You’ve likely seen, by now, the call that ended last night’s Pirates/Braves extra inning circus. Julio Lugo, breaking home on a groundball put in play by Scott Proctor, appears rather clearly to be tagged out by catcher Michael McKenry, but was called safe by home plate umpire Jerry Meals, ending the game in a walkoff win for the Braves. Clint Hurdle nearly had an on-field aneurysm. The Pirates clubhouse was, by all reports, apoplectic. Twitter and forums alike exploded with excoriations of Meals, umpires in general, and Bud Selig’s continued obstinance on the issue of expanding instant replay. A funny hashtag, #JerryMealsSaysItsSafe, was born. All of this was justifiable. The call was almost certainly wrong, and it’s perfectly obvious to any baseball fan with a memory that the MLB will take no corrective or disciplinary action towards Meals, and we’ll all have to be satisfied with, at best, some empty-hearted mumblings from Bud Selig and his usual assurance that he’s talked to some of his “baseball people” that aren’t concerned about the state of officiating. Anger is a perfectly reasonable reaction.
This morning, Rob Neyer put up a post on Baseball Nation entitled Jerry Meals Might Have Been Right:
I’m sorry, but I still have not seen a conclusive replay. I’ve read a lot of Tweets from people claiming the replays or screen-captures are conclusive, but I’m looking at the same things and I’m just not seeing it. I’m not seeing a for sure in any of them.
Yes, the throw beat Lugo by 10 feet and that’s usually an automatic out. And hey, don’t we get pissed off at umpires who assume outs, just because the throw’s there in plenty of time? I do.
It might not be likely, but it’s possible that Jerry Meals saw something, something real, that none of the cameras were able to see. If there was an eighth of an inch between Michael McKenry’s mitt and Julio Lugo’s pants, would the cameras have caught that gap? Not from what I’ve been able to tell; none of the cameras were placed in just the right place to see that gap, if there was one.
Poor Neyer stumbled right out onto the proving grounds, alongside the effigy of Jerry Meals. He immediately became the new target of a heap Twitter and comment invective. I think it’s important to note, first of all, that Rob Neyer is a good writer, and an honest one. He doesn’t take positions just to be iconoclastic, to gin up page views, or really for any other reason than the search for truth in analysis. It’s almost hard to believe given the sorry lineup of national columnists getting the most attention these days, but Neyer isn’t being disingenuous, and he doesn’t have any ulterior motive. He has simply taken in all of the still images and alternate angle videos that are available (and there are many), and has failed to find some unassailable piece of proof that the tag was made. And considering just that assertion, I think he is right; we will always be missing some amount of information. Take the below image, one of the more popular screenshots of the call that I have marked up a bit:
In the upper left is a basic three dimensional Cartesian coordinate system (it’s difficult to render, but imagine the z axis as perpendicular to your monitor, pointing right at you, running along the “depth” of this picture). Thanks to Einstein, we know we exist in four-dimensional spacetime, which is to say that, to specify the exact location of any particle or point or whatever, you need to give four pieces of information: coordinates on the above-marked x, y, and z axes, which together pinpoint your position in space, and a fourth coordinate that notes your position in time. Let’s arbitrarily designate the black dot above, marking the point where McKenry’s glove appears to touch Lugo’s leg, as (0,0,0,0) — that is, x=0, y=0, z=0, and t (time)=0. This is not to say that the universe and all of existence as we know it originated about 3 feet from home plate at Turner Field, but that it’s best, for ease of reference, for us to calibrate our coordinates at that point. Anyway, for us to conclusively say a tag has been made, we must be able to prove that any part of McKenry’s glove and any part of Lugo’s body occupied the same point in this system (all four coordinates of each being equal) at any instant. If that is the case, they are touching, and it is a tag, and Lugo is out.
With this in mind, look at the above image. We know that the value of t for both the leg and glove are equal, since this is a snapshot of a single instant in time and it’s safe to assume there are no bizarre distortions of the fabric of time going on here. One down. It is abundantly clear that the glove and leg intersect on the x, or horizontal axis. Indeed, a good portion of Lugo’s leg is already past McKenry’s glove in that dimension. That’s two! It’s also pretty obvious that on the y, or vertical axis, glove and leg intersect. There are folds in Lugo’s pants that visibly rise above lower portions of the glove, if only barely. But now, attempting to audit the z axis, the “depth” dimension of this image, we don’t have enough data. It is only, after all, a 2-dimensional image, and we’re attempting to evaluate 3 spatial dimensions. The third is hidden from us, and its data is impossible to retrieve. There could be a substantial gap on the z axis between the glove and the leg, and that would be impossible for us to determine. It’s illustrated conceptually in the below image:
On the left is the perspective of our screenshot. The lines appear to intersect. On the right is what it might look like if we were able to rotate to a vantage point above McKenry and Lugo, looking down — the smallest gap on the z axis may be present. McKenry could be a tourist in Egypt, posing for the usual photo wherein he appears to be plucking the Great Pyramid of Giza from the ground with his fingertips, a parlor trick of perspective. In the format (x,y,z,t), his glove could be at (0,0,0,0) and Lugo’s leg may be at (0,0,2,0). The only way we could attempt to retrieve the z axis information is if we had two extremely high resolution screen shots, from two different perspectives, taken at the exact same instant in time. It is possible that the television crews of the respective teams have the capability to generate these, but we do not. We could approximate it by taking screen shots of different feeds, but it would be impossible to ensure with precision that we captured the same instant in time, making the data useless. It could be possible to indirectly infer the information we need — for example, if McKenry’s glove is visibly deformed it would be reasonable to conclude that the cause was Lugo’s leg, and that the hypothetical z axis gap does not exist after all. I’ve seen a few people assert that this effect is visible in the screen shot we used above. Here it is again, without my markups:
McKenry’s glove appears to be misshapen here. It looks as if part of the glove is obscured by Lugo’s leg, and that the middle portion of the glove is therefore being pushed in by the leg. This is the result of an anomalous spot on Lugo’s pants though. Look at this shot, taken an instant later. Circled is the brown spot that made it appear that McKenry’s glove was mashed up against Lugo’s leg. But now McKenry’s glove is raised substantially above Lugo’s leg, and the spot is still there. It’s either a shadow or dirt. So we can’t make any inferences from the shape of McKenry’s glove, at any rate.
The point, Neyer’s point, is that any and every two-dimensional screenshot or replay will have this same deficiency; we will never be able to evaluate with authority all four coordinates for both the glove and the leg, so we can never say for certain that they were touching. We will always lack one dimension worth of information. That being said, we can reasonably conclude, as most everyone has, that they were probably touching. I’ve arrived at that conclusion myself. That does nothing to indict Neyer’s message though, which is that Jerry Meals might have been right. Out of everyone, it was Jerry, standing a few inches from the origin, who collected the most information about all four axes. He probably interpreted the data incorrectly. But he might not have. Perhaps more importantly, there is no provision for him to review this data and reevaluate his conclusion, and there likely won’t be for quite some time.