(No more) tinkering with Liriano

Just after we published the Release point notes, in which Francisco Liriano was shown as an example of pitchers who deliver from different sides of the rubber according to batter handedness, something started to change.

Liriano stumbled out of the gate this season, losing his first two games, never surviving later than the fifth inning.
His ERA after those outings was 7.71. Bad luck could hardly be blamed for that, as he rode a very healthy BABIP, both compared with the league average and to his previous years values. He allowed eight walks and struck out the same number of batters; he barely succeeded in throwing a strike every other pitch (97 for 187, 52 percent) a rate way lower than his established 63 percent mark.

During the April 13 game, pitching expert Bert Blyleven declared from the booth that pitching coach Rick Anderson had instructed Liriano to stop his dancing on the rubber, requesting him to throw from the same spot, against both righties and lefties.

As you can see in the plots below, Liriano had consistently started his pitching motion from opposite positions depending on which side of the plate his opponent was positioned.

That did not change in the first two bad starts of 2011.

After the not-so-brilliant beginning of the season, the pitching coach’s suggestion was implemented. The altered approach is more evident in the April 18 start.

During the aforementoned broadcast, Blyleven explained the rationale behind the fix. Since Liriano had trouble finding the strike zone, and since he is not one of those pitchers making his living on the black, delivering from the opposite ends of the rubber wasn’t much useful for him, while starting his motion from the middle of the plate should have helped his control.

Unfortunately, one and a half games is not enough of a sample size to judge the effect of the transition. Liriano lost the April 13 game against the Royals, despite a healthier 68 percent strike ratio, and actually got his first win on the 18th, defeating the Orioles the first time he fully used the single delivery spot approach.

That turned out to be also the last time he used that style; he reverted back to his previous technique on the April 27 game against the Rays.

With all the tinkering (and the necessary caveats due to a tiny sample size), no changes were apparent in Liriano’s walks to strikeouts ratio: He was 6:6 when following Anderson’s advice and, upon returning to his former modus operandi, he recorded another tie (4:4 on April 27), then changed for worse, issuing six free passes and striking out two White Sox on May 3.

Wait, that was his no-hitter!

It’s hard to talk about a pitcher’s struggles after he has delivered a no-hit, no-run game. Probably the feat will put confidence in the 27-year-old Liriano, and he will get over whatever trouble (even if that’s samplesizitis) has been preventing him to pitch brilliantly.
However, after a bunch of starts, some differences from the 2010 version (and the previous ones as well) of Liriano are apparent.

His K/BB is unusually low, a result of both his increased walk rate and the strikeout rate below league average for the first time in his career. His BABIP is also low, meaning that his terrible season (no-hitter notwithstanding) might have been even worse.

Despite the higher percentage of pitches resulting in a ball, Liriano has not altered the percentage of pitches delivered in the strike zone (38.6 percent this year, 39.0 percent in 2010); however, it seems he is having troubles locating his first pitches, as he was able to find the strike zone over 46 percent of the time in the past, but he has been able to do that just about one third of the time this year.

The percentage of pitches outside the zone that batters swing at is also significantly lower (25 percent this year, 36 percent in 2010), probably a byproduct of the higher numbers of plate appearances beginning in a hitter’s count. Liriano has slightly increased the number of off-speed pitches thrown to begin the confrontation with the batter, but that does not seem to be the main reason his first pitch strikes have dropped, since the percentage of offerings resulting in strikes has diminished for fastballs, sliders and change-ups, without distinction.

According to FanGraphs Pitch Type Run Values, the slider is the pitch that has always earned Liriano’s money and, while both the fastball and the change-up show slight lower value this year, the slider has experienced the most significant drop. While part of the decreased value can again be explained by the difficulty of throwing first pitch strikes (thus creating fewer pitcher-friendly counts and in turn decreasing batters’ inclinations of going fishing), it looks like the characteristics of the pitch have slightly changed.

Here are the physical values, 2010 vs 2011, of Liriano’s sliders.

	Velocity	Vert mov	Hor mov
2010	85.7		0.05		-0.45
2011	85.6		3.32		-0.02

The above numbers should not be considered as conclusive evidence of a different slider coming out of his left hand, as changes in MLBAM classification algorithm might have caused them. However, the above table, combined with the drop in efficacy shown by FanGraphs (whose classification is not based on PITCHf/x) run values, suggests something is not quite right with Liriano’s biggest weapon.

What’s ahead?

As Ol’ Satchel Paige used to say “It’s funny what a few no-hitters do for a body,” thus Liriano probably will make us forget his inauspicious April.
We’ll not likely see the single-spot-on-the-rubber approach anymore, since it did not bring any change. Once again, one-and-a-half games is not enough to make an ultimate statement on that; but since the customary style brought a no-hitter, Liriano will hardly switch back to the new technique.

Before the historic game it was reported the Twins had discovered inconsistencies in Liriano’s release and had shown him videos of it. PITCHf/x data (adjusted as explained in Fine tuning PITCHf/x location data) do not expose any of this, but it’s very possible that, despite a consistent release point, the pitching motion was erratic.

The game-for-the-ages was not different from the rest of the season, as Liriano found himself chasing batters 62 percent of the time, not exactly what you would expect from a pitcher who has good command of his stuff.

If you believe in the psychological factor, this would be a good time for it to show up.

Note: Liriano started on May 10 against Detroit, but left the game after three innings due to flu-like symptoms. Data from that game have not been considered for this article.

References & Resources
StatCorner
FanGraphs
Baseball-Reference
Joe Lefkowitz’s PITCHf/x tool
TexasLeaguers