Is Sportvision ruining baseball?

This morning the Baseball Think Factory newsblog published a piece by Diane Grassi in which she details her beefs with the use of PITCHf/x data to grade umpires and worries about the impact that forthcoming ball tracking technologies from Sportvision will have on the effectiveness of scouting.

To extent that the adoption of new technologies always results in the degradation of skills with older technologies, she probably has a point. The advent of the typewriter and the word processor have combined to deal a heavy blow to the art of penmanship amongst the masses. Better automated ball tracking will probably render some currently essential skills in the baseball industry obsolete or quaint over time. It does not follow, however, that the human element will depart the game of baseball along with it. Did the value of good writing go out the window with the quill?

In addition to my disagreement with her conclusions, I would also like to set the record straight on some errors of fact about the PITCHf/x, HITf/x, and FIELDf/x systems in her article, particular some of the erroneous facts she states in support of her argument that the use of PITCHf/x for umpire grading is fatally flawed. Her information on the Sportvision systems seems to come from an interview with Ryan Zander, director of business development at Sportvision, and perhaps an unnamed source at Major League Baseball.

She alleges that the umpires have been graded against an inconsistent system, newly introduced and not applied evenly across all stadiums.

During the 2008 MLB season, the PITCHf/x camera system was installed in every major league park – with certain exceptions made for the last year of Yankee and Shea stadiums in New York, as both the Yankees and Mets relocated to new stadiums in the 2009 season. The object of the PITCHf/x system was to gather data from the stadiums in order to composite requisite information for the camera system technology to go live in 2009.

Data was collected during the 2008 season by the PITCHf/x system that included tracking nearly all pitches thrown for the entire season for supposedly all 30 teams, totaling approximately 700,000. And that data is now being used as the base measure to evaluate MLB umpire accuracy for 2009. – Unfortunately, the umpiring data for the new Yankee Stadium and the Mets’ Citi Field was not included; unaddressed publicly by MLB.

In fact, the system was installed and brought live in all parks but two, Baltimore and Washington, during the 2007 season. This included installations in old Yankee Stadium and Shea Stadium in 2007. Baltimore and Washington were added to begin the 2008 season. PITCHf/x data for 2009 does include new Yankee Stadium and Citi Field. I can’t see any reason why MLB would choose not to include that data in the umpire grading data; if Ms. Grassi has a source that says they don’t, I’d love to know.

She then turns her argument against umpire PITCHf/x grading to allege that Sportvision and MLB don’t understand the rule book strike zone definition.

PITCHf/x takes 25 pictures of the ball in flight between the pitching mound and home plate. Sportsvision® software then uses a ‘best fit’ algorithm in order to calculate compensation for different variables of the ball’s flight path, including the position of the ball when it crosses the plate.

But here is where the disparity arises, as a strike is not called at the front of the plate but where it crosses the plate as it makes its way into the catcher’s glove. The camera, however, starts reporting data 5 feet in front of home plate; reminiscent of the ill-timed traffic light camera that incorrectly tickets a driver for going through a red light while traveling through the tail end of a yellow caution light in an intersection.

Here again she is simply incorrect. It is true that MLBAM reports the pitch location at the front of the plate for its entertainment-focused Gameday application. However, the data used for grading umpires contains knowledge of the whole trajectory of the pitch, and Sportvision’s umpire grading does take into account the 3-dimensional nature of the zone over home plate. In fact, the umpire grading system offers the umpires a measure of leniency, giving them a two-inch margin around the 3-D zone and considering factors such as the position of the catcher’s glove in counting calls in the umpire’s favor.

She also has some misunderstanding about the naming, nature, and capabilities of the two newest systems from Sportvision: HITf/x, which is the calculation of initial batted ball speed and direction from existing PITCHf/x camera footage, and the as-yet-unnamed but popularly-called FIELDf/x, which will use new cameras mounted to capture a view of the whole field in order to track ball and player movements throughout the whole game.

For after PITCHf/x, the upcoming HITf/x will be used for scouting in the not too distant future by MLB teams and it also will be a supposed tool that will measure every aspect of every player’s mechanics. Such technology will put sabermetrics to shame and will again rely upon technology which again, the naked eye cannot see on its own. “Every moving event within an actual game will be tracked,” according to Sportsvision’s General Manager of Baseball Products, Ryan Zander. It will track the pitcher, the ball and the fielder with individual stats.

HITf/x is already in existence, and the so-called FIELDf/x is coming, but neither measure a player’s mechanics. FIELDf/x measures a player’s location on the field over time.

It appears Ms. Grassi’s not quite clear on what type of scouting these systems could be used for. For scouting of players already in the major leagues, yes, whether for advance scouting of upcoming opponents or possible trade targets or coaching and improvement of a team’s own players, this system does have scouting applications. However, it has no use in the sense she uses scouting in her article, that is, finding future players like Derek Jeter on the high school ball fields around the country. Sportvision is not encroaching on the domain of the amateur talent scout.

