Fluke Watch – Dillon Gee

Dillon Gee is not a pitcher on most people’s radar. Considered at best a fifth starter throughout the minors and the last two years in the majors, he’s one of the many pitchers who exist in real baseball but practically do not exist in fantasy unless someone is in a ridiculously large NL league. And if you just look at his ERA this year—well you’d think nothing has changed.

But if you look at xFIP—which should in theory be a leading indicator of a pitcher’s future performance*—things look different. In fact, Gee is actually in the top 10 in the majors in xFIP at 3.16! Does this mean this so-called No. 5 starter is likely to break out any time soon? Could he be a buy-low candidate? Well let’s find out.

*To an extent – it should indicate where a pitcher is getting lucky/unlucky and should regress, but it obviously can’t predict whether a pitcher can keep up his peripherals in the first place. Of course that’s what this type of article aims to figure out.

Gee’s pitches

Gee has thrown six pitches over the last two years in the majors:
{exp:list_maker}A four-seam fastball that averages 90 mph with totally average movement.
A two-seam fastball with decent tail but and only decent sink (but the sink compared to the four-seamer is nice).
A change-up that has traditionally been Gee’s best pitch and go-to out pitch, which averages 82 mph but has nice sink and drop compared to the fastball.
A curveball with a large amount of drop and horizontal movement (essentially an 11-5 curve).
A slider at around 80-82 mph with not special movement
A cutter at around 86-87 mph with nice sink. {/exp:list_maker}

Now, Gee dropped his slider in 2011 around June in favor of the cutter, but he appears to have switched back to the slider this year. (It’s hard to tell at times if he is still throwing the cutter occasionally.) Otherwise, his pitches’ movement and velocity have remained the same this year as compared to last year. His improved peripherals are not explained by suddenly found movement or velocity. Nor are they explained by pitch location: Gee’s basically aiming each of his pitches in the same general areas as they were aimed at last year. No change here is responsible for his improvement.

Gee’s pitch usage

Gee is a right-handed pitcher, so he faces an even mix of left and right handed batters. And his improvements seem to have come against both types of batters.

Against lefties, in 2011, Gee relied heavily on his change-up, throwing the pitch 28.8 percent of the time. Gee also threw an even mix of his fastballs (28 percent two-seamers, 25 percent four-seamers) to lefties. He did not use his two breaking pitches often (counting the cutter as a breaking pitch) often. The cutter was used only nine percent of the time, while the curve was used only 8.6 percent. Before Gee switched from the slider to his cutter in 2011, Gee never used his slider against lefties, throwing only three sliders to lefties in April-May.

In 2012, Gee’s pitch distribution has changed slightly. Change-up usage is down around three percent, which could simply be a sample size fluke. In its place, Gee has relied more on his curve, throwing the pitch 13.8 percent of the time against these batters. The curve is still the fourth most used pitch in his arsenal against lefties, but it’s more of a factor. In addition, despite Gee losing his cutter this year, Gee has been willing to throw his slider around 7.6 percent of the time against these batters.

These changes can also be seen on a count basis: In 2011, the change-up was clearly Gee’s out pitch. He’d use it frequently in every count, but in two-strike counts the change-up usage would rocket upward. That made sense, as it was a good swing-and-whiff pitch. This year, the change-up usage has actually declined on two- strike counts compared to other counts. Instead, Gee is relying on his four-seam fastball more as an out-pitch and is also using both of his breaking balls in this role.

Against righties, in 2011 Gee again relied mainly on fastballs and change-ups. The fastballs, again pretty evenly split in usage, were used a combine 58 percent of the time, with a change-up coming around 21 percent of the time. The remaining 21 percent was split evenly between the curveball and the slider/cutter.

In 2012, Gee, again, Gee’s pitch usage has changed. The biggest change has been the monster increase in slider usage—Gee went from using the slider/cutter 10 pdercent of the time in 2011, to 22 percent in 2012 (all the slider now). It’s now Gee’s primary off-speed pitch, having surpassed the change-up. Gee’s fastball usage has dropped to being used less than 50 percent of the time, with most of that drop going to the slider. Gee’s change-up usage has also decreased to 16 percent.

This can again be seen in how he uses his pitches in each count. Previously in 2011, Gee would use his fastball mainly, or his change-up as a second option, to get outs on two-strike counts. In 2012, the slider has really taken a place as a real out pitch.

Gee’s results

Okay, so we now know a clear change has occurred in Gee’s pitch usage. Do Gee’s individual pitch results make sense given that change?

First, Gee’s change-ups have gotten basically the same exact results as last year: 19 pdercent swinging strike rate, with a 53-57 percent groundball rate. Those are pretty darn good numbers. But remember, Gee’s improvement has come as he’s actually decreased his change-up usage.

Instead, Gee’s improvement has come from basically every other pitch. His slider, now used at an increased rate and as an out-pitch against righties, has clearly received improved results. Last year, the cutter got swinging strikes nine percent of the time; this year it’s 14.2 percent. Last year, the pitch got ground balls 42 percent of the time; this year the slider’s at 58.3. Using the pitch on two-strike counts, particularly 0-2 and 1-2 counts, has resulted in batters being extremely confused.

Gee’s curve is also receiving massively improved results from last year, when it was a negative pitch. The pitch now gets a swinging strike 12.5 percent of the time, up from 9.5, and gets 64.7 percent ground balls, up from 29.6 in 2011. That’s a huge difference.

Gee’s fastballs are also both improved slightly.

It’s hard to say if these improved results are sustainable?, since they come from, if anywhere, the interaction between Gee’s own pitches. This is essentially game theory, and it’s incredibly hard to predict. Looking individually at pitches for example:

The curveball’s groundball rate seems destined to regress as a fluke of small sample size (only 17 balls in play) as the pitch has remained exactly the same. On the other hand, the increase in ground balls could be explained by the fact that the slider’s presence might make batters swing over the top of the curveball more. The slider’s improvement could be from being used just on two-strike counts to surprise batters, but one might think this improvement therefore wouldn’t last as batters get updated scouting reports.

Conclusion

Overall, it’s really hard to know whether Gee’s xFIP and peripherals are a good indicator of a good fantasy option in the future. There are no obvious improvements in his stuff that would cause his improved performance, but his change in pitch usage may explain the improved performance. This is a gray area in this method of PITCHf/x analysis—we can tell there’s a change, but we can’t nail down the effects of that change.

As a result, I’d probably recommend staying away from Gee for now (which you’re probably already doing). There’s no clear sign pointing toward him keeping up these improved peripherals and he is on a team with poor power, the Mets, that won’t net you many wins.