A PitchSight To See

This week saw the professional debut of both Stephen Strasburg and Aroldis Chapman. (Although this is the third professional debut for Strasburg I’ve heard about, after the Arizona Fall League and the start of Spring Training.) Both games were available for viewing on the Internet, and no doubt well-attended. Ultimately though, the Nationals and Reds could rely only on radar gun readings, scouting reports and the game’s statistics to judge how their highly valued prospects performed.

In the future, that might not be the case, owing to the development of PitchSight. Created by L-3 Communications, PitchSight is based on the same technology that allows for QuesTec, but is designed instead for teams to develop and scout their players through all levels of baseball.

While the data available to us from technology like Pitch f/x is remarkable, it is also thus far limited to major league parks. PitchSight, designed to be installed in bullpens and other smaller venues, will allow teams not only to see that same data for all their pitchers—in both games and during bullpen sessions—but also to have a consistent, objective view rather than relying exclusively on the limited resource of scouts and pitching coaches.

I had a chance to speak with Ken Riddle, L-3’s Director of Business Development and Tom Glavine, who of course needs no introduction, about the product. Glavine was quick to emphasize that while PitchSight can absolutely supplement what pitching coaches and others see in a player, it cannot simply replace them. According to Glavine, the data provided by the program is helpful, but ultimately it is up to “guys who know what they’re seeing” to judge how near or far a player is from the majors.

Riddle explained that the company sees PitchSight as technology for major league organizations to install through the minor leagues, in addition to Division I colleges. In addition to its scouting potential, it can also be used to track those players recovering from injuries. While Glavine does not see, as I had wondered, that the tools could be used to foresee injuries, he does believe that it can aid those pitchers returning from injuries, even more so in cases when the pitchers have already used the system and have a baseline set of numbers for comparison.

image
Stephen Strasburg pitches in Spring Training, a game that PitchSight might someday cover. (Icon/SMI)

Of course, besides the obvious question of how much a system like PitchSight will replace scouts and coaches and how much it will supplement them, it raises other points. For one, prior technology to this effect has been used almost entirely in game situations. For PitchSight, which will be used to take data in practice situations, is there a risk of having conflicting data when pitchers reach “the next level” in a game situation?

Glavine, who as a 300-game winner and long-time disciple of Leo Mazzone presumably knows of what he speaks, does think there will be an inevitable difference for game performance compared to that of practice. Despite this, he believes that since a bullpen session is about consistency and improving (or maintaining) mechanics, the data will continue to be useful.

Another question is that if, perhaps literally, every pitch thrown by a pitcher in his professional career is tracked and stored in data, would pitchers be concerned about batters gaining an advantage. Glavine seemed a particularly appropriate person to ask this question, since he (along with Greg Maddux) once raised objections to the camera contained inside a catcher’s mask, for fear it would reveal his release point.

Despite this concern, Glavine thinks it unlikely the technology in PitchSight would provide hitters with an edge, especially with the amount of video that batters now can view. (Interestingly, this was the response of television producers to Glavine and Maddux’s concerns, who told the pitchers that they had “14 other cameras that show your release points a whole lot better than this one-inch camera.”)

It will probably be some time before PitchSight, or similar technology, is universally adopted in professional ballparks across America. But it is doubtless a first step in expanding the use of pitch tracking. The next time a Stephen Strasburg or Aroldis Chapman is making a debut, their teams will of course want to keep scouts in the stands and pitching coaches in the dugout. For equipment like PitchSight will also be present, providing a new set of data and making the game that much smarter.


22 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Richard Barbieri
13 years ago

As I understand it, the PitchSight is tracking the ball rather than anything the pitcher is doing. So if a pitcher was compensating, but in doing so producing the same movement, velocity, etc. on his pitches, then it wouldn’t necessarily pick that up.

On the other hand, that’s something a pitching coach presumably would, which shows how this technology goes hand-in-hand.

Alan Nathan
13 years ago

From what I know (and it is not very much), PitchSight is an outgrowth of QuesTec.  The latter is no longer used for umpire training/evaluation/etc. in MLB, having been replaced by PITCHf/x starting in 2009.  My guess is that PitchSight is really the same as PITCHf/x in that it uses fixed video cameras (at least two are needed) to track pitched baseballs.  All of us who have been analyzing PITCHf/x data for the last few years know its capabilities.  There is no doubt that anyone with enough smarts can create a similar system that would work as well, since the technology is well established.

It looks like PitchSight will be in direct competition with both PITCHf/x and TrackMan (the latter is a portable Doppler radar system for tracking the baseball) for tracking pitched balls in a “training” environment (bullpen, minor leagues, etc.).

Alan Nathan
13 years ago

Found this marketing video about PitchSight:

http://www.expandinteractive.com/2009/09/03/marketing-video-for-l3-pitchsight/

They emphasize 5 parameters that they measure that are key to pitching evaluation:  release point, release speed, location, release angle, and break.  If you look at the charts in the video, you will see plots of each of these.  To my eyes, it does not appear that the “break” values produce the nice tight clusters that we are used to seeing from PITCHf/x.  As I said in my previous post, there is no new technology being used, but the implementation of that technology might be different from PITCHf/x.  In that regard, note in the video the cartoon showing the camera locations and angles.  It looks like the field of view only includes the pitcher half of the pitcher-home plate region.  If such is the case, then such data would not be as precise as PITCHf/x for things like break and location, since that would require a large extrapolation.

If anyone out there knows more about the system, please post it here.

Tim
13 years ago

Keith: Some students at Northeastern University have developed an underarmor type shirt that monitors movement and other bio related things that reportedly will be able to see the type of compensation you are talking about. Could be a valuable tool in combination with PitchSight (can’t locate link just now…)

Frank
13 years ago

PitchSight appears to be based on the technology originally licensed to QuesTec and developed by the same company that developed it originally.  This technology was fielded years before Pitch/fx was introduced.

