Changes in the Baseball

Examinations of the Physical Ball

The University of Rhode Island Study

(I currently have emailed queries in to a member of the research group, and hope to have fuller data soon, but what we already know is sufficient.)

About the Study

This was a five-month study conducted by a team comprising:

• Dennis Hilliard, co-director of the Forensic Science Partnership and director of the Rhode Island State Crime Laboratory at URI
• Otto Gregory, URI chemical engineering professor
• Christopher Brown, URI chemistry professor
• Michael Platek, electrical materials engineer, URI Department of Electrical/Computer Engineering
• Linda Welters, URI professor and chair of the Department of Textiles, Fashion Merchandising and Design
• Margaret Ordonez, URI textiles professor

The team examined five major-league baseballs, one each from the years 1963, 1970, 1989, 1995, and 2000. They initially did simple bounce tests with the balls, then dissected each for detailed examination. During the whole-ball bounce tests, the two balls from 2000 and 1995 gave strikingly similar results; but those results were quite different from those of the other balls. Later tests with the bare "pill" (ball core) showed very similar patterns.

(Here's a video about the study.)

After the bounce tests, the dissected balls were given extensive laboratory examinations. URI chemists, engineers, and textile scientists closely examined various critical components of each ball to assess their possible effects on ball liveliness.

The Ball Winding

Professors Welters and Ordonez concentrated their efforts on the woolen windings. According to the Rawlings Major League Baseball specifications as listed on its web site, all three layers of windings should be woolen; there is no mention of synthetic materials in the description of these layers. But here's what Welters and Ordonez found:

That is not a trivial matter. Wool absorbs moisture very easily, and humidity can deaden a ball by relaxing the tension and reducing the elasticity of its fibers: Adair (The Physics of Baseball) once showed that balls stored in a humid environment bounce only half as high as dry balls. Polyester, on the other hand, doesn't absorb water at all, as Welters noted:

The Ball Core

Age and Study Validity

Attention was next turned to the balls' pills. Mr. Hilliard made an important statement:

It is important because it runs quite contrary to statements by Mr. Patrick Drane, assistant director of the Baseball Research Center at UMass-Lowell, which was created by and is funded by MLB. Mr. Drane has said about testing older baseballs:

If we have to choose which of those two opinions to believe, we can look at the URI team's results (the UMass team has no results from older balls).

Resiliency results

Platek and Gregory conducted another series of bounce tests on the pills; the scientists conducted about 20 drops with each. They found that the average height of the bounces from the 1995 and 2000 pills, when dropped from a height of 182 inches, was 83 and 82 inches respectively. None of the pills from 1963, 1970, and 1989 averaged over 62 inches. The newer pills each bounced at least 1/3 higher than the older ones.

So a five-year difference in age, from 1995 to 2000, seems to have had essentially no effect whatever on the pills. That does not necessarily extend to being a proof that 11 years has no effect, but it seems wildly unlikely (to me, anyway) that 5 years would produce zero effect but 11 years would produce a 25% fall-off in resiliency. In short, I know which of the two labs I believe about aging, and cannot see how, on the facts, anyone could disagree.

Gregory and Platek also did compression tests, using engineering instrumentation. Gregory reported that the older pills were indeed, as expected, stiffer (that is, less resilient--the more the resilience, the more the ball carries when hit):

To put some icing on the cake, infrared spectroscopy and digital photography were used to examine the materials of the ball cores, with up to 40 hours spent working with each of the pills. The analyses of the cork pills showed a huge variations between the three newer Rawlings balls and the two earlier Spalding ones. The digital photos showed that even the pink outer layer of the 2000-ball pill is markedly different than all the others:

In short, though the sample was small, just five balls, the margins were so huge and so consistent with what major-league power stats urgently suggest that it seems impossible for any intellectually honest reviewer to doubt that the ball has--by what way or ways is not crucial, though several seem apparent--been "juiced" in modern times. No one is arguing that any such juicing was deliberate--it very probably was not--but in complex manufacturing processes, things do happen over time.

That change in the ball from the 1977 switch to the Rawlings ball seems indubitable. So also, just on this one study's evidence, is a change sometime from 1990 to 1995--which fits perfectly the 1993/1994 time frame that actual MLB power stats point to.

The Pennsylvania State University Study

About the Study

This study comprised a series of CT (computed tomography) scans conducted by Universal Medical Systems (UMS) with the assistance of Dr. Avrami S. Grader and Dr. Philip M. Halleck from The Center for Quantitative Imaging at Pennsylvania State University. Beyond that release, there is a nice discussion of the matter available in a Jay Jaffe "Unfiltered" column at the Baseball Prospectus site.

The study took cross-sectional images of major-league baseballs produced from 1915 to 2007. The method uses the differences in x-ray absorption of various materials to create contrast and reveal density in the images. The researchers used multi- and single-slice Philips CT scanners calibrated and tuned specifically to the size and material content of baseballs.

