Barry Zito has had good stretches before. As recently as 2010 he strung together four starts in May which were even better than the four he made this April. So, what follows should not be mistaken as a declaration of renewed trust in the One-Hundred Million Dollar Enigma. I’ll declare a Barry Zito “renaissance” around the same time Ryan Theriot wins a batting title. What Zito accomplished this past month was definitely aided by his opponent’s BABIP of .188 (5th best among NL starters) and strand rate of 91.4% (3rd in the NL).  His FIP is more than two runs higher than his ERA. So, to put it lightly, some regression is imminent.

That said, sometimes a pitcher does “make his own luck” and Zito’s success this April has been founded not only on good breaks, but on a conscious change in approach. I believe this change has helped him cut his walk rate in half, while generating much less hard contact. His line drive rate currently sits at 13.8%, compared to 19.8% for his career, and over 21% since he became a Giant.

This new approach, I believe, is built upon three major adjustments. First, he has learned (again) to trust his slider. (Grant Brisbee discussed this is some detail last week.) Second, he is throwing a cut fastball. And, finally, thanks in part to these new additions, he is varying his pitch selection from start to start, based upon matchups and upon the quality of his various pitches on any given night.

Backing up this impression with statistics has proven somewhat difficult, because Zito’s new repertoire has thus far been a challenge for pitch-charters. Distinguishing the pitches in Zito’s arsenal can be very difficult because he does not have a talent for, how shall we say it, changing velocities. Zito’s changeup, when he overthrows it, looks a little like a flat slider. His two-seam fastball, his slider, and his new cutter all have overlapping ranges of both velocity and movement. His mistake pitches, especially, are hard to classify, which might make some of his pitch data unreliable, particularly in small samples.

However, no matter whose data you use (or if you watch the games), it is clear that Zito has made changes, it’s just difficult to tell exactly what they are. The following chart show three sources of pitch type data:

As you can see, there is little consensus as to what is a cutter and what is a slider. In fact, as of last week, when I began trying this analysis, Brooks Baseball was listing the cutter among Zito’s pitches. They have since given that up.

With the above caveats in mind, I want to nevertheless offer some analysis which might be interesting to consider as we view tonight’s start against the Marlins.

According to the PitchFX linear weights, the slider has been Zito’s most effective pitch since 2009. However, one could question the accuracy of that data, because prior to this season Zito never leaned too heavily on the pitch. With infrequent use, he didn’t risk overexposing it as much as, say, his curve. However, this season, the slider has become Zito’s bread and butter.

Regardless of whose data you believe, it’s evident he’s throwing the slider far more often than he has in recent years. For the first time since he became a Giant, Zito has a pitch that he can throw for strikes which doesn’t look like batting practice. Zito throws the slider with better command and control than either his fastball, curve, or change-up. According to Brooks Baseball, 62% of the time he puts it in the zone or in play.

What’s great about Zito’s slider is not quality, but consistency. Although he uses the slider with regularity in every start, it is never his most commonly used pitch. In Zito’s best starts, he doesn’t need it all the much. When he shutout Colorado in his first start, he only threw sliders about 14% of the time, as he pounded the Rockies with the sinking two-seamer. However, unlike that pitch, which Zito struggles to command from game to game, the slider shows up every night, thus giving Barry a pitch he can count on when he’s behind in the count, while he figures out what else in working on any given day.

Unlike the slider, Zito’s cutter is a brand new pitch, developed this offseason. When it’s working, the cutter breaks much like the slider, but has fastball velocity, and comes from the same release point as both pitches. The cutter has thus breathed new life into Zito’s mediocre fastball, both by making it less necessary for him to throw it, and making it’s “straightness” surprising when he does. Zito’s called strike percentage has gone way up on both his fastball and his sinker as a result. I believe this improvement stems directly from opponents getting caught expecting a cutter or a slider to move down, away, and off the plate against lefties, and instead standing and staring at something right over the plate.

As important as what Zito is throwing (cutters and sliders), is what he isn’t throwing (four-seamers and curveballs). PitchFX has Zito throwing the four-seamer only 11% of the time. His career rate is 44%. As discussed above, the straight fastball has been more or less totally replaced by cutters and sinkers.

The curveball has not been replaced, but it has been rejuvenated. What happened during Zito’s Giants tenure, prior to this season, was that as his faith in his fastball decreased, he became more and more reliant on his Bugs Bunny curveball. He’s thrown it more than 20% of the time since becoming a Giant. As hitters got more looks at it and Zito started using it in more predictable patterns, it was less effective.

That’s changed this season. Zito throws the curve less (~15%) and thus it is again generating whiffs and weak contact. Consider some anecdotal evidence from last week’s start in Cincinnati:

Zito got off to a rough start, giving up a single and two walks in the first inning. With two outs and two on, Zito faced Jay Bruce. He got him to strikeout swinging on a 73 MPH curveball down and away. It was only the second time Barry had thrown his curve during the 25-pitch opening frame.

Zito ended up throwing only 15 curves in the game, seven of them to Jay Bruce, who went 0-for-3 and twice struck out to end an inning with men on base. He reserved the curve for situations in which it would be most effective and did not condition the Reds (save, perhaps, Bruce) to waiting back on it. Contrast his most recent start (6% FB, 14% CB) with Zito’s last trip to Cincinnati, during which he relied almost exclusively on fastballs (46%) and curves (20%), one of which Jay Bruce pounded into the left-field bleachers, driving home three of the five earned runs Zito yielded in five innings.

The existence of the slider and the cutter allow Zito to keep the curve in reserve, for when he needs it most. It has also allowed him to be less reliant on his sinker, which is frequently his best pitch, but is also unreliable on a game to game basis. In his four starts this season, according to PitchFX data, he has thrown the sinker as much as 30% of the time and as little as 10% of the time. Somewhat similarly, he’s thrown the changeup as much as 20% and as little as 10%. In four starts, he has never used an even remotely similar pitch distribution, as shown by the following table: