This is the article by Branch Rickey from LIFE magazine written in 1954. Baseball Think Factory used to have this posted but I don't think they have it any more.
But The Smithsonian has an article about it by Eric Hintz. See Branch Rickey, Baseball Innovator: By introducing several innovations, baseball executive Branch Rickey transformed America’s national pastime.
Google books also has the complete article online. It is basically a photo copy of the original magazine article.
Here is the entire article:
The 'Brain' of the game
unveils formula that statistically disproves cherished myths and demonstrates
what really wins
by Branch Rickey
As the man who guided the St. Louis Cardinals to six
National League pennants and the Brooklyn Dodgers to two. Branch Rickey,
currently general manager of the Pittsburgh Pirates, is considered baseball's
brainiest and most successful executive. He was among the first to use such
revolutionary practices as the farm system and the mass tryout camp, the first
executive to see the value of using baseball statistics in putting together and
running his teams. Now he has developed another approach to the game which LIFE
here presents.
Baseball people generally are allergic to new ideas. We are slow to change.
For 51 years I have judged basebal by personal observation, by considered
opinion and by accepted statistical methods. But recently I have come upon a
device for measuring baseballl which has compelled me to put different values
on some of my oldest and most cherished theories. It reveals some new and
startling truths about the nature of the game. It is a means of gauging with a
high degree of accuracy important factors which contribute to winning and
losing baseball games. It is most disconcerting and at the same time the most
constructive thing to come into baseball in my memory.
The formula, for I so designate it, is what mathematicians call a simple,
additive equation:
The symbols, familiar to all baseball fans, are explained in the caption to
the picture above. The part of the equation in the first parenthesis stands for
a baseball team's offense. The part in the second parenthesis represents
defense. The difference between the two—G, for the game or games—represents a
team's efficiency.
Can this bizarre mathematical device be put to any practical use? It can
indeed! It can be applied to any manor league club for any season or part of a
season to diagnose points of weakness and strength. For example, the formula
tells a strange story when applied to this year's National League pennant race.
Why are the New York Giants, who trailed by 35 games last year, leading with
the season more than half over?
The formula reveals that the Giants' offense until the last few weeks was
not appreciably stronger than in 1953. If anything it was a shade weaker. Even
with Willie Mays in the lineup the Giants were getting a few less men on base
than last year, their power hitting was off a shade. These two factors were
offset by a slight rise in ability to get base runners home with runs. But the
formula establishes that the real reason for the Giants' spectacular success
this season has been pitching an the key to pitching success is the staff's
ability to keep men who get on base from scoring.
The formula is designed principally to gauge and analyze performance on a
team basis. But certain elements in it provide a yardstick for measuring
individual talent. It can show a manager how and why certain players are
helping the team and how and why others are failing. A complete understanding
of the formula could influence player trades. It can cause an intelligent
manager to alter some features of his tactical approach to the game.
If the baseball world is to accept this new system of analyzing the game—and
eventually it will—it must first give up preconceived ideas. I had to. The
formula outrages certain standard that experienced baseball people have sworn
by all their lives. Runs batted in? A misleading figure. Strikeouts? I always
rated them highly as a determining force in pitching. I do now. But new facts
convince me that I have overrated their importance in so far as game importance
is concerned. Even batting average must be reexamined.
What Johnny Whosit hit
There are people who pride themselves on their ability to quote what Johnny
Whosit hit the year of the big flood. Among fans it is the accepted standard of
excellence at bat. Why? Principally because it is easy to figure. Even the
professionals lean upon it. But batting average is only a partial means of
determining a man's effectiveness on offense. It neglects a major factor, the
base on balls, which is reflected only negatively in the batting average (by
not counting it as a time at bat). Actually walks are extremely important. Ted
Williams, a student of batting values, bragged more about the 162 gases on
balls he got five years ago than about his .343 batting average or his 43 home
runs.
Statistics, of course, cannot tell the whole story. They fall short of
bridging the gap between human expectancy and fulfillment. They cannot measure
such intangibles as intelligence, courage, disposition, effort.
