Developments in Evaluating Pitchers...Elementary, Watson

File this under "New Technology--Pitcher Evaluation," Dr. Watson.

Sherlock Holmes: "Data, Data, Data. I cannot make bricks without clay."

Most of you know about Pitch f/x  If you're like me, you look forward to David's evaluation of pitch f/x metrics for Astros' pitchers.  Sport's Illustrated's Tom Verducci had an intriguing article about new technology that may provide some pitching data beyond Pitch f/x.  According to his article, a Danish company, Trackman, has used Doppler radar technology to track pitches in three dimensions, at one or more American League ball park.  I don't know enough about the technology to say whether it's a significant improvement over Pitch f/x.   But the firm's interpretation of the data results makes for interesting discussion, and perhaps provide some pieces for the jig saw puzzle of evaluating major league pitchers.  By measuring the pitch from the moment it leaves the pitcher's hand, Trackman provides an additional source of data for understanding why pitchers are successful.

Watson: "Is there any point to which you would wish to draw my attention?"
Holmes: "To the curious incident of the dog in the night-time."
Watson: "The dog did nothing in the night-time."
Holmes: "That was the curious incident."

One of the curiosities of pitcher performances is when a pitcher's fastball seem to be more effective than its velocity would indicate.  Sometimes the effectiveness may be due to movement of the pitch.  But for some pitchers, the fastball is perceived by batters to be faster than the actual measured velocity.  Scouts sometimes refer to this as "sneaky fast" or say that the fastball has a "late hop."  In some cases, this may be the result of deception by the pitcher.  But the Trackman data suggests that a major cause of this effect is the span on the pitcher's extension in releasing the ball.  Velocity, in the pitch f/x or radar gun sense, is based on speed of the ball, measured in miles per hour between discrete points.  However, if the pitcher can reduce the distance that the ball travels to home plate, then the flight time for the ball will be faster than expected.  A pitcher's momentum and stride can produce a release point closer to the batter.  The Trackman data showed that pitchers whose extension is approximately one foot longer than average can add around 2 mph to their effective velocity.

In effect, for these long extension pitchers, the ball seems to get to the batter faster than the visual appearance of the speed of the pitch.  As a result, the necessary reaction time for the batter is less than expected, and in plain terms, the batter's timing can get messed up. The article shows that the release extension is related to strike outs and swing and miss rates.  A long wing span will help a pitcher release the ball farther forward.  And the listing of pitchers in Verducci's article include a fair number of tall, gangly pitchers.  But the longest extension for pitchers at the AL park was a 5 ft.-10 in. Yankees relief pitcher, Jack Robertson, who has a measured velocity of 93 mph, but an effective velocity (from the point of release) of 95 mph.  Not surprisingly, Robertson has a reputation for a "lively" fastball.  Pitching mechanics play an important role, and allow shorter stature pitchers to be "sneaky fast."   Verducci's article discusses only AL pitchers, but it would be interesting to know the "effective velocity" for Tim Lincecum and Roy Oswalt---two short stature pitchers known for the long stride of their delivery.

J.A. Happ and Wandy Rodriguez are two Astros' starting pitchers who are often called "sneaky fast." Happ is a tall angular pitcher whom we would expect to get a long extension on releasing the ball.  Wandy is not tall, but he is generally described as deceptive in his delivery.

Holmes: "when you have eliminated the impossible, whatever remains, however improbable, must be the truth."

Verducci's article cites Trackman's conclusions regarding the spin rate (revolutions per minute) of the pitch. But, unlike most pitch f/x analyses,  the focus is not on the amount of break or movement of the pitch. The spin rate of a breaking pitch does not dictate the amount of break.  (The orientation of the spin axis, or spin angle, are involved in determining horizontal or vertical break, the most discussed components of a pitch f/x analysis.)  But the absolute speed of the spin rate for both curveballs, and to some extent fastballs, are correlated with lower batting averages and slugging rates, and higher swing and miss rates, according to the Trackman data.  Verducci suggests that a faster spin rate makes the pitch harder for batters to identify the pitch type, because seams cannot be seen as clearly.

