When deciding on the grade for a pitcher’s breaking pitch, a scout relies on the pitch’s velocity and movement (although command can be factored in as well).  These factors are combined into a single number on a scale from 20-80, with major league average as a 50 and a standard deviation recorded at 10.

Clayton Kershaw’s curveball has long been regarded as one of the best in the business, yet by my systematic calculation factoring in velocity, horizontal movement, and vertical movement, his curveball rated among the bottom third of curveballs.  Its below-average velocity and below-average horizontal movement held it back despite its above-average vertical movement.  While I was pondering this conundrum, I remembered another fact–Kershaw’s fastball happens to have a lot of rise.  What if movement was recorded by the difference between the pitcher’s fastball movement and the breaking ball movement instead of the breaking ball’s movement compared to an arbitrary point?  I was about to find out.

(This paragraph is solely methodology, so skip it if you wish.)  Using Baseball Prospectus’ excellent pitch f/x leaderboard, I selected all pitchers who threw at least 200 four-seam fastballs and at least 100 curveballs in 2014.  A breaking pitch’s horizontal and vertical movement was recorded as the difference between the pitch’s raw movement and the pitcher’s four-seam fastball’s movement.  I calculated the z-scores for the curveball’s velocity and the z score for the combination of the z scores of the curveball’s relative movements.  (I gave a 150% weight to vertical movement over horizontal movement).  Then, I combined the z scores of the velocity and combined relative movement to calculate a relative pitch score.  (I gave a 150% weight to combined relative movement over velocity).  Finally, I calculated a scouting grade on the 20-80 scale based off the relative pitch score.  Below is a table showing my results.

While I may not have solved the difference of evaluation among Kershaw’s curveball, the relative scouting grade at least opens discussion on how movement and velocity of pitches should be evaluated.  Is it better for a breaking pitch to be faster, or is it better to create a wider difference in velocity between the fastball and breaking ball?  Is it better for a pitcher’s breaking ball movement to be as different from the fastball as possible, or do some similarities create greater deception because a hitter can’t recognize the pitch earlier?

After reading Kiley McDaniel’s articles on explaining the scouting scale, I thought that I would take a different approach in trying to explain it–namely the approach of gifs.  Although the scale is normally reserved for players who haven’t lost their rookie status in the majors, perhaps a visualization can better illustrate what “major league average,” “plus,” or “below average” looks like.

To show the differences between curveballs in different positions along the scouting scale, major league curveballs must first be graded.  To do this, I pulled up Baseball Prospectus’ pitch f/x leaderboard and set it to filter for pitchers who threw at least 100 curveballs in 2014.  The way in which pitch movement is measured and the fact that lefties’ curveballs’ horizontal movement appeared to be measured lower than righties forced my methodology.

I split up righties and lefties into separate groups for analysis.  Pitch movement was recorded based on the inches a pitch broke more than the the pitch with the least break.  Total movement was recorded as the square root of the combined horizontal and vertical movement squares (C² = A² + B²).  Z Scores were recorded for each curveball’s velocity and movement, and then were added (with movement receiving a 1.5 times greater weight) to form a grade.  The grades were then transferred into Z Scores with a median of 50 and each standard deviation being 10.  Finally, the righties and lefties were combined to make a final scouting scale.

Name / Throws / Total Movement Z Score / Velocity Z Score / Scouting Grade

For our first gifs, we’ll look at two of the top curveballs on the by-the-numbers scouting scale.  At an slightly above average velocity of 79.7 mph with the highest amount of break above than the baseline (sorry, Fernando Salas), Garrett Richards’ curveball is a sight to behold and grades out as a 73:

Next, we’ll look at Tyler Skagg’s curveball, which, features great horizontal and vertical movement while maintaining average velocity.  It grades out as a 69:

With the plus-plus-type (70) curveballs out of the way, we’ll take a look at some plus (60) curveballs.  Wandy Rodriguez fits this category.  His curveball has slightly less, but similar, movement as Skagg’s but it’s lesser velocity makes it a lesser-quality pitch.  Notice how it has defined break, but it appears a bit loopy due to its velocity:

With Kevin Jepsen, we see a curveball that is graded similarly but looks much different than Wandy’s.  Although its break looks sharp because of its velocity, the total movement is slightly below average.  This pitch may actually be a slider, but Jepsen’s breaking pitches tend to run together so consider it a representative picture of his curveball.

