$Unique_ID{BAS01287}
$Pretitle{}
$Title{Appendix:  Phantom Ballplayers}
$Subtitle{}
$Author{
Kachline, Clifford S.}
$Subject{Phantoms Phantom Ballplayers Olsen mystery boxscores boxscore typos
typo}
$Log{}

Total Baseball:  Appendixes


Phantom Ballplayers

Clifford S. Kachline


Garbage in, garbage out is an expression that gained currency with the advent
of the computer age. The logic behind the catch phrase, however, prevailed
long before the electronic marvels came into being. Baseball, in fact, has had
its own version of the maxim almost since the game's earliest days, largely as
a consequence of recordkeepers who sometimes unwittingly entered erroneous
data into the record books.
     Numerous examples of the garbage in, garbage out principle have been
discovered in baseball's statistical archives through the years. But
statistics aren't the only area where the phenomenon has shown up. Another
involves what researchers of the sport refer to as "phantoms"--players
credited with having performed in the major leagues but who in reality never
did appear in a big league game.
     Phantoms are hardly a recent phenomenon. They have existed almost as
long as boxscores have been published. Some were the product of
misunderstandings by--or misinformation given to--official scorers or the
parties who compiled the boxscores. A few of these crept into the leagues'
official records. Others were created by mistakes on the part of telegraph
operators or by typographical errors and appeared only in newspaper boxscores.
     Little if any attention was paid to the situation until the 1950s. The
original edition of The Official Encyclopedia of Baseball, published by A.S.
Barnes and Co. in 1951, provided fans for the first time with a supposedly
complete alphabetical tabulation of every man who ever played in the majors,
together with his basic yearly big league stats. The publication stimulated
the interest of the sport's researchers. When editors Hy Turkin and S.C.
Thompson deleted the names of some players--most of whom were previously shown
as having a one-game career--from subsequent revised editions, the matter of
phantoms started to become a source of fascination.
     Who really were the "impostors" that were listed in earlier editions?
What had prompted the editors to include them in the first place? And what
evidence had been found to prove conclusively that they never played in the
big leagues?
     The introduction of The Baseball Encyclopedia by the Macmillan Publishing
Company in 1969 focused additional attention on the subject. In compiling
data for that publication, David Neft and his crew of Information Concepts,
Inc. researchers continued the process of purging phantoms from the records.
Since then, further investigation by Neft, Pete Palmer, Bill Haber, Al
Kermisch, and others has led to the expunging of several more players.
     In many instances confirming the status of a phantom was a complicated
chore. The task sometimes was made more difficult because the player's name
was included in the official league statistics. In other cases, especially
those involving nineteenth-century performers, the fact that official league
records no longer exist compounded the problem because it made it impossible
to determine who was credited by the official scorer and/or league
statistician with appearing in the game or games in question. (The American
League's official game-by-game player sheets of 1901-1904 reportedly were
destroyed by fire decades ago, while the official National League data also
vanished for the pre-1902 period except for the 1899 season.)
     An example of a phantom who appears in the league records is Albert W.
Olsen. For thirty-five years he was carried in the Encyclopedias and shown as
participating in one game--as a pinch hitter--for the Boston Red Sox in 1943.
Olsen did train with the Red Sox that spring, but he was shipped to San Diego
of the Pacific Coast League before opening day and spent the entire season in
the minors. In addition, it was his exploits as a lefthanded pitcher, not as
a hitter, that originally attracted the Red Sox.
     Nevertheless, American League records list Olsen as playing for Boston in
a game in Chicago on May 16, 1943. Newspaper boxscores credited him with
drawing a walk while batting for pitcher Dick Newsome and subsequently
stealing a base. However, research has confirmed that the Red Sox
pinch-hitter on that occasion definitely was not Olsen. Instead, it may have
been outfielder Leon Culberson, who had just been called up from Louisville of
the American Association, or possibly another Boston rookie outfielder, Johnny
Lazor.
     And why is there still uncertainty as to who the pinch-hitter really was?
How could this mistake have occurred? First off, the exchange of roster data
among major league teams during that period was quite limited. Consequently
scorecards often contained the same player names and uniform numbers for the
visiting team, especially early in the season, as the team listed in spring
training. For example, the scorecard for a Red Sox-Senators game in
Washington approximately a week before the incident in question showed Olsen
on Boston's roster with uniform number 14, even though he was pitching for San
Diego. Second, it's unlikely there were phone lines from the pressbox to the
dugouts in those days, and in the absence of being able to call down to check
on the identity of a player, the media and official scorers relied on outdated
roster information.
     Several years ago researchers felt they had cleared up the Olsen mystery
upon discovering that one Boston newspaper identified the pinch-hitter as
"Culbeson." This seemed to settle the matter. After all, Culberson had
joined the team that day and was installed as the Red Sox' leadoff batter in
the second half of the afternoon doubleheader, collecting a single and triple
in five at-bats. But the mystery was revived when Ed Walton, a Red Sox
historian, discussed the situation with Culberson at a Red Sox Old Timers
affair in 1986. According to Walton, Culberson contended he did not pinch-hit
in the twin-bill opener, but rather made his debut in the second game. He
added that manager Joe Cronin said he wanted the young newcomer to sit beside
him (Cronin) through the first game to get a feel of the big leagues.
However, the records reveal Cronin played third base during the entire first
game. All of which raises the questions: Was Culberson's memory playing
tricks on him and was he indeed the pinch-hitter listed as Olsen? Or was it
Lazor, who is shown as wearing uniform number 14 later in the year but who was
batting a mere .136 (3 for 22) at the time? Because the passage of time makes
memories hazy and also because the incident was of no particular significance,
the mystery may never be solved.
     A manager's pique led to a phantom known as J.A. Costello getting into
the early Encyclopedias. Curiously, the incident also took place in Chicago
and likewise involved a player's big league debut. It occurred in the morning
half of a holiday bill between Cleveland and the White Sox on July 4, 1912.
Late in the contest, Indians' Manager Harry Davis, still burning over an
umpire's decision, sent the newcomer into the game to replace center fielder
Joe Birmingham and instructed him to announce himself to umpire Gene Hart as
"Costello."
     In reality, the player was Kenneth Nash, who had only recently joined
Cleveland from Boston University law school. He subsequently appeared in ten
additional games with Cleveland that season under his correct surname, mostly
as a shortstop, and then played with the St. Louis Cardinals for part of
1914. Nash later became a prominent state representative, state senator, and
judge in his home state of Massachusetts.
     The 1903 official National League records contain a phantom who long
baffled researchers. Among the Pittsburgh player sheets is one headed "George
Gray" with entries for two games as an outfielder--on May 28 and May 31. Four
years earlier the Pirates had a pitcher named George "Chummy" Gray, and it's
possible the scorer or league statistician remembered him while filling in the
player's first name. When the first Official Encyclopedia appeared in 1951,
the player was identified as William (rather than George) Gray, a native of
Pittsburgh.
     It turns out that the two George (or William) Gray entries properly
belonged to not one, but two different players. The Pirates' left fielder in
the May 28 game really was Romer Carl "Reddy" Grey, younger brother of
novelist Zane Grey. He had been obtained on loan from the Worcester club of
the Eastern League to fill in that afternoon in the final game of a series in
Boston.
     On the other hand, while boxscores in Cincinnati newspapers listed the
Pittsburgh left fielder for the May 31 game in Cincinnati as "Gray," the
player actually was Ernest Diehl, a Cincinnati sandlotter who had been
recruited by the injury-riddled Pirates. The game represented Diehl's first
appearance in the major leagues and his only game that season, but he played
12 more games with Pittsburgh the following year. It should be noted that
ever since the original Turkin/Thompson tome, the Encyclopedias have credited
Diehl with playing one game with Pittsburgh in 1903, but until recent years
they also carried William Gray with two games.
     A majority of baseball's phantoms were the product of typographical
errors--instances where a linotypist or a typesetter mistakenly included one
or more incorrect letters in a name, or where a printer inserted a correction
line in the wrong place. In compiling data for the Encyclopedias, the
authors/researchers relied not only on the so-called official league records
but also combed boxscores published in The Sporting News, Sporting Life, The
New York Clipper, and various local newspapers. In the process they
occasionally came upon what appeared to be a previously unlisted "new" player
who, in the final analysis, proved to be someone else. Even today,
misspellings of this type in newspapers can cause great befuddlement.
     A classic illustration of a phantom who was created by a typographical
error is the player carried in the early Encyclopedias as John P. Morgan. He
was listed as appearing in one game as a third baseman with the Philadelphia
Athletics in 1916. A study of A's boxscores for the season disclosed the
source of the mix-up: It was The Sporting News' boxscore of the August 3
game at Cleveland. (Sporting Life had dropped boxscores by this time.)  The
Philadelphia half of TSN's box has "Morgan, 2b," while on the same line in
Cleveland's half is "Gandil, lb."
     Careful examination reveals the Morgan/Gandil type slug was a correction
line that the printer inserted in the wrong place. It was intended for the
Washington-at-Cleveland boxscore of the previous day (August 2) in which the
Senators' second baseman, Harry Morgan, appeared on the same line as Chick
Gandil and contained the identical AB-H-PO-A-E figures given for each player
in the misaligned August 2 line.
     This explanation may well leave you, the reader, with several questions,
to wit: (1) Who did play third for the A's that day, and was he properly
credited in the official records? (2) Why did the Encyclopedia compilers show
Morgan as a third baseman instead of a second baseman? (3) How did they come
up with "John P." for the impostor's name?
     The answers are: (1) Lee McElwee and, yes, he was credited with this
game; (2) because veteran Nap Lajoie played the entire game at his usual
second base position for the A's, the researcher who made this "discovery"
apparently assumed it should have read "Morgan, 3b" rather than "2b"; (3) an
infielder named John P. Morgan was active in the minor leagues that season,
and the Encyclopedia editors probably figured the A's had given him a trial.
Numerous other phantoms were similarly tagged with the first names of
then-current minor leaguers.
     Typos involving misspelled names led to a number of one-game phantoms.
Two examples will serve to demonstrate the point. They are John H. Carlock
(1912, Cleveland) and a player listed simply as Deniens (no first name given)
with the 1914 Chicago Federals. Sporting Life had "Carlock, ph" for Cleveland
in an August 24, 1912, game at Boston, but Cleveland newspapers and The
Sporting News reported the pinch-hitter was Fred Carisch, a reserve catcher
with the Indians. Official AL records also credit Carisch with that
appearance. "Deniens, c" turned out to be Clem Clemens, a catcher in 13 games
with the 1914 ChiFeds. One can readily visualize how handwritten names
"Carisch" and "Clemens" could have been misinterpreted by a telegrapher or
linotypist for "Carlock" and "Deniens."
     The origins of some other phantoms are more mysterious. Take the case of
Lou Proctor. The Encyclopedias credited him with one appearance with the 1912
St. Louis Browns. The Sporting News and Sporting Life, which may have
obtained their boxscores from the same source, show "Proctor, ph" and
"Procter, pi," respectively, in a May 13 game at Boston. However, a Boston
newspaper referred to the pinch hitter as Albert "Pete" Compton, a catcher
with the Browns that season (whose real first name was, of all things, Anna).
While AL records contain no reference to Proctor, they unfortunately also fail
to include the May 13 appearance among Compton's 100 games. Rumor has it that
Proctor was a prankish Western Union telegrapher who inserted his own name as
a pinch hitter.
     The presence on one team of two players with the same or an almost
similar surname can lead to problems. The Washington Senators figured in
three such mixups. Ironically, in one instance research by Kermisch has
established that a player long labeled a phantom was, in fact, the real McCoy.
The player in question was Charles C. Conway. In 1911 the Senators had both
Conway, a rookie outfielder up from Youngstown (Ohio-Pennsylvania), and
William "Wid" Conroy, veteran infielder-outfielder, on their spring roster.
When the season began, Conroy was idled by a stone bruise of the foot.
Meantime, Conway appeared in two games during opening week, finishing up in
right field on April 15 and starting at that position three days later, before
the Senators returned him to Youngstown. Unfortunately, Sporting Life and The
Sporting News each showed "Conroy" as the replacement in the first contest,
but both had "Conway, rf" in their April 18 boxscore. Although the early
Encyclopedias listed Conway and credited him with playing two games, the
Macmillan compilers dropped his name 20 years ago on the erroneous premise
that it was the veteran Conroy who participated in the April 15-18 games.
Actually, Conroy didn't make the first of his 106 appearances that year until
April 27.
     An equally confusing puzzle centered on a 1914 Washington pitcher known
as Barron or Barton. The early Encyclopedias carried both John J. Barron
(later changed to Frank John Barron) and Carroll R. "Buck" Barton and
credited each with pitching in one game for the '14 Senators. Actually only
one pitcher was involved (and just one appearance), but which pitcher was it?
Player contract data of the period reveal that Washington signed Carroll R.
Barton in 1913 and retained rights to him while he pitched for Newport News
(Virginia League) that season and again in 1914. During the same two years
John J. Barron was pitching in the New England League. To complicate matters
further, Washington signed Frank J. Barron in 1914 and shipped him to Newport
News, where he became a teammate of Barton. While Barron posted a dismal 1-3
record, Barton was a 16-game winner that year.
     So who was the 1914 Washington pitcher? The American League player
records contain an entry headed "Barron" with data for an August 18 game, and
boxscores of the contest likewise have "Barron" pitching one inning--the
ninth--for the Senators against the visiting St. Louis Browns. The player's
correct identity was confirmed when, shortly before his death in 1964, Frank
Barron disclosed in an interview that while still studying for his law degree
at West Virginia University he was signed by Clark Griffith in 1914, was
assigned to Newport News and later pitched one inning for Washington before
resuming his law studies.
     The third Washington mix-up relates to a 1944 player identified in the
Encyclopedias as Armando Viera Valdes. Official AL records contain a sheet
for Armand (without the final "o") Valdes and note that he made a pinch-
hitting appearance for the Senators in a May 3 game at Boston, but Richard
Topp's research has disclosed that the pinch hitter was Rogelio "Roy" Valdes,
a fellow Cuban but no relative of Armando.
     With many major leaguers lost to military service in 1944, scout Joe
Cambria lined up several Cubans to fill voids on the Senators' roster. Early
in the year he signed an outfielder who was listed in the 1944 American League
Red Book as Armand (without the "o") Valdes. Several weeks thereafter--too
late for inclusion in the Red Book--Cambria signed Rogelio Valdes, a catcher.
Both spent the early weeks of the season with Washington (each was later
optioned to Williamsport of the Eastern League), and presumably the official
scorer and/or league statistician picked up the incorrect first name from the
Senator roster in the AL Red Book.
     As in the case of Charles Conway, another player once regarded as a
phantom has turned out to be a legitimate athlete after all. The
Turkin-Thompson tomes listed him as William Krouse, a second baseman in one
game with Cincinnati in 1901. Compilers of the Macmillan encyclopedia decided
he was an impostor and dropped him, crediting his appearance to Bill Fox, the
Reds' regular keystoner. However, research has revealed that the "Krouse, 2b"
for Cincinnati in the July 27, 1901 game at Chicago was a recently-released
minor leaguer whose correct name was Charles "Famous" Krause. Krause, who was
on his way home to Detroit at the time, was given a chance with the Reds
because Fox was sidelined with a split finger, but the newcomer performed so
poorly that he was dumped after that one appearance.
     Recent research has revealed that one player long listed as a
phantom--Ivan Bigler, who was shown in one boxscore at first base with the
1917 St. Louis Browns when George Sisler actually played there that
day--really did make an appearance with the Browns as a pinch runner on May 6
that season, and thus he's been restored to the all-time list of major
leaguers.
     The accompanying table lists the phantoms who have been eliminated from
the all-time roster of major league players since the first Turkin/Thompson
Official Encyclopedia of 1951. In the absence of an official clearinghouse
for such data, no claim is made that the list is complete. Where it is
available, brief information on the reason for the deletion of the player is
given. Unfortunately, documentation by Turkin/Thompson and the ICI group that
compiled the 1969 Encyclopedia disappeared years ago.
     With today's sophisticated technology and record-keeping procedures, the
margin for error in the identification of players--and in the official
statistics--has been reduced considerably. Still, a slipup that occurred in
the 1984 official American League averages (Alvaro Espinoza was omitted
completely, even though he appeared in one game with Minnesota) emphasizes
that mistakes still are possible.
     A factor that poses the potential for error is the increasing frequency
in recent years of teams having two or more players with the same or nearly
identical surname--and often performing at the same position. In 1992 and
1993, for instance, the Los Angeles pitching staff included brothers Ramon and
Pedro Martinez as well as Kip and Kevin Gross, who are not related. Among
other teams with pitching pairs of the same name in 1993 were: Houston--Matt
and Curt Jones; Philadelphia--Mitch and Mike Williams; Cleveland--Matt and
Curt Young; and San Diego--Gene and Greg W. Harris (not to be confused with
Boston pitcher Greg A. Harris). The season also witnessed numerous other teams
having players with identical last names who played similar positions,
including outfielders Rickey and Dave Henderson, Oakland; outfielders Bernie
Williams, New York Yankees, and infielders Edgar and Tino Martinez, Seattle.
     Then there's the case of twin brothers Jose and Ozzie Canseco, both
outfielders, playing simultaneously for a time with Oakland in 1990, and such
earlier potentially baffling situations as Atlanta's pitching staff of 1987
that included two Smiths (Zane and Pete) plus a Cary (Chuck) and a Clary
(Marty), while St. Louis had a duo of relief specialists that year named
Dayley (Ken) and Dawley (Bill) and also a Perry (Pat) and a Terry (Scott), two
pitchers who coincidentally were traded for each other during the season. But
the twosome that could cause even more difficulty for researchers circa 2000
is Don Robinson, both of whom pitched for San Francisco as well as Pittsburgh
in 1987.
     Couple this with the typographical errors that still show up in newspaper
boxscores, and you can see that the days of garbage in, garbage out are likely
to continue.


Tabulation of Phantoms

Allen, Robert. 1919 Philadelphia AL, 9 games as OF. Pseudonym used by Alvah C.
"Rowdy" Elliott, long-time minor leaguer.

Baldwin, ----. 1907 Boston NL, 1 game as C. One of James C. Ball's 10 games.

Barton, Carroll R. 1914 Washington AL, 1 game as P. Same as game credited to
Frank J. Barron.

Boylan, ----. 1887 Philadelphia NL, 1 game as 2B. One of Charles J.
Bastian's 60 games.

Carlock, John H. 1912 Cleveland AL, 1 game as PH. Typographical error; one
of Frederick B. Carisch's 26 games.

Christman, H.B. 1888 Kansas City AA, 1 game as C. Documentation behind
deletion no longer available.

Collins, Frank. 1892 St. Louis NL, 1 game as OF. Documentation behind
deletion no longer available.

Costello, J.A. 1912 Cleveland AL, 1 game as OF. One of 11 games played by
Kenneth L. Nash, who used pseudonym in debut.

Davis, Thomas J. 1890 Cleveland NL, 1 game as OF. One of George Stacey
Davis' 136 games.

Davis, ----. 1903 Chicago NL, 1 game as OF. Documentation behind deletion no
longer available.

Deniens, ----. 1914 Chicago FL, 1 game as C. Typographical error; one of
Clement L. Clemens' 13 games.

Drennan, K. John. 1904 Detroit AL, 1 game as 1B. Typographical error; one of
William "Wild Bill" Donovan's 46 games.

Dresser, Edward. 1898 Brooklyn NL, 1 game as SS. Typographical error; one of
Jack Dunn's 51 games.

Dugan, E. 1884 Kansas City UA, 3 games as OF. Same player as William H.
Dugan, who played 9 games with Richmond AA the same year.

Gray, William. 1903 Pittsburgh NL, 2 games as OF. One game belongs to Romer
C. Grey; other game belongs to Ernest G. Diehl.

Kerns, Daniel P. 1920 Philadelphia AL, 1 game as PH. Pseudonym used by Edward
"Ted" Kearns, later a 1B with 1924-25 Chicago NL.

King, Frederick. 1901 Milwaukee AL, 1 game as C. Game belongs to John A.
Butler, later with St. Louis and Brooklyn NL, who used pseudonym in debut.

Lane, ----. 1901 Boston NL, 1 game as 3B. Typographical error; one of Bobby
Lowe's 129 games.

Mares, ----. 1894 Louisville NL, 1 games as OF. Documentation behind
deletion no longer available.

McCauley, William. 1884 St. Louis AA, 1 game as C. Game belongs to James A.
McCauley, who also played in 1885-1886.

Meddlebrook, ----. 1884 Baltimore UA, 1 game as OF. Documentation behind
deletion no longer available.

Merson, ----. 1914 Brooklyn FL, 1 game as PH. Typographical error; one of
George J. Anderson's 98 games.

Miller, Bert. 1897 Philadelphia NL, 3 games as 2B. Games belong to Frank A.
Miller, formerly listed as Frederick Miller, who also played one game each
with Washington NL and St. Louis NL in 1892.

Miller, Henry D. 1892 Chicago NL, 4 games as P. Games belong to Harry
DeMiller, who was erroneously listed for 1 game as 3B with 1892 St. Louis
NL.

Moore, Guy W. 1922 St. Louis NL, 1 game as OF. Documentation behind deletion
no longer available.

Morgan, John P. 1916 Philadelphia AL, 1 game as 3B. Typographical error; one
of Leland S. McElwee's 54 games.

Olsen, Albert W. 1943 Boston AL, 1 game as PH. Appearance apparently belongs
to either Leon Culberson or Johnny Lazor.

Pratt, Thomas J. 1884 Baltimore AA, 1 game as OF. Documentation behind
deletion no longer available.

Proctor, Lou. 1912 St. Louis AL, 1 game as PH. Game belongs to Anna S.
"Pete" Compton, who played in 101 games that season.

Ritchie, ----. 1910 St. Louis NL, 1 game as PH. Documentation behind
deletion no longer available.

Schauer, ----. 1890 Columbus AA, 1 game as 1B. Documentation behind deletion
no longer available.

Seymour, Thomas. 1882 Pittsburgh AA, 1 game as P. Player's correct name was
Jacob Semer.

Sheehan, Timothy. 1884 Washington UA, 1 game as OF. One of John A. Ryan's
seven games.

Smith, Charles H. "Pacer." 1877 Chicago/Cincinnati NL, 34 games as 2B-OF-C.
Record belongs to Harry W. Smith, who also played 1 game with 1889 Louisville
AA.

Smith, E.J. 1890 Buffalo PL, 1 game as 1B. One of John Irwin's 77 games.

Strands, Lewis. 1915 Chicago FL, 1 game as 2B. One of John J. Farrell's 70
games.

Thayer, Edward L. 1876 New York NL, 1 game as 2B. Player's correct name was
George T. Fair.

Turbot, ----. 1902 St. Louis NL, 1 game as OF. Documentation behind
deletion no longer available.

Valdes, Armando V. 1944 Washington AL, 1 game as PH. Game belongs to a
different player, Rogelio "Roy" Valdes.

Young, David. 1895 St. Louis NL, 1 game as 3B. Documentation behind deletion
no longer available.