Last week I posted a couple of emails from Anthony McLean, who had cc'ed me on an open letter to the commissioner in support of the QuesTec Umpire Information System. At the end of the second email Anthony asked an intriguing question:

Finally, for whatever reason(s), MLB umpires seem to cling to a divine ordination to apply "signature zones". Conventional reporting on this issue seems to accept the devolved/de facto horizontal zone like it was an immaculate conception. Clearly there were responsible umpires, and presumably causal events for twenty five years of ignoring the rule book. Still, I've yet to see an explanation as to how this came to pass, or why the MLB umps believed this was acceptable. Have you? I'd love to hear any sort of rationale, however weak.

It was a difficult issue to get one's arms around given the dearth of information. There is no way other than anecdotally to determine if a pitch in a certain location would be considered a strike in 2003 or 1973 or 1953. First, there is the variance between umps and leagues, especially before the umpire pool was merged across leagues. Then there are rule changes over the years. The strike zone itself has been tweaked in 1950, 1963, 1969, 1988 (and there was de facto change for the two years preceding this change), 1996, and 2001 though the umpires have adhered to the varying strike zone rule in varying ways. So a waist-high fastball down the middle of the plate may be a strike at a given time according to the rules but not according to the umps.

Besides there is no record of pitch location and the resulting ball-strike call. All that we have are statistics that are indirectly a result of strike zone definition, such as strikeouts, bases on balls (non-intentional), and perhaps hit batsman, home runs, hits, and the like. It was a pickle.

I wrote Anthony to let him know that I was still puzzling and puzzing even though my puzzler was sore than I thought of something I hadn't before-more on that later. Anthony wrote back:

Thanks... In pursuing this, I've had lotsa e-mail dialogs with everyone, from Murray Chass at the NY Times...to Dan Patrick at ESPN.

No one seems to be able to provide an answer

Ah, a personal challenge: to outdo Dan Patrick!

I initially thought of looking at data starting in the mid-Seventies, but after reading this article by the great Leonard Koppett, the first part in a series on the current strike zone war, I thought staking out a thorough archeological dig of the strike zone was in order.

Where do we start? How about with the history of rules that have affected ball-strike calls, including of course the definitions of the strike zone. (Sources: MacMillan's Baseball Encyclopedia, 8th Edition; Baseball Library's Baseball Rules Chronology, and MLB's Historical Timeline of the Strike Zone. I had a more comprehensive source, possibly Bill James, on the development of pitching but I cannot find it just now.). I have broken them down by category:

Count of Balls and Strikes:

1858: The called strike is introduced.
1879: All pitched balls must be called strikes, balls, or fouls. Three strikes equal an out. Nine balls equal a base on balls.
1880: Eight balls equal a base on balls.
1881: Seven balls equal a base on balls.
1884: Six balls equal a base on balls.
1886: Five balls equal a base on balls.
1887: Four strikes equal an out (first called third strike does not count). The batter may no longer call for a high or low pitch.
1888: Three strikes equal an out.
1889: Four balls equal a base on balls.

Definition of Balls and Strikes

1880: Third strike must be caught by catcher on the fly.
1894: Foul bunts (except tips) equals a strike.
1899: Foul tip if caught by catcher equals a strike.
1901: Foul balls equal a strike except if two already called (NL only).
1903: Foul balls equal a strike except if two already called (AL adopts).
1909: Foul bunt with two strikes is an out.
1910: A pitch without one foot on rubber is a ball.
1926: Pitcher no longer credited with a strikeout on a third-strike wild pitch.
1955: Ball called if pitcher does not deliver ball in 20 seconds when a base is occupied.
1957: A strike called if batter hit by a ball in strike zone.

