A good many years ago I published on this site an article probably called The Strike Zone Winning Percentage, which outlined a very simple method to measure the extent to which a team wins the war of the strike zone.   There is a second purpose inside that, which is to measure and to highlight the relationship between winning the battle of strikeouts and walks and winning games. 

            This article repeats and expands upon that research, but also addresses another question that I was curious about.  Is controlling the strike zone still as important as it was years ago?   Or, as the game now rests more and more on hitting home runs, less on sequential offense, has drawing walks become less significant?  Since everybody strikes out 150 times a year now, is that no longer the problem that it once was?

The Formula

1)      The formula is real simple.  It is

Batter’s Walks times Pitcher’s Strikeouts

Divided by

The same thing, plus Batter’s Strikeouts times Pitchers Walks

In other words, you multiple the good stuff—the walks drawn by your hitters and the strikeouts recorded by the pitchers—and those represent your "wins".  You multiply the bad stuff—the batter’s strikeouts and the pitcher’s walks—and those represent your "losses".  

KZ Wpct =  (BatBB * PitSO) / ((BatBB * PitSO) + (PitBB * BatSO))

      Last year, for example, the Chicago White Sox batters’ drew 586 walks, while their pitchers struck out 1,588 opposition hitters.   On the other side of the ledger, the White Sox pitchers issued 485 walks, and their hitters struck out 1,389 times.  This makes a strike zone winning percentage of .580:

      (586 * 1588) / [(586 * 1588) + (485 * 1389)]  =  .580

With a .580 strike zone winning percentage, we would expect them to win 94 games.  They actually won 93 games. 

      The Miami Marlins had weaker numbers in all four categories:

Batter’s Walks:              450

Pitcher’s Strikeouts:    1381

Pitcher’s Walks:            529

Batter’s Strikeouts:      1553

      That makes a strike zone winning percentage of .431.  With a .431 strike zone winning percentage, we would expect them to win 70 games.   They actually won 67. 

      Sometimes it works; sometimes it doesn’t.  The Arizona Diamondbacks last year had a better strike zone winning percentage than the St. Louis Cardinals did, but the Snakes won 52 games, the Cardinals 90.   There are other things that matter in winning games, as well.

Predictive Significance of the Strike Zone Winning Percentage

      I figured the strike zone winning percentages of every major league team since 1900.   There are a total of 2,610 such teams. 

      Of those 2,610 teams, there have been 90 which had a Strike Zone Winning Percentage of .613 or higher.   Those 90 teams have an average strike zone winning percentage of .638.   Their actual winning percentage is .595. 

      There have been 109 teams which have had strike zone winning percentages of .387 or lower.  The overall strike zone winning percentage of those teams is .363.  Their actual winning percentage if .377. 

      This chart summarizes all of the data in that study:

            As you can see, teams in the aggregate tend to win about as many games as their strike zone winning percentage suggests that they should.   With exceptions, of course. 

            Of the 100 teams since 1900 with the highest strike zone winning percentages, 95 had winning records.   Four had losing records, and one was a .500 team.

            Of the 50 teams with the lowest strike zone winning percentages, all 50 had losing records.  It is hard to win if you don’t control the strike zone.

            My conclusion would be that, in most cases, the won-lost record of your team is essentially determined by their ability to control the strike zone.   Other things are important, but nothing else is AS important, and everything else combined is not as important. 

            Since 1990, these are the teams which have led the majors in Strike Zone Winning Percentages.  The won-lost records given are the team’s ACTUAL won-lost records, not their strike zone won-lost records.  This data is presented to help you evaluate the truthfulness of this statement:  that almost all teams which have very good strike zone won-lost records also have very good overall won-lost records:

1990   New York Mets (91-71)

1991   New York Mets (77-84)

1992   Cincinnati Reds (90-72)

1993   St. Louis Cardinals (87-75)

1994   Chicago White Sox (67-46)

1995   Cleveland Indians (100-44)

1996   Cleveland Indians (99-62)

1997   New York Yankees (96-66)

1998   New York Yankees (114-48)

1999   Houston Astros (97-65)

2000   Boston Red Sox (85-77)

2001  Oakland A’s (102-60)

2002   Arizona Diamondbacks (98-64)

2003   New York Yankees (101-61)

2004   New York Yankees (101-61)

2005   Philadelphia Phillies (88-74)

2006   Minnesota Twins (96-66)

2007   Minnesota Twins (79-83)

2008   Toronto Blue Jays (86-76)

2009   New York Yankees (103-59)

2010   Minnesota Twins (94-68)

2011   Philadelphia Phillies (102-60)

2012   New York Yankees (95-67)

2013   Detroit Tigers (93-69)

2014   Oakland A’s (88-74)

2015   Los Angeles Dodgers (92-70)

2016   San Francisco Giants (87-75)

2017   Cleveland Indians (102-60)

2018   Houston Astros (103-59)

2019   Houston Astros (107-55)

2020   Cleveland Indians (35-25)

2021   Los Angeles Dodgers (106-56)

The Timeline

            But has this changed over time?  I’m really wondering whether this has changed in the last 10-15 years, but in general. . .has it changed over time? 