Reggie Jackson once said that a rookie hitter, no matter how good, wouldn't last in the major leagues if he couldn't see the "red dot" of a slider.  Batters try to quickly identify the type of pitch based on what they can see from the rotating ball and red seams.  Baseball Prospectus cited an interview with Kirby Puckett and Wade Boggs for this description of pitch identification:

[A]s the ball rotates toward the plate - usually 14 to 16 revolutions per pitch - the best hitters can ''read the seams'' to learn the pitch's identity. The spinning seams create distinctive, tell-tale patterns.

A slider appears to have a red dot, often near the ''2 o'clock'' position on the ball. A curveball gives the ball a tumbling effect, with horizontal stripes rolling downward…

A fastball looks very white and clear, with the seams barely visible. A four-seam fastball is slightly darker than a two-seamer…

A forkball also has a tumbling effect, but it appears more oblong than a curveball.

Thus, according to Verducci, a higher spin rate on the pitch confuses the batter's normal decoding of the pitch type.  But it's important to note that, even though a curve ball's or slider's spin rate may make it harder to identify, the actual break on the pitch (and not just the spin rate) should be important factors relevant to pitch effectiveness.  A number of other factors probably contribute to effectiveness too, such as sequencing of pitches, differences in pitch velocities, and deceptive actions like hiding the ball.

Beyond pitcher evaluations, the technologies for comparing spin rates on pitches could be used in scouting and pitcher instruction.  For example, in the latter instance, the pitching coach could determine the technical details of a pitcher's mechanics that produce the highest spin rates on pitches.

Sherlock Holmes: "Never theorize before you have data. Invariably, you end up twisting facts to suit theories, instead of theories to suit facts."

Verducci's theorizes that fastballs with high spin rates may enhance the "sneaky fastball:"

Pitchers who don't throw hard but have high spin rates on their fastball -- such as Shaun Marcum of the Brewers and Koji Uehara of the Orioles -- post higher strikeout rates than their modest velocity would otherwise suggest.

So, I thought I would compare the spin rates for the Astros' starters' 4 seam fastballs with spin rates for the following pitchers: Marcum, Uehara, Verlander, Oswalt, and Cliff Lee.  I chose Marcum and Uehara based on the quote from Verducci's article, above.  Verlander had the highest spin rate on curveballs in the Trackman sample, and I included him as a high velocity comparison pitcher.  Oswalt and Lee are right hand and left hand pitchers who have excellent 4 seam fastballs which seem to be more effective than the measured velocity.  The data is from Texasleaguers.com and encompasses pitch f/x data from April 1, 2010 through April 12, 2011.  I have also shown 2 seam fastball spin rates, where applicable, for comparison.

Inspector Lestrade: In another life, Mr. Holmes, you would have made a excellent criminal.
Sherlock Holmes: Yes, and you an excellent policeman.

Wandy Rodriguez and J.A. Happ are both in the "crafty lefty" variety, with similar fast ball velocity just below 90, who are sometimes described as "sneaky fast."  Happ, in particular, has a high spin rate on his fastball, and perhaps this contributes to his ability to outperform what you might expect from his raw stuff.  Wandy has higher spin rates on his fastball than Marcum and is just below Uehara, both of whom were cited by Verducci's article as examples of pitchers whose spin rates allow them to achieve better than expected strike outs. 

I also wonder if a small differential in spin rates between 4 seam and 2 seam fastballs may add to the deception, confusing batters in differentiating the two types of fastballs.  Although Myers' spin rate on his 4 seam fastball is relatively low, notice that he is the only pitcher here with a higher spin rate on his 2 seam pitch.  I am speculating, but maybe this prevents batters from identifying  the "darker" clear circle of the four seam pitch versus the 2 seamer.  Simiarly,  Roy Oswalt's 2 seam fastball rotates only slightly less than his 4 seam pitch.

Any thoughts out there?