Next, we’ll look at an average (50 on the scale) curveball.  Erik Bedard gets slightly above average break, but with below average velocity.  It has solid two-plane break, but it doesn’t look very sharp:

Next, we’ll look at a below average (40 on the scale) curveball.  Jordan Lyles has good velocity on his curveball, averaging 81.75 mph, but he also averages nearly two standard deviations less break than an average curveball in this sample.  The following is a tough angle to see the break, but it lacks the sharp break of a better curveball.

Finally, we’ll look at Jered Weaver’s well below average (24 on the by-the-numbers scale!) curveball.  His curveball was the slowest in the sample and also featured below average break (much of the perceived break is from its low velocity).  It’s still an effective pitch for him, but it’s probably a result of his height and delivery combination than his curveball.  With just about any other pitcher, it would likely be ineffective.

On August 3, 2009, Neftali Feliz made his major league debut against the Athletics, pitching two perfect innings with four punchouts. In those innings, he mowed down hitters with 23 fastballs that averaged 99.45 mph, 4 changeups that averaged 91.13 mph, and 3 sliders that averaged 82.43 mph. He would end his rookie season with a 1.74 ERA (2.48 FP), a 33.3 K%, and a 6.8 BB%. As a 21 year-old, he already looked like a bonafide bullpen ace for the Rangers.

Fast forward to the end of spring training in 2014. Feliz is 25 years old and in the prime ages of his baseball career. And he’s starting in AAA. With a fastball that is 91-93 mph. Rangers’ General Manager Jon Daniels said of him, “He’s healthy and his work ethic has been solid, but he needs some work and the best place to get him that is in Round Rock right now. I expect he’ll be back as soon as he’s ready to help us.” A team whose bullpen for opening day included Seth Rosin thought that Feliz wasn’t ready to contribute for them out of the gate. Clearly something was off.

On August 1, 2012, Neftali Feliz underwent Tommy John surgery. Tommy John surgery generally requires at least 12 months for recovery, and he was back in the majors by September 2013. He averaged 94.19 mph with his fastball during his 6 games in September 2013. While his velocity was a step down from his 97+ mph heat in 2009-2011, pitchers often have to slowly build their arm strength up again to pre-surgery levels and there was no reason to believe he wasn’t on track to doing so. When his velocity failed to reach that level through most of his 2014 campaign, though, it became unclear if he would ever regain his pre-surgery stuff.

By some measures, his time in AAA was a success. He struck out 9.73 batters and walked only 2.51 per 9 innings. He produced a 3.14 ERA compared to the Pacific Coast League’s league-wide 4.64 ERA. His biggest problem was home runs—he gave up 6 in only 28.2 innings. When was he called back up to the major league squad on July 4, there were reasons to be cautiously optimistic that he could find some success again as a reliever. The Rangers noted that he was throwing in the mid 90’s some games while in others he would sit in the low 90’s.

Feliz didn’t exactly dominate during his early outings. Through July 23, he had pitched 10.1 innings with only 4 strikeouts, 3 walks, and 2 home runs given up. Yet, out of the playoff race, Texas dealt their closer Joakim Soria to the Tigers and anointed Feliz their new closer. While it’s possible that the team merely liked his shiny ERA at the time over his FIP (2.61 to 5.75), perhaps they started to see some signs of life in him. Regardless, his 1.69 ERA and 13 saves out of 14 save opportunities the rest of the way probably made them feel validated in their decision. With his end of the season performance, it appears likely that he will be the Rangers’ opening day closer.

Projecting into 2015, Feliz’s 4.90 FIP and -.1 WAR from 2014 provide red flags. His home run rate also look to be an issue. His extreme flyball tendencies (51.1 FB% versus 27.3 GB%) resulted in 1.42 HR per 9 innings despite a fairly ordinary 11.1 HR/FB% rate. His 17.2 K% and 9 BB% doesn’t exactly inspire confidence, either. Steamer isn’t a fan and projects him for .1 WAR in 65 innings.

But there are reasons for optimism, too. He kept up his low BABIP streak at .176 (.215 career) thanks to his impressive 20 IFFB% (17.8 career). Among relievers with 200 innings since his debut, he has the lowest BABIP, the 6^th lowest LD%, and the highest IFFB%. Steamer projects him for a .284 BABIP next year, but I’m willing to bet his will be much lower than that figure and will continue to let him beat his FIP by around a full run.

