“We both know that Dallas Keuchel is an ace, right?”
“We both know that Dallas Keuchel is an ace.” Or so my potential trading partner claimed. The message in question was in regards to a trade-block entry I submitted indicating my interest in making one last blockbuster move towards a title push. I was offering what I considered to be a plum set of prospects, Tyler Glasnow and Rafael Devers, plus, if necessary my early first-round pick in which all 2015 Rule 4 draftees were first eligible. My request was a legitimate ace and a quality closer. I had hoped to acquire a Jose Fernandez, Jacob deGrom, Stephen Strasburg- type but had not received the interest I anticipated. So I began to wonder, is Dallas Keuchel an ace? My instinct is to be suspicious of breakouts without solid evidence of skill growth. So, did Dallas Keuchel grow into an ace or was he just the recipient of a lucky season?
Keuchel broke into the majors in 2013 and showed no indication that he was ready to be a frontline starter. Owners that were fortunate enough to add him in early 2014 after a fast start were likely pleasantly surprised as he finished the season with an ERA of 2.93, down from a ghastly 5.15 in 2013, driven by a rebound in BABIP, which dropped from a ridiculously high .350 to a near league average of .295. Owners who believed in his breakout were richly rewarded as his stats from 2015 were absolutely worthy of being considered ace-calibur. He was the fifth overall pitcher on ESPN’s Player Rater. His strongest category was wins, a notoriously fickle statistic, but he was extremely good across the board. The increase in performance seemed to be driven by two factors, strikeout rate and BABIP.
BABIP is the easy factor to analyze. Among qualifying pitchers from 2012-2015, the average BABIP was very consistent at .290. Year by year there is very little deviation. Keuchel, however, experienced a large drop in BABIP from .295 to .269. He will almost certainly regress back towards the league average, eliminating some of his apparent gains.
The trickier factor is his strikeout rate. Any number of factors can lead to an increased strikeout rate including but not limited to velocity increase, pitch location, pitch sequencing, pitch movement, and new pitches. In Keuchel’s case, the increased strikeout rate was an effect of a significantly higher swing-and-miss rate that was driven entirely by his changeup. The swinging strike percentage on the pitch jumped from 15 percent to over 20 percent, a 33 percent increase. No other pitch showed any significant improvement.
I compiled data from all pitchers who pitched a qualifying number of innings from 2012-2015. I found 13 seasons in which a pitcher increased his strikeout rate by 20 percent or more. One pitcher was able to carry that increase to the next season. That pitcher was Clayton Kershaw. Most pitchers experienced a significant regression that canceled between 50-75 percent of their apparent gains. Keuchel’s rate jumped by a whopping 27 percent. So the question is likely not whether or not his gains are real, but rather how much of the gain is real. The analysis below tries to take a look at that.
The table below shows the velocity of Keuchel’s pitching arsenal for 2014 and 2015.
His fastball velocity is presented as ranges with error bars below. As you can see, if anything, his fastball tended to be a tick below his 2014 level. It seems safe to say that an increase in fastball velocity or decrease in offspeed velocity was not the cause of his increased strikeout rate.
The movement on his pitches did not significantly vary from 2014 to 2015, which is to be expected. It’s not easy to change what a pitch does after it leaves you hand. The chart below shows his horizontal and vertical pitch movement. It doesn’t appear that there is much difference to be seen here, either. Perhaps there was a bit of an improvement in plane change as his fastball and changeup stayed higher and his slider broke downward a bit more. I found the changeup data to be counterintuitive as I assumed more downward break would correspond with more swings and misses.
Next I considered the location of Keuchel’s pitches. His pitch location suggests a reason why batters swung and missed at an increased rate. From 2014 to 2015, he was able to relocate the majority of his changeups from middle-down to low-and-outside. With no change in the percentage of pitches swung at, it’s not surprising that batters struggled against a pitch that was generally outside the strike zone. This offers hope that his strikeout rate can be maintained, as long as hitters keep offering at near strikes.
Based on a detailed look at the inner workings of Keuchel’s season, I believe it’s safe to say that he was a more effective pitcher in 2015. That is, it was more than luck that led to his breakout. That being said, he will not repeat a BABIP of .269 and despite being quantitatively supported by pitchfx, he is likely to experience some regression in strikeout rate. Both Steamer and ZIPS have projected a strikeout rate for Keuchel that splits the difference between his 6.5 and 8.4 per nine innings. That seems about right to me. I would anticipate Keuchel being a better pitcher than he was in 2014 but not as good as in 2015. Keeping in mind that even his elevated rate of 8.4 per nine was far from elite, it seems unlikely that Keuchel can be counted on as an ace. Based on his projected statistics, I have him ranked as the 21st starting pitcher, in between Gerrit Cole and Chris Archer. Certainly very good, but not enough to carry your staff.