News
Loading...

Danny Salazar, Contextual FIP, and His Freezing Curveball

(Photo: Jerry Lai/USA Today Sports)
When Danny Salazar was demoted to Triple-A to begin the year, I was a bit disappointed but not particularly surprised. Salazar had struggled with his control in the spring, and considering his slow start in 2014 along with the value in giving Zach McAllister a chance in the rotation, it wasn't unreasonable to want him to work things out in Columbus. There was also the service time factor:

Salazar ended up being called up on April 18th, a couple weeks short of assuring that he won't become a Super Two but keeping the possibility open if he spends any more time in the minors.

Since being called up, Salazar has clearly been a dominant pitcher. He's been so good that even Rick Manning and Matt Underwood have discovered the fact that Salazar has a really, really good split-change.

To demonstrate just how good he's been, I will use a brand new stat called cFIP, which stands for Contextual Fielding Independent Fielding. cFIP was created by Jonathan Judge and was introduced in this article he wrote for The Hardball Times. You can find the cFIP leaderboard at Baseball Prospectus, along with DRA (Deserved Run Average), another context-neutral pitching stat created by Judge.

cFIP is similar to FIP in that it only looks at a pitcher's ability to prevent walks, hit by pitches, and homers while inducing strikeouts. It improves on FIP by making adjustments for the context of each plate appearance. Here is a quick summary from the Hardball Times article:

cFIP has multiple advantages: (1) it is more predictive than other pitcher estimators, especially in smaller samples; (2) it is calculated on a batter-faced basis, rather than innings pitched; (3) it is park-, league-, and opposition-adjusted; and (4) in a particularly important development, cFIP is equally accurate as a descriptive and predictive statistic.
The last point is important because it allows us to use cFIP to answer both "how well has this pitcher pitched?" and "given his performance, how well is he likely to pitch in the future?" With samples still very small, it's extra important to use stats like cFIP which separate performance from other variables.

So what does cFIP think of Danny Salazar this year?



Salazar's cFIP is 18 points (1.2 standard deviations) better than the second best cFIP, which is the same gap as between number two and number ten. That's simply remarkable, and while it doesn't mean that Salazar will continue to be the best pitcher in baseball, it makes it hard to argue that anyone has been more effective this year.

Most of Salazar's cFIP lead is due to his incredible strikeout rate. He has struck out 37 percent of batters, 17 percentage points above league average and seven percentage points above the next highest strikeout rate by a starter, which is James Shields at 30 percent.

However, raw strikeout rate doesn't even demonstrate how dominant Salazar has actually been. Instead, we can use context-neutral strikeout rate, which is calculated as a component of cFIP and adjusts for the batters each pitcher has faced, whether they faced them at home or on the road, whether they had the platoon advantage or not, what stadium they were playing in, and which catcher was behind the plate.

When we look at the leaderboard for context-neutral strikeout rate above average, the gap between Salazar and the rest of the field widens:


After making the proper adjustments, Salazar's strikeout rate is 13 percentage points higher than Shields, the same gap between Shields and the 43rd best adjusted strikeout rate on the list of 134 starting pitchers who have thrown at least 20 innings this year. He's also done a much better job of preventing walks, with a context-neutral walk rate 3.2 percentage points better than league average, 25th best among starting pitchers. Here's a visual look at how those rates have pushed Salazar to the top of the cFIP leaderboard:

Salazar simply stands alone in his ability to control the strike zone this year. The one cFIP component that is not ideal is his context-neutral home run rate, which is 1.5% higher (worse) than league average, which ranks 114th among the 134 starting pitchers who have pitched at least 20 innings.

Salazar's home run rate has increased over the last two years despite his ground ball rate rising from 34.4% in both his first two seasons to 43.1% in 2015. He's been hurt by a 20 percent HR/FB ratio, which is very unlikely to continue.