Next, we’ll look at his home run rate. His 2014 figure was the highest of his career, caused primarily by his 11.1 HR/FB% (6.9 career). What may have caused that? Well, it may have been caused at least in part by his changeup. A changeup is a pitch designed to fool hitters who are looking for a fastball: it is supposed to be thrown with identical arm speed as the fastball to make it harder to pick up, and then its velocity and/or movement difference makes it effective. For a pitcher throwing in the upper 90’s with hitters already struggling to catch up to their fastball, a changeup may be less effective because the velocity reduction may sometimes help the hitter instead of hurting him (of course, there are exceptions). When Feliz was throwing in the upper 90’s in 2009-2011, he threw his changeup just 4.4 percent of the time. When Feliz was throwing in the low to mid 90’s in 2014, he threw his changeup 12.4 percent of the time. For his career, opponents have a .212 ISO against his changeup compared to a .119 ISO against his fastball.  In 2014, hitters had a .429 ISO against the pitch, including 3 of his 5 home runs given up on the year.

His velocity provided another reason for optimism as well. While his early- to mid-season velocity wasn’t great, he improved as time went on: in July he averaged 92.88 mph; in August he averaged 93.7; and in September he averaged 95.81. The ISO against his fastball decreased each month as well (from .107 to .107 to .053), even as he increased his usage of his fastball (from 64.7% to 77.36 to 77.78).  His velocity increase had an added bonus as well: it allowed him to use his changeup less (from 14.72% to 11.32 to 3.17). The biggest question is whether he can maintain his September velocity, or even improve upon it.

Overall, I don’t think Neftali Feliz is a safe bet to be great in 2015. But I do think that he has a real chance to be much better than the projections project him to be. To end this post, I’ll post a few gifs of Feliz at his best in 2014:

* All pitch usage, velocity, and movement numbers are obtained from Brooks Baseball. All pitch results numbers are obtained from Baseball Savant.

Rogers Hornsby was once quoted as saying, “People ask me what I do in winter when there’s no baseball. I’ll tell you what I do. I stare out the window and wait for spring.”  While staring out of a window for several months may be a worthwhile way to spend our limited time on Earth, I wish to propose a marginally more enjoyable choice: watching gifs of top pitching prospects.

With the 2014 Arizona Fall League featuring several top pitching prospects, and with the plentiful high-quality videos posted on Youtube by our good friends at MLBProspectPortal.com, I was able to create a collection of gifs that showcase these pitchers’ talents.

Kyle Zimmer

Zimmer finished the AFL with an incredible 41.7 K%.  Unfortunately, due to injuries, he only made three starts.  When he has been healthy, though, he has been known to throw mid 90’s fastballs and biting curveballs.  Below is a gif of his curveball thrown to Pirates’ outfield prospect, Josh Bell.  It should be noted both that Bell missed this curveball by a lot, and that he rarely swings and misses versus left-handed pitchers (9.9 K% against them).

In the pitch following the curveball, Zimmer threw a high fastball that Bell couldn’t catch up to:

In the pitch following the curveball/fastball combination, Zimmer threw either a changeup or two-seam fastball that had Bell out in front:

Tyrell Jenkins

A former 1st round supplemental draft pick by the Cardinals in 2010, Tyrell’s statistics have been erratic in the minors, including his 13.3 K%/7.4 BB%/4.31 FIP in High A in 2014.  His athletic frame and power stuff are still coveted, though, and the Braves acquired him in the recent Jason Heyward trade.  Below is a curveball from Jenkins to Addison Russell, who happens to be one of the top prospects in all of baseball:

Below is a high fastball to Dalton Pompey that resulted in a strikeout.

Archie Bradley

Archie Bradley is likely the most well-known name on this list, and some sources had him as the #1 pitching prospect entering this season. FanGraphs’ own Kiley McDaniel gave him a future 70 grade on his fastball, along with a future grade of 65 on his curveball.  Below is one such curveball thrown to Addison Russell:

Not content to merely acquire a strikeout against Russell, Bradley went with his fastball with good arm-side run to create a double play-inducing weak grounder:

Tyler Glasnow

Tyler Glasnow posted a league-leading 31.9 K% and .171 opposing batting average  in 2014 in High A.  The third highest K% in the FSL was 23%, and the second lowest opposing batting average was .216.  These numbers may seem incredible to you, but they’re actually a step down from his 2013 season in Low A where he posted a 36.3 K% and a .141 opposing batting average.  You may have guessed that Glasnow has good stuff.  You’d be right.

Below, the batter sticks his bat out for a bunt and then spins back as if the pitch were going to hit him.  It was called a strike.

Finally, we have a called strike three on a curveball.  Notice the defeatism in the batter.; he has been bested by the pitcher but is glad that he has been allowed to keep his life.

*Credit for all of the videos goes to MLBProspectPortal.com.