Wednesday, June 9, 2010

Galarraga's Almost Perfect Game and the Flash Lag Illusion

There's been a big sports story that has been getting a lot of press lately.  Armando Galarraga of the Detroit Tigers was on his way to pitching a perfect game, that is, he had pitched 8 and 2/3 innings and hadn't let a single person get on base (no hits, no walks, just 3 batters up and 3 down, every inning).  Philadelphia Phillies pitcher Roy Halladay recently pitched a perfect game against the Florida Marlins, but don't be fooled... his perfect game was only the 20th in the history of Major League Baseball!  (definitely no small feat!)  Galarraga's would have been the 21st perfect game ever, but the 27th batter hit a little infield blooper and was called safe on a very close play by first base umpire Jim Joyce.  In the old days this would have led to a lot of debate that would have lasted for years over "what might have been", or eve "what should have been", but we live in the age of instant replay, and just seconds after the call, the instant replay showed that the runner was actually, and quite clearly, "out".  Just like that, Galarraga had been robbed of his perfect game, by a bad call.  Of course, part of the appeal of this story to the media, and to many of us watching, is the grace and sportsmanship with which all of this has been handled.  Galarraga, for his part, rather than blowing up and screaming, or complaining, simply laughed and shrugged it off, finishing the game and getting the win for his team.  And the umpire, Jim Joyce, rather than making excuses or stubbornly claiming he was right, or that it was a tough call (which it was), owned up to making a mistake and costing Galarraga his perfect game.

Well, I would like to suggest that, grace aside, there is another reason Galarraga was right not to get upset because the blown call may not have been entirely Jim Joyce's fault.  In fact, if anything is to blame, it is our flawed human perception, or perhaps Major League Baseball for not using instant replay or other computer and video asssisted techniques for officiating, but rather relying on human umpires whose perceptions are flawed.  I would like to propose this instance as an excellent example of where having a human umpire will almost invariably lead to bad calls, and hypothesize that the reason for this is a simple visual illusion that most of us are susceptible to.  That's right, Jim Joyce, take solace in the suggestion that (I believe) many umpires may have made the exact same call in your place.  And what is to blame you might ask?  The Flash-lag illusion.
We are all familiar with visual illusions.  They occur when what you perceive in your "mind's eye" doesn't agree with what you see with your actual eyes.  For example, we have all seen images like this one that appear to be moving even though everything in the image is perfectly still.

The flash-lag effect, or flash lag illusion, is a similar illusion in that the brain is "seeing" (perceiving) something different from what is actually there.  Probably the best way to describe it is to say that when you are watching something that is moving, your brain is paying attention to that movement and anticipating where the object will move to next.  In this way, your brain can compensate for the time it takes the "message" of what is being seen to travel from your eyes to your brain by perceiving that the object is slightly leading in front of where it is actually located in time and space.  Normally, this wouldn't be a problem, except when you need to compare something that is moving to something that only happens in an instant.  In the flash-lag effect, you are distracted for a split second by an abrupt, instantaneous event, usually a flash of light, and your brain tries to fill in the information about the moving object that it has missed while being distracted.  As a result you "see" the object leading its actual position, or, you see it moving farther than it actually has.  To help illustrate the point, watch the following YouTube video.  You will see a yellow dot (the "flash") appear, and momentarily block your view of the moving baseball.  After the video plays a couple of times, the still images of each frame will play in slow motion, and you will be able to check and see if the baseball is where you think it is.



(If the illusion worked, it should look like the pitch drops off to the side after the yellow dot because your brain perceived the ball as being closer to you than it really was.  If you need more of a tutorial, check out this page, or this brief video.)
Now that we understand what the flash-lag illusion is, let's try to figure out if it might explain what happened to Jim Joyce in his mistaken call.  It has been proposed that umpires, or rather referees, experience something similar to the flash-lag illusion when they are trying to make a call about  tracking something that is moving and something else that happens in an instant.  For example, in the case of Jim Joyce calling the runner safe at first, he had to be paying attention to the runner's feet moving pretty quickly as they charged toward first base, BUT, he also had to look up for a brief instant to see when Galarraga caught the ball.  This check to see when, or if, the ball was caught could act much like the yellow dot in the example above which would result in Joyce "seeing" the runner's foot lead its actual position, and appear to be over the base before it actually was.