            Domination of the strike zone is quite significantly more important to a team’s overall success now than it was 100 years ago.  The change is probably due mostly to the increasing number of strikeouts.  As strikeouts have increased, the importance of strikeouts as a factor in winning games has also increased.  This would be my conclusion. 

            But first, documenting the change.   I graded the "match" between a team’s strike zone winning percentage and their actual winning percentage on a 10 point scale.

            If the difference between a team’s winning percentage and their strike zone winning percentage was 21 points or less (three games, essentially), I graded that as a match of "10". 

            If the difference was 21 to 42 points, (four to six games, in most cases), I graded that a "9".

            If the difference was 42 to 63 points (.042 to .063), I graded that an "8".

            If the difference was .063 to .084, I graded that a "7".

            If the difference was .084 to .105, I graded that a "6".

            If the difference was .105 to .126, I graded that a "5".

            If the difference was .126 to .147, I graded that a "4".

            If the difference was .147 to .168, I graded that a "3".

            If the difference was .168 to .189, I graded that a "2".

            And if the difference was .189 or larger, I graded that a "1".

            Ties going to the LOWER grade.  That is, if the difference between a team’s actual winning percentage and their strike zone winning percentage was .0210000, I would have graded that a "2", rather than a "1".

            I set up the chart based on the fact that the largest discrepancy in history between a team’s actual and strike zone winning percentages was .210, so I started with that and split it into 10 units.  In retrospect, this wasn’t the best way to do it.  A better grading system would have been to use spaces with a  width of .01111, rather than .021.  Only 8% of the scores are in groups 1, 2, 3, 4 and 5 combined; only 3½% are in groups one through four.  There hasn’t been a team in Group 10 in more than half a century.  Half the chart is kind of wasted in very small groups of teams, but I didn’t know that at the time I set up the chart.   But I concluded that that mistake impacted the power of the study, but not the direction of the conclusion.   It is unlikely if not impossible that the conclusion would change if I re-did the groups and re-calculated everything, so I didn’t do that.  I will note that if I had done it the other way, Group 10 would have been teams with a disparity no more than 1.8 games between their strike zone winning percentage and their actual winning percentage.  There would have been 370 of those teams, whereas Group 1 would have been "anything larger than a discrepancy of .100", and there would have been 244 of those teams. 

            Also, the standard deviation of actual winning percentage (.082 for all teams 1900-2021) is much larger than the standard deviation of strike zone winning percentage (.064), which causes the measured gaps between actual and strike zone winning percentages to be larger than they might to be. It would be very simple to make mathematical adjustments to Strike Zone Winning percentage which would make them match up a little bit better, but I always prefer simple to complicated.   I don’t want to mess up the simple formula for KZWpct. More on this later.

            Anyway, this chart describes the number of "10s", "9s", "8s" etc. in each decade in baseball history.   G10 is "Group 10", which is the number of teams with a very small difference between their actual winning percentage and their strike zone winning percentage.

            Running your fingers up and down the chart, you can see that in the early decades, 1900 to 1949, there are almost as many teams in Group 7 as in Group 10.  Group 7 is a difference of about 10 or 11 games between the strike zone winning percentage and the actual winning percentage.  Group 10 is a difference of 3 games or less.  Up to 1950, there are about 25% more Group 10s than Group 7s. 

            In modern baseball, there are more twice as many teams in Group 10 as in Group 7—meaning that, in modern baseball, it is much less common than it was 100 years ago for there to be a significant discrepancy between a team’s strike zone winning percentage and their actual winning percentage. 

            First dividing baseball history into four periods, 1900-1930, 1931-1961, 1962-1992, and 1993-2021, the average "match score" was 7.73 in period one (1900-1930), 7.83 in period two, 8.16 in period three, and 8.54 in period four (1993-2021).  Meaning that the game is much, much more about the strike zone NOW than it was a hundred years ago.  

            This chart reports the scores decade by decade:

            That chart shows the connection heading DOWN in the last two decades, which is sort of interesting, but that’s caused in large part by including the 2020 season, when teams played only 60 games because we didn’t all want to be killed by the virus.  Playing only 60 games, the "actual" and "strike zone" winning percentages did not have the normal time to converge on a common point.  If you exclude the 2020 season from the data, the score for the 2010-2021 period is 8.52, rather than 8.42. 