But somehow baseball's intangibles balance out. They reflect themselves in
other ways. Over an entire season, or many seasons, individuals and teams build
an accumulation of mathematical constants. A man can work with them. He can
measure results and establish values. He can then construct a formula which
expresses something tangible, and that is why this formula was devised.
It was a slow and tedious task. When LIFE first asked me six months
ago if the development of such a formula was possible, I expressed serious
doubts. In the past little effort and almost no thought has been spent on
separating the basic elements of baseball and giving them a relative value. Mr.
Connie Mack was once quoted as saying that pitching was 70% of the game. My own
feeling, before compiling and examining any evidence, was that pitching was at
least 50% of the game. I considered it baseball's most important single
department. Give me four superb starting pitchers—Christy Mathewson, Cy Young,
Walter Johnson and Dizzy Dean—and I felt I could win a pennant with Humpty
Dumpty at every other position. George Weiss, general manager of the New York
Yankees and a man whose opinion I respect, put pitching at about 35% of the
game.
What were the facts? To help assemble data that would lead to facts I
brought in Allan Roth, who prepares and refines statistics for the Brooklyn
Dodgers and who, in my opinion, is the top statistical specialist in baseball.
The problem had to be approached cautiously. It became apparent after
several false and frustrating starts that baseball statistics were indeed
puzzling. They tried men's patience. Only after reverting to bare ABCs was any
progress noted. We knew, of course, that all baseball was divided into two
parts—offense and defense. We concluded further that weakness or strength in
either of these departments could be measured in terms of runs. Thus offense
was equal to the number of runs scored by a team and defense was equal to the
number of runs scored against it. It was the only practical way to figure it.
These were the facts.
The next step was to find out how to use them. Perhaps, by taking the runs
scored per game by major league clubs over an entire season and measuring them
against opponents' runs, some pattern could be determined. We compiled these
figures for both major leagues for each season over the past 20 years. We got
other statistics and combinations of statistics. The pile became bulky and
finally overwhelming. It was obvious that it needed processing by mathematical
experts.
We took the figures to mathematicians at a famous research institute. Did
they know baseball? No, but that was not essential.
Their job was to take our figures and our guidance and, by the process of
correlation analysis, see what relation one set of figures had to another. Was
the pennant winner the club with the best pitchers or greatest batters or
surest fielders? Did the finish of a season's race reflect accurately the total
number of hits or runs made by a club?
After six weeks the findings came back. Among them was
one which constituted a framework around which to build a formula. The
mathematicians discovered that by subtracting opponents' runs from the runs
scored per game by a team over a season they got a column of figures which
correlated strongly with the final standings. In the National League last year,
for example, it related to the order of finish as follows:
In other words offense (which we shall refer to as O) minus defense (which
we shall refer to as D) indicates the number of games a team was behind the
leader. The equation: O-D=G—the condensed version of the big formula shown at
the beginning of this article. It proved 96.2% accurate when applied to final
major league standings over the last 20 years.
This, of course, was just the beginning. By using O-D=G as a guide, it was
possible to jump off into unexplored territory, testing the footing to learn
where we were on solid ground and where we sank in. If we could separate the
measurable component parts of offense and relate them one to another, we would
have half of the formula. If we could do the same thing with defense we would
have all of it.
What were the factors that went into the scoring of runs? There were a
barrel of them. There was ability to get on base by any means possible. There
was power, the ability to hit for extra bases rather than just singles. Then
there was speed, daring on the base paths, timeliness of hitting and making the
most of opportunities.
Of course, there were the unmeasurable factors. In my experience probably
the most important single thing in batting has been the mental attitude of the
hitter going to the plate. The most gripping moment in any field of sports
comes when batter faces pitcher. Batter and pitcher eye each other.
Psychologically one or the other is in command before a ball is thrown. But can
you measure this? Could you measure the arrogance of a Rogers Hornsby as he got
ready to take his cut. Walter Johnson's utter indifference to the identity of
any batsman who ever faced him?
