Wednesday, October 05, 2011

Learning Technology in TV Sports Viewing

Over the years I've written from time to time a critique, or a how to, or an illustrative use post about learning technology. All of these posts come from my own explorations with the technology, what I've uncovered based on that, and a reflection about potential future use. A handful of the posts still get a fair amount of hits for which I'm grateful, since it shows others do get some use value out of what I've written and that provides some sort of validation for the activity in the first place. Of course I realize that many of the hits are driven by how Google prioritizes its searches and where my post fits in on the topic in question. I view Google as a de facto dictator in this respect, but largely of the benevolent kind.

I'm going to try to make the same sort of posting based on my experience as a sports fan watching on TV. I will focus on golf, tennis, football, and baseball, simply because that's what I've watched the most as of late. It's pretty obvious that other than reviewing a particular segment of an event via the DVR, as a fan there aren't experiments to be performed this way, so my learning about the technology is not quite the same as the learning I've had with ed tech. Also let me note at the something else pretty obvious. The technology in sports viewing can be used in two different ways. First, it can be there merely to help the fan visualize what is going on. In this respect the technology has no impact on the play itself. Second, the technology can be brought in as an additional referee, presumably an entirely objective one, to help in making close calls or in reviewing calls made by human referees that are contentious.

One example of the first use come from football, where it is now standard to have virtual markers on the screen for the line of scrimmage and the first down marker. The camera angle is not perpendicular to the field, so it helps the viewer to keep those yardage markers fixed on the screen. Further the announcers are not particularly reliable in determining whether on a close play a first down has been achieved or not. The announcers suffer from the same issue of parallax as does every other viewer of the game.  The virtual lines help the viewers make their own judgments about the success of a particular play and they also make it clear what the immediate goals are for the offense.  The humans who referee the game retain discretion regarding where to "spot" the ball, but given the spot whether a first down has been achieved or not is immediately evident to the viewer.  So this particular technology is very helpful and I give it high marks for improving the fan experience.

Another example of  this type of technology use comes from golf, where Protracer now seems ubiquitous (although maybe that is because I watch the major tournaments but not much other golf).  The camera has a tough time picking up the flight of the ball and/or while in flight the ball is only slightly bigger than a speck on the screen, which makes it very hard to view.  (In contrast, the camera is excellent at showing the player's swing, which is in full.)  The new technology, by illustrating the entire path of the ball, gives much more sense of the flight of the ball, the spin that was put on the ball, and whether the player hit a good shot or made a miscue.  Like the football example, this technology enhances the viewing experience and I appreciate it as part of what is shown.

Now let me turn to an example from baseball, where I believe the technology use is less beneficial, possibly even pernicious.  The technology use reminds me of when I used to play stickball at then JHS 74, in the school yard there.  We would draw a rectangle on the concrete wall behind our imagined home plate.  The rectangle would serve as the strike zone.  We'd color the rectangle in with chalk, with the thought that if the ball hit the rectangle some chalk would get on the ball and that way we could tell if it were a strike or not, should we dispute the call.  The technology today seems to do something similar.  A rectangle is shown on the screen to represent the strike zone and an image of the ball is positioned on the screen.  Balls in the rectangle are supposed to be strikes.  Those outside the rectangle are balls.

There are two different problems with this sort of rendering.  The first issue is that the method is inconsistent with the rules of the game.   While the width of the strikes zone is given by the width of home plate and that is invariant, according to the rules the height of the strike zone is from the knees to the letters (armpits) when the player is in his usual stance.  That height is therefore idiosyncratic from player to player.  The strike zone is not a fixed entity but rather something that varies with the batter.  A pitch that is high for one player can properly be a strike for another player, and likewise for low pitches. The technology, however, treats the strike zone as fixed, not flexible. 

The second issue is that the strike zone should be conceived of as a three dimensional object, with the third dimension the depth of plate.  At issue then is whether the trajectory of the ball crosses through that three dimensional object or instead remains entirely outside it.  For a pitch with a substantial curve or sink, this can be a big deal.  The right answer depends not on where the catcher ends up catching the ball, but rather where the ball crossed the plate.  A two-dimensional representation entirely obscures this distinction.

These points would seem theoretical only except that with a fair degree of regularity the fan sees both what should be strikes as they are represented on the screen be called balls by the umpire and what should be balls as represented on the screen be called strikes by the umpire.  The consequence is for the fans to consider each umpire as highly idiosyncratic himself and possibly unfair to one team or the other.   Fans may very well have that impression anyway, but here the technology makes things worse, in my view.

Given the progress with the technology in the other sports and how much money is tied up into baseball viewing, I've got to wonder whether this situation can be improved.  The view of the pitch is from centerfield looking toward the plate.  Imagine a different view (one that is not shown on TV at present) perpendicular to the standard view that shows the player standing at the plate and in ultra slow motion shows the pitch as it crosses the plate, in particular it's vertical trajectory.  I don't know whether such a view is really possible to obtain.  The time lapse is a very small fraction of a second only.  But if we had that view, it would really help inform whether the ball is in the strike zone or not.  Absent that view, however, I'm wondering whether what is shown is helpful or misleading.

Let me turn to the case where technology is used as referee.  The best example, I believe, comes from tennis where the Hawk-Eye ball tracker is now used to determine whether a shot is in or not.  As in other instances of using technology as an objective referee, this is done only when review of the play is requested.  It isn't performed otherwise, because it is time consuming to review.  One player challenges the call and there is a quota on the number of challenges per set, so a player simply can't slow things down with bad challenges.  The incentives are there to use challenges to get the calls right.  But the technology is different in Tennis than it is in Football, because the evidence that is reviewed in tennis is a snapshot of the ball hitting the ground.  The question of camera angle doesn't come up with that.  (Though I'm not sure why not and nor am I sure why the technology seems so accurate.)  In contrast, with Football review the video is watched by the human referees and whether the video provides a good view of the play can be hit or miss.  In part this difference emerges because in football there are many potential variables to be determined - fumble, movement on the line,  the position of the ball when the knees touched down, time on the clock, etc.  In tennis, the review is used only for in or out calls.  With this more limited scope the review can be good at providing that particular determination.

I would like to see something similar used in baseball, but not for balls and strikes.  Checked swings are a better candidate, one where regularly the home plate umpire surrenders judgement either to the first base or third base umpire (depending on whether the batter is left handed or right handed).  There is a parallax problem in determining whether the bat crosses the vertical plane defined by the front of the plate.  Wouldn't technology devoted to just that question be used as Hawk-Eye is used in tennis be an improvement on what we have now?  (My impression is that the vast majority of the time the player does "go around" but it is called less frequently than that.) 

I wonder if other sports fans would agree with my assessment.  I also wonder if baseball broadcasters, in particular, might experiment with alternative technologies for judging balls and strikes.  The announcing has gotten better in the way it discusses the mindset of the players from one situation to the next (and with pitchers in the booth then suggesting what the next pitch selection should be).  The technology is supposed to provide the fan with more objective information.  It would be good for there to be some critique done (other than my post) on whether the technology actually achieves that end. 

No comments: