02 March 2016

Ubaldo Jimenez's Adjustments Keyed An Improved 2015

Ubaldo Jimenez’s 2015 season was markedly better than his 2014 one. Year-over-year he lowered his walk rate by 5.3 percentage points while increasing his strikeout rate by 0.2 percentage points. His ground ball rate also shot up by 7.8 percentage points. As a result, even while his HR/FB rose 2.1 percentage points his ERA-, FIP- and xFIP- all dropped by 23%.

photo by Keith Allison under CC BY-SA 2.0 license

Jimenez himself credits the success to working with Ramon Martinez (brother of Pedro). The Orioles hired Martinez in March and he, along with pitching coach Dave Wallace and bullpen coach Dom Chiti, helped Jimenez address his complicated mechanics. Said Ubaldo:
[Martinez is] a smart guy. He knows how to be a good teacher. He learned from Dave, so they have a good chemistry. If you don't learn with them, you're not going to learn with anybody else. 
Along the way to these improvements, Jimenez did something he’d never done before. In 2015 he posted the highest O-Swing% of his career: 27.4% according to PITCH f/x. His previous high was 27.1%, which he achieved when he was still averaging 96 MPH with the Rockies in 2009. He is a far cry from that pitcher today. With a fastball that averages only 90 MPH today, Jimenez is still learning to pitch with a diminished repertoire.

A high O-Swing% is very good for a pitcher. When a batter swings at a pitch outside of the zone, he cannot walk. Second, he has a reduced chance of making contact. Last year the league O-Contact rate was 62.9% compared with an 87.1% Z-Contact rate. Finally, if a batter does make contact, its impact is reduced. Last year the average batted ball out of the strike zone left the bat at 83 MPH, compared to 91 MPH when the batted ball was in the zone.

Indeed an examination of the 1,283 pitcher-seasons since 2007 shows that as a pitcher’s O-Swing% increases, his xFIP goes in the other direction:

All of this is just to show that a career-high O-Swing% for Jimenez is a good thing, just like it is for any pitcher.

How did he get there? The following table breaks down his O-Swing changes by pitch:

While Jimenez’s O-Swing% was up across the board, the charge was led by his split-fingered fastball. The following graph tells part of the story, that hitters swung at the pitch much more often than last year:

Unfortunately for hitters, Jimenez’s splitter is out of the zone more than ⅔ of the time. So if a hitter swings at it, that means he's been fooled.

What led to this improvement? First, Jimenez threw the pitch more than a mile per hour harder than he did in 2014:

Gaining an entire mile per hour is a significant change. Jimenez also tightened up his release point, which perhaps had two effects: helping him land his pitches in the same spot, and hiding the splitter better. The following images show his fastballs’ release points in 2014 and then in 2015:

In 2015 the release points for his fastballs overlap much more. There still is about an inch and a half of variance, but if you're a hitter looking at Jimenez’s arm angle to decide what pitch is coming, you’ll now have a tougher time guessing. Given that you most likely have two strikes on you anyway, so you have to protect, you're going to have a harder time making solid contact.

In fact, Jimenez made his release point more consistent across the board:

Overall he reduced the variance in his vertical release by a whopping 50%. That is a nice improvement in his ability to fool hitters. It probably makes up for the fact that the variance in his horizontal release point actually grew by a third of an inch.

As a result of throwing it from a lower slot, the splitter appeared to drop more:

This increase in vertical drop, combined with the increased velocity and more consistent release point, made the splitter a more effective pitch in 2015. Although hitters’ BA against the pitch increased from 0.175 to 0.224, their ISO against it dropped from 0.138 to 0.129.

Now far removed from his flamethrowing days, Jimenez continues to grapple with being a sinker/slider pitcher who relies on deception rather than heat. 2014 wasn't his year but 2015 was a step in the right direction. He showed that, even after ten years in the big leagues, he can adapt and adjust to new circumstances. That’s good because Orioles fans have at least two more years of watching his funky delivery every fifth start.

All data from FanGraphs, Baseball Savant, and Brooks Baseball.


Anonymous said...

Oops, bad start!

Anonymous said...

Thanks for providing another another metric to look for when evaluating the performance of pitchers. I like to look at that kind of data both because I enjoy playing in fantasy baseball leagues and because it gives me a better understanding of why various pitchers are failing or succeeding.
Now when I read fangraphs I'll look for O-Swing %, in addition to pitch/fx data.
One thing I'm curious about: where do you find information for league rates on data like O-contact and Z-contact rates? I don't know where to find that information.

Roger said...

Did you guys see the Fangraphs piece on extreme groundball pitchers? Are we concerned that both Jimenez and and Gallardo could flame out really really quickly? Seems like the Gallardo medical is a foreshadowing of things to come. I had no idea that Bill James had come down so hard on groundball pitchers.

Anonymous said...

Yes, I saw that fangraphs piece too. Actually there was more than one. Interesting stuff, as usual. The one I remember in particular was the one that showed a correlation between a heavy GB profile and higher than expected HR/FB rates.
Camden Yards would be a confounding factor, right? After all, it consistently yields more HRs than all but 5 or 6 ballparks.
I was looking at Marcus Stroman's data on fangraphs about a month ago and I noticed that what happened to him last year seemed to confirm this: he had a GB rate above 60% but a HR/FB rate that was greater than 14%. But then again, he pitches in Rogers Center---which yields a lot of home runs---and the AL East which, with the exception of the Trop in Tampa Bay, has numerous hitter friendly venues.
Didn't that article conclude (or speculate) that it might have something to do with how the batter swings at pitches lower in the strike zone?

Jon Shepherd said...

Uppercut swings yield more homeruns with groundball pitchers because flight paths of ball and swing paths of bat mirror each other. The answer of course is to pitch to upper cutting hitters differently.

I have not read the FanGraphs piece, but the above is something the Athletics were doing about four years ago...matching swings and ball paths.

vilnius b. said...

I guess that's why the high fastball has become so popular. Everybody knows how successful Bumgarner is using the high fastball, but other pitchers are doing it more as well.

I noticed, from watching Orioles games, Drew Smyly of Tampa Bay was doing a great job using the high fastball to good effect. I don't remember him doing that with Detroit. I guess the Rays coaches told him to start using it more.

And I'm pretty sure I read an interview with Brandon Moss on FanGraphs, in which he said that one reason he struggled so much his last season with the A's (he also had injury problems) was that pitchers figured out he couldn't handle the high fastball. He once thrived on low fastballs but he's been seeing fewer of them the last couple of years.