Your definition of fun may vary. But Yu Darvish and his eight-pitch mix are going to make life interesting for catchers, hitters and even PITCHf/x analysts. Here’s a picture from his Cactus League debut. The pitch in red was a strike three splitter to end an inning. The axes show movement during the flight to home plate from the catcher’s perspective.

There’s an 86 mph cutter that’s got less drop and more hook than the other, faster, cutters. What’s up with that? Looks like two change-ups on the left side of the change-up and splitter group. And you can clearly see the slow curveball.

This was a situation where game video and post-game interviews helped out. Dan Brooks looked at my original rough classifications and suggested some improvements. You can see those on Darvish’s player card. The charts include his World Baseball Classic appearances, but you can filter the tables by year. We plan on providing yearly movement graphs in our next update to the site.

Simon Campos said...

The work you guys do is just amazing. Thank you.

Peter Jensen said...

Harry – What does 0 on the Y axis represent on your graph?

Harry Pavlidis said...

the y-axis is actual vertical movement, so a 0 would be a pitch that left the pitcher’s hand and flew straight as an arrow without departing from it’s original trajectory (no such pitch exists since no one can spin the ball enough to overcome gravity, the best four-seamers still have a few inches of drop on the way to the plate).

Joe said...

his stuff looked great…im thinking about investing in him this year.

Ed Frank said...

Darvish will make pitch classification fun

MARCH 8, 2012 BY HARRY PAVLIDIS

Harry, on your reply to Peter Jensen:

I know the point of this diagram is Pitch-type determination, but I am trying to understand the details of PITCHf/x, its data and the various ways that people display that data.

The (positive and negative directions of the) horizontal x-axis in your diagram are identical to those in PITCHf/x.

To be specific, the left-hand side of your diagram is in the direction of the catcher’s glove-hand (left) and the right-hand side of your diagram is in the direction of the catcher’s throwing (right) hand. As you say, the diagram is from the catcher’s perspective [or Point of View (PoV)].

For purposes of your diagram, the vertical “y-axis” (Peter’s term) in your diagram is actually a modified version of the PITCHf/x z-axis, which has its zero on the ground in front of the catcher at the “point” of Home Plate.

Your “y-axis” and the PITCHf/x z-axis have the same positive and negative directions – positive “upward” and negative “downward”.

The PITCHf/x “y-axis” has its zero on the ground in front of the catcher and its positive direction “toward the mound, away from the catcher”.

In other words – for PITCHf/x – (x,y,z) = (0,0,0) is on the ground in front of the catcher at the “point” of Home Plate.

Also, am I correct in assuming that the unit for both horizontal and vertical axes is “inches”?

As a reference, consider that the baseball diameter is very close to 3 inches and the width of Home Plate is 17 inches. Laying in a bold, 17-inch horizontal line indicating the (horizontal) location of Home Plate might help to answer my question below. A reference circle indicating the size of the baseball on this scale might also be useful.

Also, can I assume that the indicated pitch arrival locations are all in the “front”-door plane where each point’s PITCHf/x coordinates are (x,y,z) = (x,+17,z), where x and z both vary for each point at constant y?

OK, now I’d like to re-ask Peter’s question in a slightly different way:

Using his (x,y) notation, what is the significance of the (0,0) point in your diagram?

It seems to me that there are (at least) 3 possibilities:

(1) Your (x,y) = (0,0) is “normalized” as the “center” of the tight cluster of PITCH f/x data called the “Release Point” which – for a RHP, catcher’s PoV and expressed as absolute (but approximated here) PITCHf/x location in inches – is (x,y,z) ≅ (-12, +660, +72) or about a foot to the left of the center of the rubber, 55 feet toward the mound from the “point” of Home Plate and about 6 feet up in the air (we can quibble about these values), or

(2) Your (x,y) = (0,0) is actually normalized from PITCHf/x location (x,y,z) = (0,+660,+72); that is, centered with respect to the “point” of Home Plate.

(3) The (x,y) values for each point are actually relative to some un-specified reference trajectory. For example, a pitch thrown (with or without spin, I guess) in a gravity-free vacuum would maintain the direction of its initial velocity at the Release Point and travel in a straight line to the “front”-door.

(1) Seems more likely based on the listed pitch-type locations, since you say “… the y-axis is actual vertical movement … “

Do I have any of this right?

Thanks for your patience,

Ed Frank