They say they found that the density of the materials used to make official major-league baseballs has increased several times throughout history--for example, that the core density increased rapidly after Rawlings became the ball maker, in 1977. Because there is a direct relationship between increases in density and a ball's capacity to travel when struck, a denser ball means more apparent batter power (and more home runs). The researchers moreover found that the traditional cork center was at some point replaced with an enlarged rubberized core. (Pure cork is used in all minor-league baseballs.)

This study has been reported and re-reported a great deal. Perhaps the longest article about it is "Getting Inside Baseball" (Matthew N. Skoufalos, Image magazine, March 19, 2007). From it, one thing that emerges is the combative personality of David Zavagno, president of Universal and the driving force behind the study. Zavagno is and long has been convinced that the ball has clearly been juiced.

Meanwhile, MLB of course denies everything. Their mainstay in the juicing issue has been, as we saw farther above, the Baseball Research Center at UMass-Lowell, which pooh-poohs Zavagno's case.

What is a disinterested observer to believe? And why?

Zavagno may be his own worst enemy here, in that he is so vehement that his manner tends to predispose some to disbelief independent of what his actual, demonstrable facts may be. He seems to have a personal grudge, and that suggestion of bias never helps. On the other hand, the Center does not exactly come to the table with clean hands, either: it was founded by and is chiefly or wholly supported by money direct from MLB (and now also Rawlings, which makes the baseballs). Each side looks aghast at suggestions that they could be anything other than dispassionate scientists, but--as that article in Imaging weekly put it--"sometimes only the appearance of wrongdoing is enough to taint observers' opinions."

Stripped of rhetorical excesses, here is what the two sides seem to be saying.

Pro:

Con:

I think it important here to have some of the actual words used by Patrick Drane, assistant director of the UMass Baseball Research Center:

OK, help me out here. How can he say the ball is "no different" right after saying there was "a different material" with a "different density property" used at differing times? Has my comprehension of the English language deserted me here? Different = No different? Say what?

In any event, the 1999/2000 difference the Lowell lads examined completely lacks interest, being merely a straw man erected by MLB and Rawlings: no one at all is surprised that there were no perceptible differences, since even the most ardent ball-juicing advocates have nothing to say about that brief period. What is--or would have been--of interest is the 1992 - 1994 period, which is when, it is widely believed, there was a material change. Why didn't the Lowell lads look at that period?

Well, we've seen that discussion above concerning the URI study. Let me quote Professor Drane again:

Now my engineering degree is electrical, not mechanical, but am I supposed to believe that in 2000 balls from 8 years back would have experienced aging effects drastically differing from those in balls 6 or 7 years old? That's not what the URI study showed. (OK, OK, the gap there was 5 years--but would another year or three really make a sudden, drastic difference?)

Meanwhile what it sounds like is that "without any changes in method" is being used as casuistry to cover differences in materials, though I really, really, really doubt that there have been no changes whatever even in "methods", since there is testimony available that in fact there have been such changes; even if they might be minor ("might" meaning we can't be sure, owing to the bizarre secrecy that surrounds these processes, including locked "secret" workrooms), the fact of any changes rather cripples the definitive Mr. Drane's credibility.

I think that part of the "pro" camp's problem is their insistence that the balls are observably different from the nominal MLB specifications. That may well be true but also be irrelevant, in that MLB's utter inability to produce a "specification" that would not make any businessperson accustomed to technical specifications have an apoplectic fit of laughter does not mean that the balls themselves are constantly varying. As even MLB's pet laboratory noted, under MLB's specs as written, two balls both within spec could have travel distances varying by almost 50 feet under identical conditions. But, as they also noted, no real balls exhibit anything like that sort of variance (though even real balls vary by up to 7 feet in travel).

The real focus is not on matching a grotesquely ill-written spec, it is on the degree to which the ongoing usual practices of manufacturing have or have not changed in some way, possibly quite minor-seeming to the maker (Rawlings), that has some material effect on travel distance. And the evidence--mark that word, evidence--is that they have.

Summing What the Labs Tell Us

Two respectable groups of well-credentialled scientists, using state-of-the-art tools and investing substantial amounts of time, have come to pretty much the same conclusions after examining actual baseball from different seasons. And their conclusions are also pretty much in line with each other and with what the actual statistics of major-league baseball strongly suggest.

The URI study showed that baseballs from two well-separated years in the post-1993 era (1995 and 2000) were very like each other, yet very different from balls from pre-1993 years; it also showed significant categorical differences between the older Spalding and newer Rawlings balls. The Penn State study showed a marked increase in zip from 1977 on. All that agrees with the stats-indicated and common-sense indicated belief in a big jump when the ball maker was changed between 1976 and 1977, as well as the belief that there was a subsequent jump starting in 1993 and in full force in 1994.

The UMass-Lowell studies at MLB's lab of the 1999/2000 balls signifies, as expected, nothing, since there was never any indication that the ball has changed since the 1994 season began. Those much-hyped results were, whether by design or ineptitude, simply a meaningless straw-man exercise.