Salazar likes to throw fastballs up in the zone, where his velocity can be more effective. This is going to lead to some very hittable pitches when he misses his spot, so some homer-heavy spells are to be expected. Luckily Salazar has been so dominant at controlling the strike zone that he can afford to give up a homer here and there.

Since being called up in 2013, Salazar has mostly relied on his fastball and split-change (which we'll call a splitter). Let's take a look at starting pitchers who, like Salazar, have an average fastball velocity of at least 95 mph. The following shows each pitcher's Fastball Rate (fastballs per pitch) and Offspeed per Non FB Rate (changeups and splitters per non-fastball pitches, which include sliders and curveballs). I excluded 2015 from the sample so we can see how Salazar has changed his pitch mix this year.



Salazar has a relatively unique pitch mix, throwing his fastball with the second highest frequency and his changeup for the third highest frequency. Rubby De La Rosa, who pitched for the Dodgers and Red Sox and now pitches for the Diamondbacks, has the most similar pitch usage, but even he throws his fastball six percentage points less often and his changeup ten percentage points more often.

That strategy worked pretty well for Salazar, as his splitter recorded the third highest swinging strike rate of any splitter in MLB and his fastball was reasonably effective. His slider was clearly the weak point in his repertoire, as he allowed a .506 slugging percentage on the pitch, mostly due to leaving it over the plate.

Here's how Salazar has used his pitches this year:

He's cut down the fastball usage significantly and thrown more splitters when he strays from the fastball. He's leaned on the splitter in particular with two strikes, going from throwing it 26% of the time in 2013-2014 to 46% of the time in 2015. It's a beautiful, dominant pitch that plays extremely well off his fastball. It's continued to be unhittable, with batters whiffing on 32% of total pitches, a higher rate than any other pitch thrown by a starter this year.

One great thing about a fastball-splitter or fastball-changeup combination is that the two pitches often look quite similar out of the hand, a phenomenon called "tunneling". Salazar does an excellent job of keeping his fastball, slider, and split-change in the same tunnel. Here's a look at the average pitch path of those three pitches (graphing tool created by Harry Pavlidis).


His slider and changeup in particular have remarkably similar vertical movement, with the two pitches diverging significantly when they start to break horizontally. This suggests that as long as Salazar gets his slider command to be decent, the pitch will improve because it will be tough for batters to pick the pitch up out of his hand.

This year, Salazar has added a curveball to his repertoire. This pitch has a much different path, with a big vertical bump upon release that is easier for a hitter to pick up.


Perry Husband, who came up with the theory of effective velocity, calls pitches that come out of a pitcher's typical tunnel "freeze pitches". Freeze pitches are meant to surprise the hitter and catch them off-guard, leading to called strikes. I'm not sure if Salazar has deliberately tried to use his curveball as a freeze pitch, but the results speak for themselves.

Here are locations of Salazar's curveballs:

And here are the results, courtesy of Baseball Savant:  


Salazar has not quite figured out how to command his curve yet leaving some far out of the zone and many right down the plate. Hitters have not been eager to swing at it, with just six swings total. He's only thrown it six percent of the time, which isn't surprising considering the fact that he is still learning the pitch.

That said, due to the pitch looking so remarkably different compared to his other pitches, I wouldn't be surprised if Salazar continues to limit the usage of his curveball. It's a great pitch to keep hitters honest and get some cheap called strikes here and there, and his splitter is dominant enough that he can rely on that to finish batters off. If batters start to be more aggressive on the pitch, Salazar may need to start to command it better and not leave it over the plate all the time.

Salazar has elicited a lot of emotion from Tribe fans, and that will continue as he either crushes the competition or falls short of expectations. He's got two great pitches, one that could be good, and another that will keep batters on their toes. It's remarkable to think that he is only the third best pitcher on this pitching staff, though by the end of the season he could be considered the best.



Share on Google Plus

About Nick Wheatley-Schaller

Under Construction