definition of Strike Zone

1870: The batter may call for a high or a low pitch.
1876: "The batsman, on taking his position, must call for a 'high,' 'low,' or 'fair' pitch, and the umpire shall notify the pitcher to deliver the ball as required; such a call cannot be changed after the first pitch is delivered."
High - pitches over the plate between the batter's waist and shoulders
Low - pitches over the plate between the batter's waist and at least one foot from the ground.
Fair - pitches over the plate between the batter's shoulders and at least one foot from the ground.
1887: - "The batter can no longer call for a 'high' or 'low' pitch.
"A (strike) is defined as a pitch that 'passes over home plate not lower than the batsman's knee, nor higher than his shoulders.'"
1907: "A fairly delivered ball is a ball pitched or thrown to the bat by the pitcher while standing in his position and facing the batsman that passes over any portion of the home base, before touching the ground, not lower than the batsman's knee, nor higher than his shoulder. For every such fairly delivered ball, the umpire shall call one strike."
1950: "The Strike Zone is that space over home plate which is between the batter's armpits and the top of his knees when he assumes his natural stance."
1963: "The Strike Zone is that space over home plate which is between the top of the batter's shoulders and his knees when he assumes his natural stance."
1969: "The Strike Zone is that space over home plate which is between the batter's armpits and the top of his knees when he assumes a natural stance."
1988: "The Strike Zone is that area over home plate the upper limit of which is a horizontal line at the midpoint between the top of the shoulders and the top of the uniform pants, and the lower level is a line at the top of the knees. The Strike Zone shall be determined from the batter's stance as the batter is prepared to swing at a pitched ball."
1996: The Strike Zone is expanded on the lower end, moving from the top of the knees to the bottom of the knees.
2001: No rule change but MLB asserts that umpires will call strike zone in rulebook.

Pitcher's Mound

1845: The pitching distance is forty-five feet, marked by 12-foot line.
1865: The pitcher's box appears: 12' x 3' box. Must pitch within box.
1866: The pitcher's box enlarged: 12' x 4'.
1867: The pitcher's box: 6' x 6'.
1868: The pitcher's box: 4' x 6'.
1869: The pitcher's box: 6' x 6'.
1881: The pitching distance: 50'.
1886: The pitcher's box: 4' x 7'.
1887: The pitcher's box: 4' x 5.5'.
1893: The pitching distance: 60'6" (present distance). The pitcher's box replaced by the rubber (12" x 4").
1895: The pitcher's rubber: 24" x 6" (present size).
1904: The height of the pitcher's mound established at no higher than 15".
1950: The height of the pitcher's mound exactly 15".
1968: Pitcher's mound: 10".

Restrictions on the Pitcher

1858: Can make a short run while delivering pitch. Pitch from below waist.
1863: No steps in his delivery. Both feet must be on the ground when he releases the ball.
1867: Can take as many steps as they like in their delivery.
1872: May snap wrist during delivery. Must pitch below waist.
1879: Must face the batter when he pitches.
1883: May pitch up to shoulder height.
1884: May pitch with any motion as long as not higher than shoulder height. May take one step before delivery.
1887: Must keep one foot on the rear line of pitcher's box. May take one step before delivery. Ball must be visible to ump.
1893: Must place rear foot against the rubber.
1897: Intentionally discoloring or injuring the ball is punishable by a $5 fine. The ball is replaced.
1898: Balk called if pitcher does not make throw to base after making motion to it.
1899: Balk for pickoff throws only.
1908: Forbidden from scuffing or soiling a new ball.
1920: Spitball banned. Two pitchers per team may be designated to be grandfathered in for year.
1921: 17 spitball pitchers granfathered in for their careers.
1925: Rosin bag allowed.
1940: Balk on fake throw to unoccupied base. May take two steps-one forward, one backward-as long as his pivot foot remains in contact with the rubber at all times.
1968: Balk if pitcher goes to mouth with men on base.

Other Rules of note

1845: Ball weighs 3 oz.
1854: Ball weighs from 5.5 to 6.5 oz. and is from 2.75" to 3.5" in diameter.
1858: Game set at nine innings (not ended when first team scores 21 runs).
1859: Bat limited to 2.5" in diameter.
1863: Bats must be round-no more flat cricket bats.
1868: Bats: no longer than 42".
1872: Ball weighs from 5 to 5.25 oz. And has circumference from 9" to 9.25".
1877: Home plate is relocated to its present spot.
1881: The pitcher is fined for deliberately hitting a batter with the ball.
1887: A batter hit by a pitched ball is entitled to first base and not charged with a time at bat Home plate is to be made of rubber and is to be twelve inches square. A base on balls is scored as a hit and counted as a time at bat.
1888: A hit batsman is awarded first base and credited with a hit.
1895: Maximum diameter of bat: 2.75".
1900: Home plate changed from 12" square to 5-sided, 17"-wide shape.
1910: Cork-centered ball officially adopted.
1920: Australian yarn introduced, creates "lively ball". Ball's "gloss" removed by ump.
1926: Cushioned cork-center ball adopted.
1934: Both major leagues adopt same brand of ball.
1950: Starting pitcher must go five innings for a win.
1971: Helmets required at bat.
1973: DH adopted in AL (becomes permanent in 1976).

You'll notice that aside from an odd DH rule or three, baseball has for the most part been just slightly tweaking its rules since early in the twentieth century. However, the redefinition of the strike zone every decade or so has been the one exception to the homeostasis in baseball rules, and it's more an issue today than ever. Just the fact that baseball had to assert that the rulebook would actually be followed is telling enough.

Now, Let's assume a few things. First, that a number of the changes above such as changes to the strike zone, the pitching distance, the number of balls in a walk, etc., affect the statistical record in a discernable way. You just have to know where to look for it.

If the umpires have been altering the strike zone, there should be some account of it in the statistical record. Maybe it won't be as sudden a change as a rule change, but it should be perceptible over time. Also, if de facto strike zones are adopted by umps in one league first, then there should be a perceptible difference between the leagues' statistical record.

I made a table of strikeouts, walks (unintentional), hits batsmen, hits, and home runs per plate appearance per league per year since 1871. I also calculated the percentage change per league each year and the annual differences between the leagues. For the sake of brevity-like that's a concern for me-, I'll just post strikeouts and unintentional walks here (By the way, there are inaccuracies between the pitching and batting data for pre-1962 data. For instance, no strikeouts were record for batters for the first decade of the 1900s. I have chosen the larger of the two to cast the widest net. Also, for interleague era data, I have opted to list the batting data. Neither would give a complete picture.):

OK, so it appears that strike zone definition affects the statistical record much less than other rule changes or even changes in the size or quality of the player pool.

The largest variance occurred in 1877 when home plate was moved to its current position though it was still square (K/PA increased 79.94%; BB/PA went up 50.33%). The foul strike caused huge increases in strikeouts in the NL in 1901 (59.38%) and the AL in 1903 (56.68%). The one-year switch to four strikes in 1887 caused a huge swing in strikeouts (NL: 1887 down 38.46% and 1888 up 37.29%; AA: 1887 down 35.37% and 42.87% in 1888). The largest change in walk occurrences happened in 1889 when the majors went to 4 balls per walk (65.98% in NL, 34.17% in AA), 1886 when it became 5 balls for a walk (49.65% in AA, 17.41% in NL), etc. But years like 1872 (65.98% increase in walks) and 1878 (48.95% increase in strikeouts), in which no rules changes were implemented but the quality of the play varied greatly, witnessed a great deal of variability in strikeout and walk calls.

The years in which the strike zone was redefined produced various results:

1887: Baseball defined the a unified strike zone for the first time as batters were no longer able to call for a high or low pitch. However, it is difficult to say how much the strike zone changes affected the statistical record given that baseball was still monkeying with lots of rules at once. For example, only in 1887 four strikes were required for a strikeout, the first time that a hit batsman was given first base was 1887 (at least in the NL; the AA had established the rule in 1884), and in 1887 the pitcher's box was shrunk and the pitcher was required to keep his foot on the back line of the box (thereby moving the release point back at least a couple of feet).

The end result was a huge dropoff in strikeouts (35.37% in AA and 38.46% in NL), walks held steady at least for that era (3.73% increase in AA and 7.03% in NL), and hit bastmen went up (38.02% in AA; first introduced to NL). If the desired effect was to stimulate offenses, it worked: hits per plate appearance went up 11.62% in the AA and 5.72% in the NL and homers went through the admittedly very low roof (35.18% up in AA, 44.02% in NL).

In 1907, the strike zone rule was re-written but remained basically the same. It was knee to shoulders but now the ball could pass over any portion of the plate and could not be called a strike if it bounced. The slight changes caused very little change in the statistical record: strikeouts were down slightly (AL 2.83%, NL 6.44%), walks were up ever so slightly (AL 0.73%, NL 0.75%), hit batsmen were up slightly (AL 2.35%, NL 0.39%), hits were down slightly (AL 0.21%, NL 0.66%), and home runs went up in the NL (11.84%) and down in the AL (24.88%). Overall, not much of a change occurred perhaps because the loss of the bounce strike and the addition of the "corner" strike balanced each other out.

In 1950 the big change to the zone was the loss of the shoulder-high strike. This strike zone redefinition is of special interest because it appears from the statistical record to have taken three years to become fully implemented. This may the first time that the rules and the umps had a parting of the ways at least for a short time. Another explanation could be that it took pitchers three seasons to adjust fully to the change though I think that the league stats starting to converge is a good indication that the umps in each league took extra time to arrive at a uniform strike zone throughout baseball.

In 1950 this was an increase in strikeouts but the AL's increase was disproportionately low (or the NL's was disproportionately high, depending on your point of view): a 3.18% increase to a 9.35% strikeout-to-plate appearance ratio in the AL and a 9.76% increase to 10.49 in the NL (the first year above 10% since 1916 except for a post-World War II spike in the AL due to returning players). The disparity between the leagues was the largest since the pre-Federal League 1912 season, 1.14% (again except for the one-year aberration in 1946).

The rule change affected the league differently as far as walk rates in 1950 were considered. The AL dropped 4.77% to 11.12% walks to plate appearances. This was still the second highest figure to that date for walks in a league-the highest being the AL the year before, 11.67. The unusually high and the quickly expanding strike totals in the AL may have been the impetus to redefine the zone. The NL increase by 3.57% to a 9.51% walk ratio, still 1.61% lower than the AL total.

The rule had an immediate effect on hit batsmen: a 21.39% increase in the AL and an 11.17% increase in the NL (and only a difference of 0.02% between leagues). Hits-per-plate appearance held steady (3.54% increase to 23.54 in AL and a 0.54% decrease to 23.22% in the NL: only 32 percentage points separated the leagues). Home runs went through the roof: 24.14% increase in the AL (2.00% HR-to-PA ratio) and an 18.30% increase in the NL (to 2.30%).

In 1951, almost all of the categories decreased: walks (AL down 8.57, NL down 4.08), home runs (AL down 12.63%, NL down 7.27%), hits (AL 1.95% and NL 0.36%), and strikeouts in the NL (down 5.58% though up 2.15% in the AL). More importantly the distinctions between leagues was decreasing for all ratios: strikeout difference only 0.35% from 1.14% in 19650, walks only 1.05% from 1.61 in 1950 and 2.50% in 1949, HBP only 0.01%, hits only 0.06%, and HRs only 0.39% (though up 8 points from 1950).

1952 appears to be the putative end to the mini-revolution. Strikeouts were up sharply (AL 12.61% up; NL 12.77% up-it's also the first time since the Federal League that both leagues were over 10% in same year). Walks continued to drop (AL down 4.92% and NL down 2.82%; only 0.80% separated the leagues as well).Hit batsman were way up (AL 9.24% increase, NL 8.12% increase), but hits and especially home runs were down (hits down 3.34% in the AL and 2.52% in NL; HRs: 4.99% AL decrease, 9.53% NL decrease). The end result was that the leagues were more closely aligned in terms of these ratios in a non-war year for the first time since the lively ball.

In 1963, baseball went back to the pre-1950 strike zone and decimated offenses for the rest of the decade. At least the effect was immediate. Strikeouts went way up (AL 9.70%, NL 7.95% and for the first time ever they were both over 15% in K-to-PA ratio) and walks went way, way down (AL decreased 11.79%, NL 14.13%-the NL's 6.65% BB-to-PA ratio was the lowest total in thirty years). Offenses were down slightly (except a 14.35% dropoff in HRs in the NL) and would continue to drop for another five years.

In 1969 baseball redefined the strike zone to the 1950 definition in an attempt to jump-start offenses. Home runs increased sharply: 20.86% in he AL and 35.54% in the NL. Walks did, too: AL up 18.32%; NL up 29.14%. Strikeouts dropped by 8.98% in the AL and increased slightly (0.53%) in the NL.

Bill James said in his original The Historical Baseball Abstract that the 1988 redefinition of the strike zone, which removed move of the upper part of the zone, was presaged by a de facto change that started three years earlier. I see nothing that supports a change in 1985 (strikeout up 2.25% in Al but down 2.55% in the NL; walks up 5.02% in AL, 2.55% in NL). However, home runs jumped by over ten percent in both major leagues (AL 10.50%, NL 11.18%), but that rise is still dwarfed by the one three years earlier (AL 28.02%, NL 19.52%).

In 1986, strikeouts spiked (AL up 10.37%, NL up 8.93%) and so did hit batsman (AL 18.78%, NL 11.23%). It appears that there was some sort of battle for the plate between the batters, who were hungry to join the power onslaught of the period, and the pitchers, who did want to give in to the batters.

In 1987, like two years earlier, strikeouts (AL up 2.30%, NL down 0.62%) and walks (AL up 2.75%, NL up 1.13%) stayed about the same but homers skyrocketed (AL 14.31%, NL 18.93%).

When the rule was officially changed in 1988, strikeouts dropped somewhat (AL down 6.97%, NL 3.74%) but walks dropped substantially (AL down 8.48%, NL 11.77%-at 6.86% the walk ratio was the lowest since 1968). In both leagues home runs plummeted by at least a quarter (AL 16.76%, NL 28.68%). So in the end strikeout rates were still slightly higher than before the zone was shrunken, and walks too came down. By 1988, home runs were down to almost exactly the 1984 rates in each league. Hit batsmen rates were also much higher than before the period.

This seems to be the story from the data available: home runs went way up in 1985 but the strike zone remained about the same. In 1986, home runs flattened slightly as strikeouts and hit batsmen soared. Apparently, here is where the outside strike was ceded to the pitcher to allow him to combat the homer brigade. Batters sensed this and crowded the plate to cover the outside corner, thereby getting beaned more often. 1987 was a repeat of 1985: lots of home runs but little change in the strike zone. Then the rule change in 1988 forced fewer walks and strikeouts as pitchers gave up on the high strike altogether. Batters still crowded the plate and given that and the fact that strikeouts dropped less than walks after the rule change, I am tempted to think that umps continued to give the pitchers the outside strike, ignoring the new (or any other) strike zone definition.
It seems to me that the accommodation, the outside strike, that umps made for pitchers started right there. How else would a smaller strike zone cause more strikeouts? I understand that hitters may strike out more often if they are trying to hit more home runs. However, the strikeouts stayed higher after the home runs went back to "normal". Pitchers stopped trying to get high strikes because they were dangerous. Batters crowded the plate to get to the outside strike that was ceded to the pitchers and got plunked more often.

By 1996, strikeouts and walks had both increased slowly over time. Home runs were up about 50% and hit batsmen were up about 100% from the early Eighties. Also, the strikeout and walk rates varied greatly per league (1.74% difference in strikeout rates and a 1.20% difference in walks in 1995). The 1996 redefinition appeared to be an accommodation for the existent lower strike zone. It made very little impact: AL had a 1.98% increase in strikeout and a 0.88% decrease in walks; the NL, 1.90% and 0.60%. Strikeout rates were at an all-time high, but that had more to do with a steady increase starting around 1982.

By 2001 strikeout rates had actually started to fall but walk rates were at their highest since the early Fifties, when baseball first started experimenting with the strike zone. The 2001 rule enforcement (again no rule change was actually made) had a significant impact: strikeout rates climbed by at least 5% (AL 5.21%, NL5.12%). Walk rates dropped by at least 12% in each league (AL 15.05%, NL 12.70%). This had an even larger impact on walk rates than the 1963 strike zone expansion. I would call the move a major success if not for one issue. Hit batsmen rates soared in 2001. In the AL that year a batter would be plunked more than once per one hundred plate appearances (1.05%). This was the highest rate since 1900 for any league. Also, the percent increase in hit batsmen in the AL in 2001 was 38.96%, the highest increase since 1890. The NL also experienced about a 10% increase to almost one hit batsman per one hundred plate appearances (9.86% increase to 0.97%).

In 2002, strikeouts dropped possibly die to the outside strike being taken away even more perhaps because of fear of the dreaded, computerized QuesTec system. Hit batsmen also dropped significantly (7.24 in AL and 7.62 in the NL). Home run rates have been dropping in both leagues since the 2001 rule enforcement.

So as far as I can tell, the drama that is being played out between owners, umps, and players regarding the strike zone and the QuesTec system, started in the mid-Eighties with umps ceding the outside strike to pitchers and almost simultaneously, batters crowding the plate and getting hit more often. But it seems impossible to disentangle those two events into cause and effect. It's a chicken-or-the-egg type Ghordian knot. What is clear is that as a result the batter's box was obliterated. I believe if the umpire requires a well demarcated batter's box, the hit batsmen and potentially the outside strike issues are somewhat mitigated.

The owners have become increasingly unhappy with the umpires strike zone since the Eighties. Their attempt at reconciliation in 1996 ("if we give you the low strike, will you give us the high strike and get rid of the outside strike?") failed miserably. Owners are not the kind of men who allow employees (i.e., umpires) to disobey orders. So in 2001, after a failed coupe by the umpires' union, the owners try to rein them in again. This time it's a success but not a complete one. Also, the owners are still distrustful of the umps, and start implementing a computerized strike zone guardian.

What happens next? I am left with Leonard Koppett's comments from the article above:

The whole ball-strike idea boiled down to one simple principle: A pitcher had to make a hittable delivery in fairness to the hitter, and the hitter had to swing at it in fairness to the pitcher.

What is a hittable pitch?

It seems incredible but after over 125 years of trying to codify that question baseball is still struggling to answer it. In one sense the answer is written down in black and white in the rulebook. However, as an average fan can tell you, what an umpire may call a hittable pitch varies from pitcher to pitcher, batter to batter, and inning to inning. It may even change within an at-bat.

Does this make umpires bad people or bad employees? I think not. They did make an honest effort in 2001. The fact that the owners are distrustful of them and imposing a computerized babysitter after the strides made in 2001 irks the umps doubly. Then again they did bring it upon themselves by re-interpreting the strike zone on their own at least as early as the mid-Eighties.

To me it boils down to a management issue. 2001 got them three-quarters of the way there. Now if the owners had used QuesTec to enforce the strides made in 2001 and retrain and empower (yeah, I said it) the umps instead of becoming oppositional with it, maybe they could get to the promised land , that is a strike zone according to the book. I'm afraid now that Bud and MLB's inability to properly motivate their employees will cause more bad than good as neo-Luddite umps attempt to circumvent the QuesTec system and bring the issue to a head.

It appears that the owners will implement QuesTec throughout baseball in the next few years. Whether that system can significantly improve ball-strike calling (I have my opinions) and whether umpires will continue to succumb to its use and its strictures remains to be seen.