“There is no doubt that anyone with enough smarts can create a similar system that would work as well, since the technology is well established.”

My understanding is that it took SportVision several years and several million dollars to develop Pitch/fx, which just reproduces what QuesTec had already done.  Maybe it’s not as easy as it seems.

Alan Nathan
13 years ago

Re Frank:  For sure the implementation of PITCHf/x is very different from the implementation of QuesTec.  The latter only tracked the ball in the vicinity of home plate, whereas the former attempts to get as much of the full trajectory as possible.  As a result, QuesTec could measure pitch location very well but could not measure anything associated with the release or with the break.  Tracking over a longer spatial region is harder, since one must be much more careful with the camera calibrations, distortion, etc.  And finally, PITCHf/x has to do it all in real time whereas QuesTec did not.  So, the two systems really are different, even though the general principle is the same.  And it looks like PitchSight is much more like PITCHf/x than like QuesTec.

If I were a MLB club and were interested in using such a system (PITCHf/x, TrackMan, or PitchSight), I would want to know as much as possible about how it worked, how well it worked, what its limitations are, etc.  Due to the openness of Sportvision, we all know a lot about PITCHf/x.  Less is know publicly about the other two.

Harry Pavlidis
13 years ago

It should be noted PITCHf/x is installed in some minor league parks. Surprise Stadium and Peoria Sports Complex have data available via Gameday.

Other minor league parks have installations but the information is not published.

Frank
13 years ago

Re Allen: I think you are mistaken about QuesTec.  I was very interested in the technology at the time and they tracked the entire trajectory and did it in real time. Some changes were made for the umpire system but that was the second generation of the technology. I don’t know where you got your info but it doesn’t seem accurate.

Frank
13 years ago

Of course, I meant “Alan”, not “Allen”.

Alan Nathan
13 years ago

My information actually came from Marv White (formerly of Sportvision) and he was specifically talking about the umpire system.  I admit that I have no independent knowledge of the QuesTec system, and it is quite possible that they developed a full tracking system also.  But no matter.  The major point I was trying to make earlier was that using video technology for tracking pitches is neither new nor groundbreaking.  Mont Hubbard did an experiment at the 1996 Atlanta Olympics using such techniques (J. Applied Biomechanics, vol. 17, pp 63-76, 2001).  I am pretty sure his experiment predates QuesTec, even though the publication date may not.  And of course particle and nuclear physicists have been using particle tracking techniques for decades.

But for me, the real question is how well does the PitchSight system work.  I think we already know how well PITCHf/x works.  I would very much like to know more about PitchSight so I can form that judgment for myself.

Alan Nathan
13 years ago

No problem, Franc (er…Frank) smile

Alan Nathan
13 years ago

Frank…since you do seem to know more about QuesTec than any living person (or at least any person willing to talk about it), perhaps you could contact me privately so we can chat.  Way back in the late 1990’s, I tried very hard to get information (and even data) from the company, but was never successful.

Frank
13 years ago

Alan (got it right!): I don’t know anything more but there were plenty of articles and even a TV segment about it at the time.  This started 10+ years ago with virtual replay on TV of pitches during games. They showed the whole pitch and right after they were thrown.  People just assume it’s new. I don’t know anything about PitchSight beyond today’s press release and video you already know about. Try Google.

Alan Nathan
13 years ago

Richard:  In your article, you mentioned that you talked with Ken Riddle.  Do you have contact information for him or anyone else in the company?

KeithSchafer
13 years ago

Would something like this show the difference between the motion before and after an injury to show a pitcher where they are still compensating for that injury to help them get the pitcher back on track before another injury occurs from the compensating adjustment?

Alan Nathan
13 years ago

Peter…thanks for the tip (which I had overlooked).  I called and learned that the “Marv White equivalent” for PitchSight is Paul Baim.  A quick perusal of my old e-mails reveals that I have several interactions with him way back in 1999 in the context of QuesTec.  So, I have sent him an e-mail to try to learn more about his system.

Peter Jensen
13 years ago

Alan – Did you notice that Sportvision has finally hired a permanent replacement for Marv?  If you find out any more information about PitchSight I would appreciate it if you would send me an email.

Peter Jensen
13 years ago

Alan – There is a phone number at the end of the video on PitchSight.  Have you tried that?  Ken Riddle works at the main office of L-3 Communications in New York City if that helps.

Northern Rebel
13 years ago

One thing I love about this forum, is that the authors are accessable, and respond to comments by the readers. It shows that they respect our intelligence as fans, and I hope this trend continues as The Hardball Times, expands, and inevitably grows HUGE!

Thanks guys!

Alan Nathan
13 years ago

I talked with Paul Baim from PitchSight late last week.  He told me some things about QuesTec that I didn’t know.  The original system, which pre-dates PITCHf/x, was a 4-camera system.  Two cameras were high up looking down, as with PITCHf/x.  Two others were at field level, typically near the dugouts.  Initially the system tracked the full trajectory. I am not sure who used that information.  Broadcasters?  Teams?  For sure the data were not publicly available.  However, when used for umpires (the Umpire Information System, or UIS), it only tracked in the vicinity of home plate.

It looks like PitchSight is more or less the same as the original QuesTec system, which is more or less like the current PITCHf/x.  I did learn that the Boston College baseball team has a PitchSight system.  You too can own one for the low cost of $30k.  And I am serious when I say “low cost”.  That is very low, given what it does.

I am trying to contact the BC people to learn about the system, how easy it is to set up and use, what they are using it for, etc.