The Houston Astros have been talked about lately as a very corrupt organization with their acts of sign-stealing during the 2017 season - and postseason more specifically. However, the player I’m going to talk about today wasn’t a part of that team, but was a result of the Astros advanced analytics group that seemingly can change any player into a stud. Gerrit Cole was coming off an average season with the Pirates, who were the polar opposite of Houston analytically as they’d mold pitchers to get groundball outs. When the Astros got Cole, they knew that he was not a groundball pitcher, he had nasty stuff and could strike guys out. I will explain what changes the Astros made to Cole and how that translated to better results.
In Pittsburgh, Cole wasn’t playing to his strengths. He was a fireballer and had amazing movement on his pitches, but he was treated as if he needed to get the ball on the ground instead of swings and misses. The sinker was a huge part of what Pittsburgh preached as he used it 11.1% of the time there, which had a bad wOBA against of .325. So when he left, the Astros wanted to cut it out completely and he hardly ever used it again. With that gone, he was able to use his nasty slider and curveball more often, throwing them 3.4% more and 7.4% more respectively (via Baseball Savant).
Not only did he start using those better pitches more, he also improved those pitches through better pitch movement. His fastball, slider, and curveball all changed significantly, as shown from the charts by Brooks Baseball below:
In Pittsburgh (on the left), his slider dove down and, at times, almost looked like his curveball, his fastball’s movement profile also did the same thing with his changeup. This means that hitters were really seeing two different movement profiles even though he had four different speeds. However, when the Astros (right) management team changed him, you can notice that there is an easy differentiation between all of his pitches, as none of his pitches mesh together (besides the changeup and sinker, but he didn’t throw his sinker much in Houston). Hitters were not able to figure out where each ball was going and so they swung and missed more often. Cole also found a way to make his curveball and fastball add more depth because he added more spin to each pitch, with about 300 RPMs added to both his curveball and fastball.
The way pitch movement sees change is the spin axis of the ball, which is the direction spin as the ball is traveling towards the batter. A fastball, for example, travels up compared to a curveball that travels down, so they should have about an opposite spin axis of each other. I will go more in-depth about that in another post. But when you see these graphs, from Brooks Baseball again, you will notice a similarity between them and the movement graphs:
The Y-axis of these graphs is the vertical release point, just to keep the spin axis points close to each other. The similarity of these graphs is that when he was in Pittsburgh (left), his spin axes were blobbed together for his slider and curveball and for his fastball and changeup. This shows that the Astros (right) changed the spin axis of his pitches to change the way they moved. That lead to the differentiation between his pitches and he found a way to confuse hitters.
Another way that made it hard to tell the difference between Cole’s pitches was that he was able to tunnel effectively. Tunneling is throwing pitches in the same arm slot consistently to not give away what pitch is being thrown if there is a difference. The difference between his arm slot from Pittsburgh to Houston is pretty noticeable, also depicted on graphs from Brooks Baseball:
The two blobs of dots shown both have different sizes, showing how wide and tall each arm slots are. In Pittsburgh (left), the blob looks decently large and you are able to depict the different pitches that he is throwing, while for the Astros (right) it is smaller than before and this tight blob makes it hard for batters to read pitches coming out of the hand.
You probably know what happened after all of these changes, he became one of the best pitchers in baseball and was the runner-up for the Cy Young last season behind teammate Justin Verlander. Also, many of his stats improved. His ERA in 2017, 4.26, dropped to 2.88 and 2.50 in 2018 and 2019 respectively. He didn’t strike very many hitters out in a Pirates uniform with a 19.4% and a 23.1% in his final two years. In Houston, though, he turned into the pitcher they wanted him to be, striking out 34.5% and 39.9% of hitters in his two great years in Astro orange (via Fangraphs).
With Cole leaving a controversial team to going to an all-time powerhouse, the Yankees, expect much of the same from him when he dons pinstripes later this month.
Comments