Eyesight is another variable factor. Babe Ruth's eyes were so quick, it is
said, that he could read the label of a phonograph record while it was
spinning. I suppose there is a way to get an optometrist's rating on hitters'
eyes, but that could not help fix a formula.
But the ability to get on base, or On Base Average, is both vital and
measurable. It is determined by this:
(We have added BB and HP to the bottom part of the fraction because they are
not included in official times at bat in present baseball statistics.) For
example, Stan Musial's OBA for last years was
This means that he got on base 43.7% of the times he faced the pitcher. This
is a far more significant figure than just batting average alone, which was
The virtue of this approach is that it gives a positive evaluation to walks.
A few years ago Virgil Stallcup walked only six times in an entire season or
once every 72 trips to the plate. Compare this to Eddie Stanky who set a league
record with 148 bases on balls in one season, or once every 4 ¾ appearances.
Although we did not try to evaluate the base on balls statistically, I would
rate it about 75% of a base hit—because a base on balls can never advance a
runner more than one base and cannot advance him at all unless there is a
runner on first.
The next measurable quantity is Extra Base Power. There are several ways of
computing this. The conventional slugging average which most baseball people
use is simply total bases over times at bat. My own formula for computing
power, which I have used for years, and called "isolated power," is
the number of extra bases over and above singles in relation to total number of
hits. It turns out that neither of these is as reliable in determining extra
base power as a formula which combines the two. This new power quotient,
determined by
is the percentage of extra bases per times at bat. For Stan Musial last year
it came out
It compares with a National League or .145 and cannot be excluded from any
formula for offense.
Now that we have men on base how do we get a formula for getting them home?
It must include speed, taking the extra base, stealing one occasionally,
managerial savvy and timeliness of hitting. All these capabilities,
unmeasurable in an individual, are reflected by one statistic which has never
been used to my knowledge and which I shall call "clutch." It is
simply the percentage of men who got on base who scored.
It is determined by dividing actual runs by the number of opportunities to
score, or base runners.* The method for determining this is
It shows that Brooklyn was by far the best in the big leagues in this
department last year with
or 43% of its base runners scoring. It is a clear reflection of the club's
general efficiency. By comparison the inept St. Louis Browns scored with only
30% of its base runners.
These were the three basic, measurable ingredients for offense: on base
average, extra base power and clutch. But how did they fit together? Were they
all equal in importance? To find out in what proportions they did add up to
runs scored we measured them against major league records for the last 20
years, seeing how one related to the other in actual offensive results. The
correlation showed that OBA went hand in glove with runs scored. When one was
high for a team, so was the other. Clutch was just as strong, but extra base
power had a lower correlation. In relating each of these ingredients to each
other in the indicated proportions we had to give extra base power less
importance. So offense became
Was this the formula for offense? It was easy to find out. We went back over
the record of every major league club of 20 years substituting figures from the
official averages for symbols and working out totals. These totals correlated
with runs per game for each club almost perfectly. The margin of error was 2%.
There had been some arbitrary decisions and values adopted in putting the
formula together, but checking back proved that the assumptions had been
correct. Since runs scored are the final expression of offensive strength, it
looked as if we had our offensive formula.
Unfortunately there was no way of applying all three of these basic factors
to individuals as well as teams. Clutch was strictly a team figure, but after
giving it a thorough trial we found there was still no place for RBIs in the
formula. As a statistic, RBIs were not only misleading but dishonest. They
depended on managerial control, a hitter's position in the batting order, park
dimensions and the success of his teammates in getting on base ahead of him.
That left two measurable factors—on base average and power—by which to gauge
the over-all offensive worth of an individual. We applied them to some of the
game's greatest hitters and arrived at a rating for each.
How the greatest hitters are
rated
Who came out as the greatest of all time? There could be only one. Babe Ruth
came out head and shoulders over his closest rival. If he had not I would say
without hesitation that we were on a cold trail with our formula. The ratings
on the opposite page are based on composite records since 1920, the year the
lively ball came into use.
I know that the order in which these hitters are rated is apt to shock some
students of batting. I was shocked myself. I found it difficult to believe that
only one American Leaguer among the first 25, Ted Williams, was still active. I
found it equally difficult to believe that the National League still has four
going strong Kiner, Campanella, Musial and Robinson. Admittedly there were
discrepancies in the ratings, partly due to park dimensions.
The list shows some great batters high in OBA and low
in power. It is unfair to great performers like Tris Speaker and Ty Cobb, and
it does not even include George Sisler, one of baseball's best batsmen, since
some of their best years came before 1920 in the era of the dead ball. Cobb
deserves to be higher because he beat you with more than his bat. He stole more
bases per season than an entire team does today—and stolen bases are not
included in the above table. He beat you with brains, aggressiveness and
opportunities, all the things that show up in clutch which we cannot
estimate for an individual player. When Cobb got on base his very presence
there upset the pitcher. It caused the infield to make errors it would not
otherwise make. Jackie Robinson of Brooklyn has the same nuisance value to a
lesser degree today, and as a runner on third he has no equal.
Now that we had the formula for offense under control, what about defense?
There were two factors involved, pitching and fielding, and only one of
them—pitching—was measurable on either a team or individual basis. There is
nothing on earth anybody can do with fielding.
Fielding averages? Utterly worthless as a yardstick. They are not only
misleading, but deceiving. Take Zeke Bonura, the old White Sox first baseman,
generally regarded as a poor fielder. The fielding averages showed that he led
American League in fielding for three years. Why? Zeke had "good
hands"! Anything he reached, he held. Result: an absence of errors. But he
was also slow moving and did not cover much territory. Balls that a quicker man
may have fielded went for base hits, but the fielding averages do not reflect
this.
Fielding then cannot be measured, although it must be admitted that, all
other things being equal, it could be the difference between winning and losing
four or five games or mean the run that wins the big game. But application of
the formula to 20 years of statistics shows fielding to be worth only about one
half as much as pitching or 15%. No team would have an aggregate fielding
efficiency of 1.000. So the variation within the 15% between the best fielding
team and the worst would be only a few percentage points and does not destroy
the general validity of the formula.
But pitching! There's something a man can get his teeth into. If there is
any one phase of the game on which I might consider myself an expert it is
pitching. Poise, control and stuff are three important assets possessed by
great pitchers. And what about his skills? What can he make a baseball do? Does
one count for more than another?
Walter Johnson and Preacher
Roe
There have been pitchers who have gotten by without one or the other of
these basic requirements, but they would have been exceptions in any league and
they are not the best. I doubt if Walter Johnson ever gave a thought to
pitching to spots, the stock in trade of current pitchers. He just wound up and
fired. If you have a strong back and a buggy-whip arm you don't have to think.
Preacher Roe, Brooklyn's oldest and most artful pitcher of recent years, is the
other extreme. Roe doesn't have much speed, but he has variety. He has control
and poise. He sets batters up to swing at a certain pitch. He keeps them off
balance and guessing.
Whatever combination of these three qualities a pitcher has, the total is
reflected by one standard—Earned Run Average. This is the number of runs,
exclusive of those made because of errors, that are chargeable to a pitcher.
There have been attempts to substitute for ERA, but none of them has been
convincing or successful. We add nothing new to ERA in our formula. But we do
learn a good deal by breaking it down. The illuminating chart on the opposite
page, which compares lifetime records of eight great ERA pitchers in baseball
today, is an example. It presents a comprehensive picture of pitching equipment
divided into specific categories.
Some features of these figures gave me a great surprise. They showed me that
the difference between a great pitcher and a run-of-the-mill thrower is
astoundingly slight. The great ones allow only two less men to reach first base
per nine innings. More surprising still, I found that the ability to strike
batters out was not a determinant of good pitching in the real sense. As you
can see, Eddie Lopat fell below average in this department and Robin Roberts
was not far above average. The strikeout kings like Bob Feller and Hal
Newhouser were not strong in the bases-on-balls column.
It turned out too that strikeouts had an extremely low, correlation figure
when measured against earned run average. All right, if strikeouts were not
important in the over-all scheme of pitching, what was? The evidence showed
that hits allowed and walks given up were strong determinants. So was keeping
base runners from scoring. Among the great pitchers, weakness in any of these
departments had to be compensated by strength in another. If Mike Garcia put a
goodly number of men on base, he compensated by not allowing many to score. If
Robin Roberts allowed a larger percentage to score, he compensated with an
exceptionally low bases-on-balls record, which kept his men-on-base average
down.
The problem was to put these variables together in a pattern which would add
up to the final expression of pitching strength as interpreted in earned run
average. We finally got one. In effect, it was roughly the reverse of the
formula for offense. First, there was percentage of hits allowed or the batting
average against a pitcher. It is determined simply by dividing hits by the
number of times at bat or
If all pitchers in the American League had been as effective as Billy Pierce
of the Chicago White Sox last season, the batting average of the league would
have been a lowly .218, for that was the batting average against him. Pierce's
totals were
To the batting average against a pitcher we add the percentage of men who
got on base because of walks and hit batsman, or
This was a means of measuring control. The great control artists such as
Lopat walked only one batter out of every 20 or 25 that faced them last season.
Lopat's figure was
We came then to the clutch figure for pitching, the percentage of base
runners scoring earned runs. The method of calculating was
A good figure in this department was
the principal reason why Ed Ford of the Yankees had an excellent earned run
average last year. He was the most effective clutch pitcher in the American
League with
He allowed only 22.9% of the men who got on base against him to score. The
average in the league was 31%.
It turned out that hits allowed, walks allowed and clutches were of equal
importance. But not strikeouts. After examining all the evidence, I was forced
to admit, and I did so grudgingly, that strikeouts contributed nothing more to
the end result than pop fouls caught by the catcher. On the basis of statistics
we valued that strikeouts at one eighth of the other three more vital factors
and got a pitching formula which read
It contains all the measurable components of pitching. The result correlates
with ERA to a high degree, the margin of error being little more than 2%. As
with ERA the lower the figure, the more effective the pitcher. It worked as
well last year as it did five years ago and when applied to the averages, since
1920, of the 12 great pitchers listed on the next page it was just as reliable.
We now have an instrument for determining the value of elements which go
into the two basic departments of baseball. We can examine with sharper insight
the performance of a team or individual over a given period. This knowledge can
be used to detect flaws that would not otherwise be noted, to give a proper
balance to baseball forces, to rearrange batting orders intelligently, to
pinpoint problems in pitching. Although the formula gives a comprehensive
diagnosis of teams and players, it has limitations. It cannot predict the
performance of a team on any given day or in any brief series because players
have good and bad days. Nor can it foresee with accuracy the outcome of a
pennant race because players do not always live up to past performances. But
the formula is a valuable tool for analysis and just think of what it will do
in the hot-stove league arguments.
There was one more important step to take. We had formulas for offense and
defense, and we knew the difference between them was very close to the order of
finish in the league standings. But were offense and defense of equal value in
determining the final result? Through the years I have felt , along with the
rest of baseball's old guard, that defense was infinitely more important than
offense. Once again I was faced by facts and forced to reverse my way of
thinking. The figures showed that offense has gradually taken over the game and
has become more important in winning pennants than defense. For the last 10
years in both major leagues the ratio of importance for pennant winners was 54%
for offense and 46% for defense, with pitching about 30% of the game.
When George Sisler, the great oldtime first baseman and author of Sisler
on Baseball, first saw the figures his reaction was one of bewilderment.
"I still don't believe it," he said. "But there it is." And
there it was, cold and irrefutable. Brooklyn, which ran away with the National
League race last season, won with offense. It scored 6.16 runs per game or 1.41
more than average for the league. But on defense, with 4.45 runs per game
scored by opponents, it was only .31 better than average. Milwaukee was the
standout on defense last season, allowing only 3.75 runs per game, but its
offense was too weak to make up the difference.
Year by year the pendulum has swung back and forth between offense and
defense. Defense won for the Boston Braves in 1948 and for the Philadelphia
Phillies in 1950. In the American League where the New York Yankees have been
on top for five straight years, they won with a preponderance of offense three
times and with defense the other two. But mathematical calculation shows
offense clearly in command over the past decade.
