Five percent of all the major league teams in history committed no balks at all during the season, and another 10% committed, as a team, only one balk, although how one can commit a balk as a team I am not exactly sure.  Another 14% of teams committed two balks on the season, which is actually the most common number for Balks by a team—two. 

            The 18 highest balk numbers of all time were all compiled in the same season, 1988.   Want to feel old?   The 1988 season was (a) after I had finished writing the annual Abstracts, and (b) NOT in the last quarter of the time period covered by this study, 1900 to 2019.   120 years covered by the study; the 1988 season was not in the last 30.    In 1988, as probably most of you know, there was an effort to clarify what was and what was not a balk.  Umpire Ron Lucchesi had said publicly that he never called a balk because he had no idea what a balk was, and there had been occasional arguments about whether that was a balk or not.   The rule-jiggerers added a few words to the balk rule, to clarify it, but the effort backfired, and the major league record for balks in a complete season (all teams) was broken by mid-May.  

            The average for all teams since 1900 is five balks.  Processing balks for this system is on the one hand simple, since the data is straightforward and the process well established, but it is in a sense fatuous, since the system is set up to deal with significant numbers of events, and 29% of teams have zero to two balks.   Stating the data as Balks per Batter Faced, because that is how we have been doing these things, the average in the era 1900 to 1909 was .00029, or one balk for every 3,420 batters faced.    Balk rates then increased slowly but steadily until 1988, and have decreased slowly but steadily since 1988.

            The standard deviation of such small numbers is a yet more fatuous number, but, you know. . .it’s part of the process.   These are the Balk frequencies and standard deviations for each decade:

              Since there are teams in every decade with zero balks, the high-end scores are limited by the ratio of the norm to the standard deviation.   In the 1900-1909 era the standard deviation was almost the same as the norm, so a team could only be one standard deviation better than the norm, even if they committed no balks.  In the last two decades the norm is essentially twice the standard deviation, so a team that commits no balks is (essentially) two standard deviations better than the norm, or scoring at 120.   But it is such a minor and un-noteworthy accomplishment that I will spare you the list of teams which have done it. 

            On the other end is the 1988 Oakland A’s, a team which committed 76 balks—19 more than any other team in history.  They are 6.1 Standard Deviations worse than the decade norm:

            6.1 standard deviations below the norm is by far the worst performance in any area within this series.   It’s weird data, kind of fake data.   The ultimate issue is "How many runs did the pitchers allow—how many did they fail to prevent—by this weakness?"   The numbers here are so small, even in the extreme cases, that there isn’t a great deal on the line.   Balks are less closely connected to wins than anything else we have studied:

            And almost all of that connection is incidental, rather than direct.   The difference between the low-walk group and the high-walk group is just 9 balks a season, just 9 bases.  The difference between the groups in wins and losses is 5 games.   Nine bases cannot conceivably lead to five losses.  It’s just that good teams tend to be good at everything, so they tend to be good at not committing balks, too. 

            The Balk Rule is the worst rule in baseball.  It’s intended to protect the baserunner, to give the baserunner a fair chance to get a lead and steal a base.  The essence of the rule is that, with a runner on base, the pitcher must come to a complete and visible stop in the middle of his delivery.  No, that’s not right; it actually requires a SECOND complete and visible stop, after taking the "stretch" position.    

            But that plays hell with the pace of the game.   The rule should be that if the pitcher has the ball and is on the mound, he can pitch—period.   He can throw from any angle, he can throw at any time, and no nonsense about coming to a stop in the middle of his delivery.   There is no reason to require or describe a "stretch" position.  If the batter is not in the batter’s box, that’s his problem.  If he’s not ready to hit, that’s his problem.  Get ready.  Stay ready.  If the baserunner isn’t ready, that’s his problem. 

            And the same for the baserunner—he can break any time that the pitcher has the ball and is on the mound. If the pitcher’s not ready, that’s his problem.   It adds a lot more tension to the game if that’s the rule. 

            Sure, you have to do SOMETHING to protect the baserunner.   You just don’t have to anything that is remotely this intrusive into the game, in order to protect the baserunner.     Thanks for reading.