This Might Be Wrong with Dallas Keuchel
Let’s boil Dallas Keuchel’s 2017 down to the most simple terms. Imagine yourself in the ideal green pasture; breeze lapping off your face as your emotions float above cloud nine. Then you trip; fall 10,000 feet into a brown pasture, with blistering winds making every step all the more uncomfortable. Not even Jeff Luhnow can save you from this unfortunate turn of events.
Through his first 11 starts of 2017, Keuchel posted a 1.67 ERA with a 3.11 FIP, consistently outperforming his peripherals with the ability to limit productive contact and generate swings out of the zone unlike any other. If we qualify first-half statistics with pitchers who pitched more than 70 innings instead of the qualified 95, Keuchel shows up atop the groundball rate leaderboard (a whopping 67.4%), with a cameo in the 75th percentile for out of zone swing percentage (31.8%). Uncomfortable at bats is a good way to describe the implications of those statistics. In case it wasn’t obvious, this is the green pasture.
In the minute sample size of two starts, or eight innings, since coming back from the disabled list with a pinched nerve in his neck, Keuchel has lost his touch. A 7.02 FIP, with a more manageable 5.53 xFIP informs us that he has gotten a bit unlucky with the home run ball, but the most disheartening is the 11% drop in ground ball rate along with a 6% cut in his out of zone swing rate. A lot of these stats aren’t representative of where Keuchel will definitely trend from here on out – batted ball data needs about a month or two of info – but there aren’t many things less encouraging than back-to-back starts of mediocrity fresh off the disabled list.
Speculating in a very rudimentary sense, one might think that Keuchel has changed up his pitch mix. Henry [Insert Last Name Here], a community contributor on Fangraphs, broke down the four pitches (predominantly) that this lefty throws and while it’s unique, not even in this small of a sample is there reason to think Keuchel has veered from his set of offerings after looking at data from Keuchel’s first 11 starts and subsequent two. Versus lefties, Keuchel relies on his sinker and slider (combined 93% usage) to hold bats to a .189 wOBA. For righties, Keuchel mixes it up by incorporating a changeup and cutter, along with the signature sinker-slider duo. The consistency between both types of hitters he faces is that if you let him get ahead, there’s a good chance you’re seeing a slider. While he isn’t as dominant versus right-handed bats, a slugging percentage of .329 is objectively fantastic when facing batters on your “worse” side of the plate.
So with this quick – and hopefully temporary – deterioration, what’s the issue?
I mentioned that Keuchel has an uncanny ability to live outside of the strikezone – an in-zone percentage of 36% is the lowest among pitchers with 70+ innings of work in the first half – and while that hasn’t changed, with only 33% of pitches in the zone in those more recent two starts, maybe Keuchel is locating less effectively out of the zone or he hasn’t gotten back the “feel” for pitches since hitting the DL.
The location argument is interesting (we’ll get to the “feel” argument later). Here is a heatmap of all 1,000+ pitches Keuchel threw in those first 11 starts, next to another map of those he threw in his last two outings.
The region with the wider “blob” of red represents his last two starts. You’ll notice particularly that instead of concentration down in the zone, Keuchel is kind of all over the place; living in the middle of the plate like he rarely tends to do. Also notice the most concentrated part of the zone – darkest red – has moved from low-and-in to lefties from the whole bottom third of the plate. This is driven most by the lapse in command of his sinker, shown below.
Once again, the wider blob are sinkers in Keuchel’s most recent two starts, while I could probably write a novel on how great his control was in the first half. This deviation from immaculate control early this season is partially due to the movement on his sinker changing. Since Keuchel’s most recent two starts were made in late July and early August, BrooksBaseball.net has already sorted out the vertical and horizontal movements of the pitch based on those starts.
In both graphs, where the line stands out in terms of movement – the most recent two months – is when Keuchel has been less effective. His all-powerful sinker has lost some of its panache, as it’s horizontal bite has dropped (less arm-side run), while the depth of the pitch has increased. This might be making it easier to read the spin on the pitch as a hitter, especially as the pitches movement trends towards that of his changeup, possibly blurring the difference between the two. In this two-start sample, the most notable deterioration is Keuchel’s effectiveness against lefties, as that side of the plate has amassed an uncharacteristic .444 average against – again, we’re working with a small sample here. But, if anything was proof of Keuchel’s sinker being “off” this would be the holy grail over a greater sample size. His elite two-pitch mix versus lefties is currently bats beating up a “different” sinker and not even allowing the lefty to get to his out-pitch; the slider.
Speaking of that slider, the most recent 36 iterations of the pitch have also meandered from his first-half dominance. This time producing an odd dual-concentration on the graph.
Keuchel might be overthrowing this – assumed by the fact that he’s not locating it at the bottom of the strike zone like he did before hitting the disabled list, but pulling it through the zone to his glove side. This “double concentration” will likely be smoothed out once we receive a larger sample from Kecuhel’s starts, but the second blot of red, further out of the zone, might be one of the reasons his out-of-zone swing rate has dropped so much; he’s likely not fooling any batters with a pitch that far out of the zone.
I’m sure Keuchel wishes I knew, but unfortunately, lapses like this aren’t always easy to hone in on. It could be mechanical – although my untrained scouting eyes didn’t catch any staunch difference between starts before and after the injury when combing through some MLB.com video.
On top of that the only real release point difference I found was an uptick in the vertical release point on his sinker in his last start (6.29ft versus his early season average around 6.18ft – BrooksBaseball). It could also be something as simple as repetition and feel. I enjoyed immensely an interview that Eno Sarris had with Andrew McCutchen, where the Pirates’ outfielder cited “feel” and “preparation” more than any of the more advanced metrics analysts punt to when trying to discern what the issue is with a hitter. Keuchel hadn’t thrown a pitch at the major league level for more than a month. The elephant in the room is that Keuchel could still be dealing with the neck issue that sidelined him for the majority of summer. A combination of the latter two points seems like the most obvious reasoning; the injury threw of Keuchel’s “feel” for his pitches, resulting in some inconsistent results for his most relied-upon pitch, the sinker.
I’m skeptical that a pitcher with such a level of unique effectiveness only a few months ago can fall-off so quickly. The Astros need Keuchel for the postseason, and betting that Keuchel figures out his issues sooner than later is an astute move. And if you actually want to bet on such success, check out what the SportsBettingExperts have to say about such speculation. Luhnow’s pivot to add another lefty in Francisco Liriano isn’t a bad move per say, but Liriano wasn’t, and will never be a viable Keuchel replacement. 2015’s Cy Young winner is an irreplaceable animal, and one of the most entertaining pitchers in the game to watch.
Photo via the Flickr Creative Commons, thanks to Keith Allison for the shot of Keuchel.
Statistics all from Fangraphs and Baseball Savant, unless otherwise noted.’s