She also seems concerned that this system uses technology that can see things the naked eye can’t see on its own, as if its secret maneuverings can be used like a hacked Diebold e-voting machine to steal an election, arbitrarily anointing good players or umpires without regard to the vast and valuable store of baseball knowledge handed down over the decades. However, these systems in fact mostly track things that the naked eye can see, like where a pitch was located, or how hard a ball was hit, or how far a fielder had to run to catch a sinking line drive.

It’s just that our naked eyes and unassisted brains are not very good at measuring and cataloging these things they see. Automated tracking systems from Sportvision allow us to remember much more accurately, find otherwise hidden patterns, and quickly query large data sets for the answers to multitudes of questions. All of this enriches the experience of baseball for many, and, I would hope, enriches the play of the game on the field as well.

Such technology does not put sabermetrics to shame; it gives sabermetricians new and powerful tools and integrates them with the flow of the game on the field in ways that were heretofore impossible and unimaginable. No longer will the accusations against sabermetricians of being a blogger in the basement or having a nose stuck in a spreadsheet hold much water. The sabermetrician in tune with these new data sources and committed to understanding the game of baseball with them will be more “on the field” than the writer in the press box. He will have the ability to gain an experience of the game as meaningful and helpful to the player as the scout sitting behind home plate. In fact, the enlightened sabermetrician will learn to converse with that scout as an equal, and the enlightened scout will enlist these new sources of knowledge to leverage his knowledge and experience of the game in new ways.

Greater collaboration between new and old, “beer and tacos” to quote Dayn Perry, will become the name of the game for successful franchises.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Padres activate Mike Adams
Next: How many home runs would Pujols hit if he was pitched to like Juan Pierre? »

Comments

  1. Nick Steiner said...

    It’s weird to hear somebody attacking the use of Pitch f/x, which in my opinion, is digitized scouting.  It’s something tangible that could be used instead of stats to take a look at player performance; I would think that guys who were against stat heads would be all over this (unless of course it’s an ego thing).

  2. Diane M. Grassi said...

    Before you discount my references, you should know that one of my references was Josh Kalk’s article from January 20, 2009 who reiterates that it was in 2008 that all MLB parks were installed with PITCHFf/x to gather data for 2009. The PITCHf/x system did not go live until this year, 2009.

    And my point about the NY stadiums, is if the data was collected in the old stadiums, it is irrelevant in terms of environmental factors. Although the dimensions of the new Yankee Stadium are the same, it has been well documented that jet streams and positioning of restaraunts, seats, etc.  are impacting ball trajectories. And Shea Stadium was completely different from the dimensions of Citi Field.

    If you have any other questions, feel free to contact me directly and I’ll be glad to take you on.

  3. Mike Fast said...

    I think you misunderstood what Josh wrote.  He said, “In 2008, a PITCHf/x camera system was installed in every major league park and gathered data on almost every pitch thrown (around 700,000 pitches).”  That’s true, and he means that 2008 was the first year that every stadium had a PITCHf/x camera operational for the whole year.

    The operational dates for the stadiums were as follows:
    April 2007: Anaheim, Atlanta, Chicago Sox, LA Dodgers, Oakland, San Diego, Seattle, Texas, Toronto
    July 2007: Arizona, Boston, Chicago Cubs, Cleveland, Detroit, Houston, Milwaukee, Minnesota, St. Louis, San Francisco
    August 2007: Cincinnati, Colorado, Florida, Kansas City, NY Yankees, Philadelphia, Tampa Bay
    September 2007: NY Mets, Pittsburgh
    April 2008: Baltimore, Washington

    Environmental factors and dimensions of stadiums have little to no effect on the pitch trajectories or ball/strike calls.  So I don’t see what relevance that has to the usefulness of PITCHf/x data for umpire grading in those stadiums.

    I don’t view this as a battle.  I just don’t understand your criticisms of the Sportvision systems or why you seem to feel they are going to be bad for baseball.  It seems to me like they are a huge plus for understanding how the game is played.

  4. theraysparty said...

    This is a new world. Technology will be incorporated into anything and everything. Baseball is the one turn based sport where taking data like this is possible and could become very important if used right.

    And agreeing with you Mike, this only hits the Major League level, possibly the high minor league levels but that is still 3-4 years before serious consideration. Young people wanting to become a Major League ball player won’t be put under this unless they work hard to actually make the bigs.

    Technology has now given us a better way to enforce and implement a near rule book strikezone. The game called behind the plate is as fair umpire to umpire as it has ever been.

  5. Hawerchuk said...

    I have to echo Nick here…This is just making the scout more data-driven.  But for those who were told there’d be no math, there’s no math!

  6. Jim Casey said...

    Anything that might push these arrogant umpires into calling a more consistent strike zone is a plus, regardless of other ramifications. When you see a rookie throw a pitch to a certain spot and it’s called a ball, and a veteran throw the same pitch to the same spot and now it’s a strike, how maddening is that? So I’m all for anything that helps standardize the strike zone.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *