|
Win
Shares
Email:
Harold
Friend
Website: Suite101
Harold is a science teacher who
loves baseball. Actually, he is a
self described "baseball fan who
became a science teacher because
he couldn't hit or throw." He has
been involved with the New York
City Education system in various
capacities since 1962 and he
received his doctorate in science
education from NYU in 1968. He
credits Casey Stengel with being
responsible for his first
baseball "degree," and in 1998,
Joe Torre and Don Zimmer saw to
it that he received the advanced
version. For that he says, "I
cannot thank them enough."
MODERN AMERICANS
LOVE to
be told what to do. "Rugged
individualism" died in the 1930s
and we now have members on
"teams" not called the Yankees or
Dodgers but named Enron or
Metropolitan Life. How many
individuals actually comprehend
the way his computer operates or
how someone hundreds of miles
away can open the door to his car
when he loses his keys? Do you
want to lose weight? Join a
gym---excuse me---a health club.
But that's not enough. Hire a
personal trainer to watch you
ride the stationary bicycle or
run on the treadmill. The
"experts" will tell you what do
because they are experts and you
don't have the time or the
inclination to learn what they
already know.
This brings us to
what is really important. Who was
better, Willie Mays or Barry
Bonds? Until recently, we could
argue forever but not convince
the Giants fan who loves Pac Bell
Stadium that Willie really was
better than Barry. Ah, but that
is until recently. Today, in our
modern technological society
where we are happy to have
experts tell us what happened,
how it happened, why it happened,
and what to think about it, we
have Win Shares to conclude for
us that Willie was better than
Barry because Willie has 642
career Win Shares (an amount that
will not change, probably) while
Barry has 611 Win Shares.
It is very
difficult and time consuming to
determine a player's Win Shares
so it must be left to the
experts. The complex Win Shares
system translates a player's
statistics into a percentage of
his team's wins. One Win Share
equals one-third of a team's
wins.
Bill James, who
created the system, avers that
one Win Share equals one-third of
a team's wins really is the
correct value. When asked why
three to one rather than some
other value, Mr. James states
that three to one is an interval
that works. "Two to one would
work, I guess. Four to one would
work. Ten to one doesn't work,
and one to one doesn't." Nothing
beats pragmatism.
Mr. James
explains that if one player were
credited with 20 Win Shares and
another with 18, we can state
that the former contributed more
to his team than did the latter
and we would be right most of the
time, but if a value of 10 Win
Shares were assigned for each
win, it would be too difficult to
distinguish between two players
who were close in Win Shares,
such as one player with 15 Win
Shares and another with 13.
Awarding one Win Share for each
win would make it difficult to
distinguish the contributions of
a batter who hit .270 with 12
home runs and a teammate who
batted .255 with 10 home runs.
Three to one works, and if it
works, use it.
Many
years ago, a kindly old gentleman
(maybe he wasn't so "kindly")
named Connie Mack, who knew a
little about baseball, was quoted
as saying that pitching is about
90% of the game.
On page 117 of
his book, Win Shares, Mr. James
explains that offense is worth
48% of a team's wins while
pitching and defense are worth
52%. "Why do we credit 52% of a
team's success to pitching and
defense, 48% to offense?
Basically for two reasons: a) I
am convinced that it is as
logical to do this as not to do
it, and b) It causes problems if
you don't. … Even as it is, even
giving 52% of the value to
pitching and defense, our values
for pitchers and fielders still
seem low."
Mr. James has
that right. Assuming that
everything else is valid, and
that is quite an assumption,
pitching and defense are a lot
more than 50% of the game. Many
years ago, a kindly old gentleman
(maybe he wasn't so "kindly")
named Connie Mack, who knew a
little about baseball, was quoted
as saying that pitching is about
90% of the game. Logically, if
one's basic premise is invalid,
all that follows must be invalid.
Few would agree that pitching and
defense contribute only 52% to a
team's success. Where does that
leave Win Shares?
Something that
purports to be scientific must be
based on objectivity and
variables must be identified and
controlled. Two of Win Shares
basic premises are based on
assumptions. It is assumed that
one Win Share equals one-third of
a team's wins because Mr. James
states that it works and it is
assumed that pitching and defense
are worth 52% of a team's wins.
Those assumptions may be correct,
but they may be wrong. And what
if they are wrong?
An illustration
using a Tom Seaver game is an
excellent example of how great
pitching can take the game to
extremes. Granted, there are few
Tom Seavers, but let us go back
to a windy April day in 1970 in a
game against San Diego when
Seaver struck out the last ten
Padres he faced en route to a 2-1
nineteen strikeout victory. The
Mets scored two runs on four
hits. Was Mets pitching and
defense responsible for 52% of
the win?
Now, how are Win
Shares determined? It is not done
quickly or easily. Mr. James
provides an approximation of the
actual process to allow one to
save time. Let's say we wanted to
discover an outfielder's Win
Shares. The following must be
done:
1) Select the
team and figure out its Win
Shares.
2) Determine the
ratio of Win Shares credited to
the offense and the ratio of Win
Shares credited to the defense
based on runs scored and runs
allowed.
3) Calculate the
ratio of Win Shares within the
defense to pitching and
fielding.
4) Calculate the
ratio of Win Shares credited to
each defensive position.
5) Select the
player and calculate his Runs
Created, which is a sum of his
offensive contributions.
6) Figure out the
runs created by each player on
the team.
7) Determine park
effects.
8) Determine the
outs used by each player.
9) Calculate
claim points used to determine
how to distribute offensive Win
Shares.
10) Determine
individual offensive Win Shares
for each player on the
team.
11) Figure out
team fielding statistics and
relate them to the league.
12) Determine
percentages at each team fielding
position, based on several point
scales that have a total of
100.
13) Calculate
claim percentages for each team
position.
14) Figure out
fielding Win Shares for each
position.
15) Figure out
claim points used to determine
how to distribute fielding Win
Shares for each
outfielder.
16) Figure out
individual Win Shares for each
outfielder.
Not too many of
us can do that nor would many of
us want to do that. The concepts
are comprehensible to most fans
but the mathematical procedures
are not. This forces individuals
to accept the "experts"
evaluations of Win Shares. There
are dangers involved.
Win Shares
accounts for the strength of a
player's team because a player on
a team of stars that win many
games will have a greater number
of Win Shares available. The 1998
Yankees had 114 wins and 356 Win
Shares to distribute among its
players while the 1962 Mets had
40 wins and 120 Win Shares for
its players. A strong player on
the Mets would receive an
accurate number of Win Shares
since there are few good players
on a team that wins 40
games.
This is fine and
Win Shares are a valuable tool,
but let us take the case of Cleon
Jones, who had 30 Win Shares in
1969. Jones batted .340 with 12
home runs and 75 RBIs on a team
that had a .242 team batting
average and scored 632 runs. The
Mets won the Eastern Division,
the National League pennant, and
the World Series because they had
outstanding pitching. It was the
great pitching that made a Cleon
Jones single in 1969 more
valuable than a Babe Ruth single
in 1930 when the Yankees batted
.309 and scored 1062 runs because
the Mets scored so few runs and
yielded so few.
The
great Mets pitching in 1969 made
almost every Mets plate
appearance with a runner in
scoring position critical. The
same situations in 1962 were
meaningless.
If the 1969 Mets
had the 1962 Mets pitchers and
Cleon Jones had identical
seasons, he would be just as good
each year, but his contributions
would not be as valuable if he
doubled in a game the Mets were
trailing by seven runs as when he
doubled in a game that was tied.
Based on that premise, Win Shares
might be valuable to determine
what a player contributed to a
specific team, but using Win
Shares to compare players is
fraught with problems.
It is July 8,
1969. The Mets are playing the
Cubs at Shea Stadium and trail,
3-1 going to the bottom of the
ninth. Ken Boswell, batting for
Koosman (yes, starting pitchers
pitched nine inning, even when
trailing by two runs), leads off
with a double off Cubs starter
Ferguson Jenkins. Tommy Agee pops
out but Donn Clendenon hits a
bloop double to left, with
Boswell able to advance only to
third. Cleon Jones promptly
doubles home both runners to tie
the game.
It was the
situation that made the double so
important. In 1962, with the Mets
trailing by six or seven runs,
the same double would be
meaningless but would take
nothing away from Jones'
statistics. The great Mets
pitching in 1969 made almost
every Mets plate appearance with
a runner in scoring position
critical. The same situations in
1962 were meaningless.
Bill James has
created an interesting and
valuable measurement. He has
acknowledged some of its
limitations but too many
individuals, especially in the
media, extend it too far and fans
unquestioningly and uncritically
accept it because they want
"experts" to think for them. The
result has been that when one is
told that in 2003 Bobby Abreu had
28 Win Shares while Vladimir
Guerrero had 18, one tends to
conclude that Abreu is a better
player than Guerrero. That is not
Bill James' problem. We all know
to whom the problem
belongs.
References:
James, Bill &Henzler, Jim.
Win Shares. STATS Inc. March,
2002.
http://www.baseballtruth.com/extrainnings/extrainnings_030904.htm
http://www.baseballlibrary.com/baseballlibrary/submit/Hoban_Michael1.stm
http://www.baseballtruth.com/leadingoff/leadingoff_043002.htm
http://www.robneyer.com/book_03_NYN.html
http://www.retrosheet.org/
| 
