’s actually swung through a higher percentage of breaking balls in the second time period, but has fouled off way fewer. But remember that the overall swinging strike and foul percentages haven’t changed. So maybe Gomez is just missing lots of fastballs instead?

Fastballs only (including sinkers):
                      Swinging strikes        Fouls
  before 4/23             9.7%                20.6%
  after 4/23              11.0%               28.0%

That does seem to be part of it. He’s fouling off more fastballs, by quite a bit, and swinging through more of them too. In addition, Gomez’s improvement on offspeed pitches has been masked because he isn’t seeing as many of them now:

Pitch distribution:
                Fastballs     Offspeed    Unclassified
  before 4/23      50.5%       43.6%          5.9%
  after 4/23       68.9%       31.1%

Now, these numbers might be biased, because about 6% of the pitches in the pre-April 23rd dataset didn’t get classified by pitchf/x–so we don’t know what kinds of pitches they were. But even if all those unclassified pitches were fastballs, we would still conclude that Gomez is getting a lot more fastballs than he did before.

Now, why would this be? You have to imagine that the scouting reports on Gomez after the first few weeks emphasized his weakness for offspeed pitches. So if pitchers are nevertheless throwing him more fastballs, that suggests it’s because he’s no longer biting on the offspeed stuff. In other words, getting more fastballs is a result of Gomez’s improved performance, not the cause of it. But the numbers suggest that maybe Gomez hasn’t completely been cured of his tendency to whiff at breaking balls, so if he starts to see more of them again he could start to struggle.

So what’s the conclusion? With only a couple of weeks of the “improved” Gomez to go on, we can’t say anything for sure. But it certainly appears that he has shown some all-around improvement as a hitter. My quick summary would be:

  1. Gomez is swinging at fewer offspeed pitches out of the strike zone.
  2. Probably because of (1), when he does swing Gomez is not as likely to foul the ball off or miss it entirely.
  3. Also probably because of (1), pitchers are giving Gomez more fastballs.

Given all this, I’m cautiously optimistic about Carlos’s future. There is certainly some reason to be concerned that this new level of performance won’t hold up. For one thing, his improved performance against offspeed pitches has been pretty dramatic, and I’m a bit skeptical that he’ll be able to sustain it. I expect to see some regression in the coming months–but hopefully he won’t regress all the way back to his early-April form.

I’ll revisit this question later, maybe around the all-star break. By then, we should be able to get a better idea about whether these results are a fluke, or whether they indicate a real step forward for Carlos Gomez.



  1. […] have any particular point to make with these. This isn’t an analytical post (unlike the Gomez post). This is more along the lines of, “I figured out how to get all this neat pitch data, so why […]

  2. […] 10, 2008 · No Comments I had already finished and posted my Carlos Gomez analysis when the G-man went ahead and became the winning run in tonight’s exciting game. So of course […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: