A few weeks ago I had a brief discussion with MLB Scout Jim Stevenson regarding Astros pitcher Dallas Keuchel. Jim posed the question: “Is this who he is? Or was this just one of those special years?” I think that most people around baseball would argue that this is probably “one of those special years”. I’d be inclined to agree with those people. However, Jim and I both agreed that Keuchel has the stuff to be a good pitcher for a long time in the MLB. Probably not a perennial Cy Young candidate, but definitely someone who could throw 200 innings and have an ERA around 3.00 for the next 6-8 years.
This conversation prompted me to take a close look at what Keuchel has done so far in his 3+ years in the majors. I discovered a pretty interesting trend, which I put into graph form and shared with the Twitter world.
Pretty astounding. Keuchel’s IP have increased dramatically each year since 2013 (153.2, 200.0, 232.0), but his hits allowed (184, 187, 185) and walks (52, 48, 51) have remained relatively constant. Obviously then, we know that his H/9 and BB/9 (and therefore his WHIP) have all decreased dramatically in that same time span. But you want to know more. What about his HR/9 and K/9? What about his actual results? What about his debut half-season in 2012? That’s a lot to ask of me, but I think I have something to satisfy all of those questions. Behold:
It’s magnificent, isn’t it? If you combine his first season and a half into one (which still only covers 239.0 IP, just seven more than he had in 2015), Keuchel has improved every one of these rate stats every year of his career (with the exception of his HR/9 from ’14-’15, which went from a miniscule 0.4 to a still-below-average 0.7). That’s pretty impressive, and clearly the results have followed, not only within the scope of ERA but also the predictors. But how did he do it? I’ve identified what I believe to be the top 3 factors in his progression: improved pitch mix, improved location, and improved secondary pitches.
When Keuchel came into the majors in 2012, his repertoire looked a bit different than it did now. He still featured a four-seam, sinker, change-up, and cutter, but he was throwing a curveball instead of a slider. However, his curve wasn’t very good, so in 2013 he decided to start throwing the slider. After 2013, the curveball was scrapped, and the slider became the primary breaking ball. This was definitely a positive change, because in 2015 his slider became his second-best pitch behind his sinker. Speaking of the sinker, he made some changes to that as well. From 2012-2013, he threw the sinker on about 39% of his pitches. For 2014-2015, that number increased to 48%, a substantial increase. The jump came as the expense of his four-seam and cutter, which are relatively mediocre pitches anyway. After he started throwing the sinker more, it’s no coincidence that his groundball rates jumped from the low-mid 50’s to mid-upper 60’s between 2013 and 2014. This provided him with three key benefits – decrease in XBH, increase in GDP (no, not the one from economics class) and decrease in home run rate. I put together a visual of those three benefits, to demonstrate the significance and immediacy of the change (if you’ve never heard of DP%, that’s fine, I made it up about 15 minutes ago – it’s the number of DP turned behind the pitcher divided by the amount of singles, walks, and HBP he allowed).
Once again, the numbers are very impressive. Conventional wisdom states that “mixing your pitches” is a key to success, but Keuchel did just the opposite. Instead of trying to improve his secondary fastballs (four-seam, cutter) he but them on the back burner in favor of his great sinker, and it paid big dividends.
Next is the location of Keuchel’s pitches. There’s not much I can say that makes sense without showing a visual of his location, so let’s have a look courtesy of Baseball Savant:
So the first two things that stand out to me are the downward and inward (to a RHH) expansion of the strike zone. Keuchel is left-handed, and doesn’t throw hard, so as a young pitcher he was likely not confident in his ability to throw in on the hands of righties. However, as he’s gotten more comfortable and become a better pitcher, he seems to have become more confident throwing in. He also has an asset that works especially well in that area:
This is a look at the sliders Keuchel has thrown in 2014 and 2015, and explains why there’s more pitches in the righties – he uses the pitch almost exclusively in the third quadrant of the strike zone. In addition to the inside pitches, we can see that he’s thrown more pitches low and even out of the strike zone in the past two years, which is likely a product of throwing more sinkers. In fact, Keuchel has thrown fewer than 40% of his pitches in the actual strike zone since 2014, thanks to his great “chase ” pitches – the sinker, slider, and changeup. Which brings me to my last topic – improved secondary pitches.
As I mentioned before, Keuchel used to feature a curveball instead of a slider, which wasn’t very good. Getting rid of the curveball and picking up the slider was the first step to making Keuchel a guy with plus secondary stuff, but it wasn’t a quick fix. According to FanGraphs’ pitch values, Keuchel’s slider was actually a very average pitch in 2013 and 2014, accruing values of -0.7 and -0.3, respectively (0 is average, the number represents runs above average for the pitch). His changeup was even worse, racking up values of -3.0 and -4.2 in 2012 and 2013. However, in 2014, his changeup improved to +2.5, and in 2015 the slider followed suit, valued at 18.0 runs above average, and his changeup again improved to +7.1. How he made those improvements is important, and I think I’ve got an idea how he did it. His changeup averaged 77.2 MPH from ’12-’13, which isn’t bad for a guy who throws 89 MPH fastballs. However, the last two years it has been more than two and a half MPH harder, at 79.8. This may seem counterintuitive – a changeup is supposed to be slow. However, by closing the gap between the velocity of his fastball and his changeup (his fastball velocity has been pretty constant over his career) may have made it more deceptive. With the slider, Keuchel went a bit different route – instead improving by changing the shape of the pitch, not the velocity. Keuchel’s slider velo actually dipped a bit from 80.8 to 80.3, but it began to move differently. From 2013 to 2014, the slider had an average of -0.46 inches of vertical movement and -4.60 inches of horizontal movement, according to PITCHf/x and Brooks Baseball (if you’re not familiar with how PITCHf/x movement measurements work, take a look here, because it’s a bit complicated). In 2015, those numbers have been -2.14 and -5.78, respectively. This means that Keuchel’s slider has increased both its depth and run since he began throwing it. He also changed where he throws the pitch most:
Keuchel harnessed the extra movement on his slider by using it more often out of the zone, which we have already identified as one of Keuchel’s strengths – getting hitters to chase. To sum up what these changes have done for Keuchel, here’s a look at the effectiveness of each of the two secondary pitches before and after their respective improvement:
So, we’ve identified what Keuchel has done to make himself the Cy Young Award winner he was in 2015 – but does it change how we answered the original question? Personally, I think that some regression toward the mean is inevitable, but discovering all these definitive improvements that are very easy to identify and link to improved success bodes well for Keuchel. Clearly, he knows how to make adjustments and is good at identifying what needs to be adjusted. As I said before, I’m not convinced he’s a perennial Cy Young candidate, but I wouldn’t be surprised to see him get into the discussion a couple more times and make a lot of All-Star games. With the way the Astros are trending now, I think their management is very happy to have this man at the front of their rotation for at least the next three years.