Had it always been thus? Not by any means. In the old days defense was
clearly dominant. The evidence simply pointed up how violently the game had
changed. I can even name the year the great change commenced. Anyone can. It
was 1920, the year the hopped-up ball with the rubber-cushioned center and
tighter-wound Australian wool came into use and the year that Babe Ruth slammed
54 home runs after having only hit 29 the year before.
That started it. Ruth revealed that the home run was good box office.
Coincidentally the rules of the game were changed. Spitballs and trick
deliveries were outlawed to lessen the effectiveness of pitchers. Fences were
here and there moved in toward home plate to put them within range of more
hitters. The stolen base deteriorated as an effective tactic. Pitchers in order
to cope with the increased scoring violated the balk rule and continue to
violate it with umpires making almost no effort to enforce it. The pattern of
the great change could almost be plotted by the constant shrinkage in stolen
bases and the steady upward surge of the homer. At its extremes it was as
follows:
Everybody in baseball was aware that the complexion of the game was
shifting. But few of them stopped to ponder how it affected the game's basic
metabolism. The change was slower in the National League than in the American.
The National League was still a "pitchers' league" until after World
War II, with teams still playing for one run, and the records disclose that as
recently as the period from 1934 to 1943 defense was still dominant in the
National League. It predominated over offense by 60-40. The offense caught up
with the defense in the National League and pushed ahead, just as it did in the
American League more than a quarter of a century ago.
The formula, applied to the stars in both leagues this season, discloses
that baseball's big bats are mostly in the National League. In the recent All
Star game in Cleveland the lineup for the National League outweighed that of
the American, according to the formula, in both hitting and pitching.
The National League, whose top batter was Brooklyn's Duke Snider, had five
men with a combined on base average and power rating of .600. The big five were
Snider (.701), Willie Mays (.686), Stan Musial (.677), Gil Hodges (.623) and
Ted Kluszewski (.621). The American League had only two—Ted Williams (.733) and
Mickey Mantle (.600)—and the average for all American League All Star hitters
was .525 compared to the National League's .551. In pitching the National
League had the edge .557 to .575.
On cold dope the National League should have won instead of losing. But if
these two squads played 154 games the formula says it would have been a
different story. One game is too short a test, especially one like the All Star
game which is an exhibition and not a blood-letting contest for league
supremacy. Even so, this one might have had a different outcome if the umpires
had seen Stone balk and had called the balk rule properly on Virgil Trucks.
Failure to call a balk when Red Schoendienst tried to steal home in the eighth
inning was crucial. Instead of being a run for the National League it was the
third out. The Nationals lost, upsetting both the form and the formula, but
that detracts not one whit from the formula's potency.
Now that I believe in this formula, I intend to use it as sensibly as I can
in building my Pittsburgh club into a pennant contender. What is wrong with the
Pirates? The formula opened my eyes to the fact that the Pirates' OBA is almost
as high as that of the league-leading New York Giants. We get plenty of men on
base. But they stay there! Our clutch figure is pathetically low, only .277
compared to New York's .397.
This could give reason for a change in the batting order, a closer grouping
in the batting order of the club's high OBA hitters. We have to get hitters who
can raise the clutch figure. Where are they going to come from? We may have
them on the farm team in Hollywood or New Orleans or Denver. My purpose is to
raise a crop of players and this thing puts a hoe in my hands and my scouts
must indeed use their eyes to find more power for clutch.
This study has been a series of surprises for me. I repeat: baseball
people—and that includes myself—are slow to change and accept new ideas. I
remember that it took years to persuade them to put numbers on uniforms. I know
a manager who still believes that iodine is the panacea for sliding burns. It
is the hardest thing in the world to get big league baseball to change
anything—even spikes on a pair of shoes. But they will accept this new
interpretation of baseball statistics eventually. They are bound to.
Goodby To Some Old Baseball Ideas by Rickey, Branch
Article in Life
August 2, 1954 (Vol. 37, Issue 5) -- p. 78, 10 page(s)
