On Friday, Kevin Gausman started for the Baltimore Orioles. He continued what has been a largely splendid season, as he limited the Seattle Mariners to one earned run over 6.2 innings. While another obligation prevented me from watching, I did take a gander at Twitter, to see this humorous observation:
The pointlessness of a pitcher’s win-loss record notwithstanding, this inconsistency struck me as odd. I decided to look into it, to see if Gausman’s peripherals were this fickle.
As it turns out, they are. Gausman has started nine games this year, with an overall SIERA of 4.44; if he had enough innings, that would rank 84th out of 93 qualifiers. In other words, he hasn’t pitched that well overall. However, on a start-by-start basis, I saw something interesting: He has never come close to that figure in any one outing.
In four of Gausman’s starts — June 18th against the Tampa Bay Rays, July 20th and June 7th against the Oakland Athletics, and July 6th against the Boston Red Sox — his SIERA was at or below 3.40. That mark would rank 26th in the majors, and would be better than the likes of Adam Wainwright and Julio Teheran. By contrast, in the other five starts — July 13th against the New York Yankees, July 25th (Friday) against the Mariners, June 12th against the Toronto Blue Jays, May 14th against the Detroit Tigers, and June 27th against the Rays — Gausman’s SIERA was at or above 4.97. That mark would rank 91st in the majors, and would be worse than the likes of Travis Wood and Eric Stults.
What has driven these dichotomous results? Primarily, strikeouts. In the four good games, Gausman fanned at least 21.7% of the batters he faced; he didn’t top 11.1% in any of the other five. His K%s and SIERAs have an R^2 of .916, indicating a very strong correlation. While his walk, ground ball, and fly ball rates correlate decently to his SIERA (with R^2s of .308, .279, and .259, respectively), we should attribute most of his SIERA variability to punchouts.
Let’s shift our focus to those, then. We can take them seriously, as Mike Podhorzer’s xK% equation proves their legitimacy. Indeed, the relationship between his xK%s and his SIERAs is actually stronger, with a correlation coefficient of .924. Furthermore, correlating the individual components to his strikeout rates, we see that S/Str% (the percentage of strikes that are swinging) accounts for most of the K% variability, as its R^2 sits at .454.
So, to summarize all of that: Gausman has had four superb games, and five terrible games. He derived most of his success from strikeouts, and he derived most of his strikeouts from whiffs. To determine the cause of the latter, we’ll delve into PITCHf/x data.
For the year as a whole, Gausman has accrued most of his whiffs on two pitches: the splitter, and the four-seamer, which account for 36 and 33, respectively, of his 81 total swings-and-misses. The former excels via its quality (it has a 24.5% swinging-strike rate against); the latter, via its quantity (he throws it 69.2% of the time). Location-wise, he generally has the most success with the splitter when he paints the lower left portion of he plate with it…
…while his four-seamer does well when it goes to the upper middle part of the plate:
If we break this down on a start-by-start basis, a pattern emerges. In the five bad starts, he did one of a few things wrong: He didn’t utilize the splitter enough; he didn’t locate it properly when he did; or, if he did the former two, he didn’t locate the four-seamer properly.
Let’s go through on a start-by-start basis — starting with the poor ones:
- In the Detroit start, he threw only seven splitters (8.1% of his total pitches), and all to the lower right of the plate.
- He made a similar error in the New York game, as he used it a mere five times (6.4%), and put it in the middle-to-upper left of the plate when he did.
- Pitching in Toronto, he implemented it 10 times (9.9%), but only four of those were in the bottom left.
- His second outing in Tampa actually saw him use it even more, at 21.1%; plus, most of those were in the lower left. However, he placed the four-seam fastball in the middle left, for whatever reason.
- The Seattle game was even worse: While he fulfilled the first two conditions handily, he threw the four-seamer seemingly everywhere but up.
Now, let’s look at the splendid ones:
- In his first bout in Tampa, he pounded the lower left with the splitter, and did so often (26.1% of his pitches).
- His second Oakland game was virtually analogous, with a 23.9% splitter rate to accompany a heavy focus on the lower left.
- The first time around the bay had him using it less (15.9%), but eschewing the top and middle, as well as the right and center.
- While in Boston, he threw it less too (14.4%) and drifted toward the middle left when he did; nevertheless, a top-heavy approach with the four-seamer bailed him out.
So the method to Gausman’s madness thus far? Throw the splitter to the lower left, and good things happen; don’t, and bad things happen. It might sound overly simple, but that’s the way it is. If he can just realize this, he could tap into Dr. Jekyll every time out, instead of Mr. Hyde.
All statistics courtesy of Baseball-Reference.