To me, this seems like a pretty classic example of the flash-lag illusion.  BUT, in order to support this idea we would have to find a way to test this.  One way might be to go back and examine lots of footage from similar close calls and look at the mistakes that were made.  If the flash-lag effect is at play in these instances, then we would predict that disproportionately more mistaken calls would be in favor of the runner being "safe" rather than being mistakenly called "out".  In a way, this is almost exactly what several researchers have done in examining the errors made by soccer referees when calling offside penalties (1, 2, 3).  Being "offsides" in soccer means being closer to your opponent's goal line than anyone on the other team (aside from the goalie) when your team has the ball, and the ball is also not closer to the goal line than you are.  (It is basically a way to prevent players from "cherry picking" the opponents' goal.)  Two competing explanations for when referees make errors on these calls were the "optical error" hypothesis and the "flash-lag" hypothesis.  In the first case, one would predict that referees would make just as many offsides calls when players were not actually offsides (a "flag" error), as they would miss calls when a player actually was offsides (a "no flag" error).  The flash-lag hypothesis, on the other hand, suggests that referees would make more "flag" errors than "no flag" errors because movement is almost always in the same direction (toward the goal line) in an offsides call. Therefore, the player in question would be perceived to be leading his or her actual position, thus appearing closer to the goal line, and more likely to be called "offsides".  This is exactly what has been found.  The unequal proportion of "flag" to "no flag" errors suggests that the "optical error" hypothesis does NOT explain the majority of mistaken offsides calls, but rather, the disproportionate number of "flag" errors is consistent with the "flash-lag effect".  Of course, this does NOT "prove" that the flash-lag effect is definitely at work, it merely disproves the idea that "optical errors" are the primary cause, and suggests the flash-lag hypothesis as a more consistent alternative. Since this type of analysis does not rule out other possible explanations for why referees might make more "flag errors" than "no flag errors", the researchers can't say for sure whether the flash-lag effect is definitely to blame, or, say, that maybe referees are biased toward wanting to be seen as proactive and authoritative, and thus are more likely to make a call (even if it is the wrong one) rather than not make a call.  Still, it is an interesting hypothesis, with some, if not overwhelming, evidentiary support (see this article for a good description).
It certainly is interesting enough of an idea that I think it would be wothwhile to conduct a similar analysis of close calls (like Jim Joyce's) in baseball to see if we might get similar results.  Additionally, I think that baseball might add a control condition that is lacking in the analysis of offsides penalties in soccer... According to my hypothesis, the instantaneous event of the baseball being caught by the firstbaseman acts as the "flash" stimulus.  But in baseball, runners are not always "thrown out".  In some cases, like when the ball is hit between the pitcher and the firstbaseman, the pitcher fields the ball and then runs over and tries to step on the base before the runner does (so the batter is "run out" or "out run").  This condition is identical to "being thrown out" EXCEPT now, 2 very similar moving objects are being tracked and compared as they move to the same location, rather than one moving object and a "flash" stimulus. I would predict that umpires would make many more mistaken "safe" calls when players are being "thrown out" than when they are "run out", and we could even analyze mistaken calls at first second and third base (and even home plate) to control for the angle of the umpire's view (which would help us also analyze the "optical error" hypothesis).  Sounds like fun... somebody get me Bud Selig on the phone... we've got science to do.

References:

1.Baldo, M., Ranvaud, R., & Morya, E. (2002). Flag errors in soccer games: the flash-lag effect brought to real life Perception, 31 (10), 1205-1210 DOI: 10.1068/p3422


2. Helsen, W., Gillis, B., Weston, M. "Errors in judging 'offside' in association football: test of the optical error versus the perceptual flash-lag hypothesis." (2006). J. Sports Sci. 24(5):521-528.

3. Catteeuw, P., Gillis, B., Wagemans, J., Helsen, W. "Offside decision making of assistant referees in the English Premier League: impact of physical and perceptual-cognitive factors on match performance." (2010) .J. Sports Sci. 28(5):471-481.

2 comments:

  1. If you have comments or criticism about my proposed experiment, please read this post first:
    http://corticalhemandhaw.blogspot.com/2010/06/this-is-why-science-is-hard.html
    Thanks,
    -Brad

    ReplyDelete
  2. I found your experiment very interesting. I was wondering if you could point in the direction of any good books on The Flash-Lag Effect. If you just post them as a comment here I will come back and check. Thanks.

    ReplyDelete