Extreme Teams

            There are four kinds of extreme teams in this study:

1)     Teams that had extremely high Strike Zone Winning Percentages,

2)     Teams that had extremely low Strike Zone Winning Percentages,

3)     Teams that did far better than projected by their Strike Zone Winning Percentages, and

4)     Teams that did far worse than projected by their Strike Zone Winning Percentages. 

1.     Teams that had extremely high Strike Zone Winning Percentages.

The highest strike zone winning percentages of all time were by the Christy Mathewson New York Giants, the Giants of 1908 (.749), 1911 (.686) and 1905 (.685).  Those are the three highest numbers since 1900.   

The 1927 Philadelphia A’s are fourth; that was the team on which Connie Mack mixed several aging Hall of Famers (Ty Cobb, Eddie Collins, Zack Wheat) with several mid-career Hall of Famers and a couple of young Hall of Famers.  Maxie Bishop was also on that team; he had a strikeout/walk ratio of 28/105.

The 2017 Cleveland Indians are fifth.  Their starting rotation was Corey Kluber (strikeout/walk ratio of 265/36), Carlos Carrasco (246/26), Trevor Bauer (196/60), Josh Tomlin (109/14), with relievers Bryan Shaw and Andrew Miller striking out 187 in 140 innings.   The 2019 Houston Astros are 8^th on the list.

2.     Teams that had extremely low Strike Zone Winning Percentages

The lowest strike zone winning percentage of all time is .287, by the Philadelphia Athletics of 1918.   When the Federal League started in 1914, competing for major league talent, salaries of baseball players increased rapidly.  Connie Mack was not wealthy enough to keep his players from jumping to the Federal League, so he sold his stars to other American League teams. 

The 1918 A’s were actually the best Athletics team of the 1915-1921 era.  Their offense was reasonably good, with first baseman George Burns, rookie Jimmie Dykes at second, Larry Gardner at third, Joe Dugan at short, and Tillie Walker and Charlie Jamieson in the outfield; don’t know if you know the names, but these were long-term major league players, mostly having decent seasons.  #1 starter Scott Perry won 20 games; he was 20-19.  But Perry’s strikeout/walk ratio was 81 to 111, and #3 starter Willie Adams struck out 39 and walked 97. 

Although the overall strike zone winning percentage is constant over time, most of the worst strike zone winning percentages of all time are from that era—the 1921 Philaelphia Phillies, the 1930 Red Sox, the 1911 St. Louis Browns, the 1910 Boston Braves, the 1915 Philadelphia A’s.  The 1969 Padres—a first year expansion team—are fifth from the bottom of the list, and the 2013 Houston Astros are eighth.  That was the Astros team that lost 111 games, triggering all the loose talk about teams tanking. 

3.     Teams that did far better than their strike zone winning percentage. 

The standard deviation of ACTUAL winning percentages is significantly greater than the standard deviation of Strike Zone Winning Percentages.  The standard deviation of team’s actual winning percentages, from 1900 to 2021, is .082.  The standard deviation of strike zone winning percentages is only .064. 

In baseball history there are 11 teams which have had .700 winning percentages.  There is only one team which has had a strike zone winning percentage of .700. 

There are 59 teams which have had winning percentages of .650 or better (101-53 on a 54-game schedule, 106-56 on a 162-game schedule.)  There are only 21 teams which have had strike zone winning percentages of .650 or better.   And all of this kind of stuff is true on the other end of the scale.  There is only one team in history with a strike zone winning percentage under .300 (the 1918 Philadelphia Athletics), but there have been 21 teams with actual winning percentages under .300.

Because this is true, the .650 and.700 teams in baseball history can’t all match their winning percentages with equally good strike zone winning percentages.  The #1 team in terms of a better won-lost record than strike zone control is Casey Stengel’s last Yankee team, the 1960 Yankees. 

Stengel’s Yankees dominated their opposition in two categories:  Home Runs, and Double Plays.  In a typical year they would outhomer their opponents by 50 or more (70 in 1960), and also turn 50 or more double plays than their opponents.  In all other aspects of play, they were a pretty ordinary team.  The same is true, although a little less true, of the Yankee teams of the 1940s and the early 1960s.  That was the way they won. 

Stengel LOVED ground ball pitchers; not high fastball pitchers, but guys who threw low strikes.  His first big success as a manager was with the 1937 Boston Bees (Braves), when he pulled two sinker-ball pitchers out of the minor leagues.  They won 20 games each, as rookies—two rookies winning 20 games on the same team—despite striking out basically nobody.  Stengel made one of those sinkerballers, Jim Turner, his longtime pitching coach. 

Not ALL of Stengel’s Yankee pitchers were sinkerball pitchers; Tommy Byrne, Bob Turley and Ryne Duren are notable exceptions.  But the thing is, sinkerball pitchers are rarely effective for more than one or two years.   They have a very high replacement rate.  A few are able to stick around, but most don’t.  I always thought that this was part of why Stengel was so good with the Yankees but not successful otherwise.  The Yankees had talent to work with.  They could swing trades if necessary with their surplus talent.  They could keep the stream of new pitchers flowing.

Anyway, Stengel’s teams didn’t have great strikeout/walk ratios by their pitchers, or even good ones.  He was very tolerant of pitchers who walked a lot of batters, as long as they kept the double play in order, and he very much liked the strategy of pitching around a lefty to get to right-hander who might hit a GB6.  So several Yankee teams of that era are on the list of teams with the won-lost records much better than their strike zone winning percentages, including the 1955, 1956, 1960 and 1961 teams. 

Another team repeatedly on the list is the Cincinnati Reds of 1969, 1970 and 1975. . .just ordinary strikeout/walk ratios, but quite outstanding teams in other respects.  Second on the list of teams with winning percentages above their K/Z is the 1927 Yankees, who had a .714 winning percentage (110-44) with just a .526 strike zone winning percentage.   They struck out a LOT.  Their batters led the majors in strikeouts by a wide margin. 

4.      Teams that did far worse than their strike zone winning percentage.

Just terrible teams, none of them from the last 50 years.  (1) the 1912 New York Highlanders/Yankees, (2) the 1962 Mets, (3) the 1952 Tigers, (4) the 1907 Wshington Senators, (5) the 1965 Houston Astros.  All lost 100+ games except the 1965 Astros, who were 65-97.

Suggestions for Additional Research

There are three questions raised by this research which should be followed up.

The first is the extent to which having a better won-lost record than your strike zone winning percentage is a persistent team tendency as opposed to a transient tendency.  If it is a persistent tendency (that is, if teams that do well in this regard one year tend also to do well again), then we would ask what the characteristics are of teams that share this trait.

The second question, following from the first, is whether there is predictive value in the discrepancy.  I thought that I saw strong indications, while doing the study, that this was true.  I’ll give you two examples. 

I mentioned earlier than the 50 teams which had the lowest strike zone winning percentages of all time all had losing records.  The worst strike zone winning percentage for a team which finished over .500 was by the Cleveland Indians in 1986, who finished 84-78 despite a strike zone winning percentage pf .373.  Indian hitters had a strikeout/walk ratio of 944-456; their pitchers, a ratio of 744-605.

And what else do we remember about the 1986 Cleveland Indians?  The 1987 Indians are a famous flop.  Sports Illustrated, buoyed by the 84-78 record of 1986, picked the 1987 Indians (pre-season) to win the World Series.   They actually finished 61-101.

Is that a coincidence, you think, that a team that wildly overperformed their expectation as established by the strike zone winning percentage in one year became a famous flop the next year?

Remember my comment about the 1918 Philadelphia A’s.  Amidst a string of terrible teams, the A’s finished 52-78, easily their best record in that era, despite having the worst strike zone winning percentage of all time.  I am certain that the A’s, at the time, saw their 52-78 record as a sign of progress.  The next year they fell back to 36-104.  They didn’t make actual progress until several years later.

There are others like these, but this is anecdotal, rather than systematic.  I should research the question of whether a discrepancy between the actual won-lost record and the strike zone won-lost record is predictive of team movement in the following season.

The third question that cries out for further research is whether the connection between actual and strike zone winning percentages increases with the quality of play in the league.  I would strongly suspect that it does.

You remember that in the early part of this article, I established that the connection between strike zone and actual winning percentages has grown significantly stronger over the course of the last century.  This COULD be simply because the increased number of strikeouts has increased the importance of strikeouts.  But it is very likely that the quality of play has also increased over the last century.  It could also be that the increasing connection between the two is something that happens when the quality of play increases.

I would bet that it does.   Let’s not call it "luck"; let’s call it "space".  There is a certain space between a strike zone winning percentage and an actual winning percentage.  In that space are four things at least:  power, speed, fielding skill, and luck.  Perhaps, as the game becomes more professional, that "space" grows smaller. 

Given the right data set, it shouldn’t be difficult to tell.   We have strikeout and walk records for college teams, AAA teams, Double-A teams, and rookie league teams.   I would guess that we would find that even if the strikeout rate is the same, the connection between winning percentage and strike zone winning percentage is lower at lower levels of competition.  I would guess that that is true because, at lower levels of competition, there is just more "room" for other things to happen, more room for differences in fielding skill, more room for differences in speed 

Thank you all for reading; I will try to remember to open this up for comments tomorrow morning.