Return to Top

Carved in Stone?
Revisiting the 1949 AL Batting Race

In 1949, George Kell won the American League batting race with a batting average of .34291, edging out Ted Williams (.34276) by a very, very narrow margin—.00015. As a result, Williams just missed winning the Triple Crown—leading the league in batting average, home runs, and RBIs.

By Bill Nowlin and Herm Krabbenhoft, March 14, 2017
Baseball Banter: Join a Chat with the author. More Details > >
George Kell holds the sterling silver bat and citation received for winning the American League batting title in 1949.

In 1949, George Kell won the American League batting race with a batting average of .34291, edging out Ted Williams (.34276) by a very, very narrow margin—.00015. While Kell of the Detroit Tigers did not have any other first-place finishes in the various batting categories, Williams of the Boston Red Sox led the league in homers (43) and runs batted in (159, tied with teammate Vern Stephens). Thus, Teddy Ballgame just missed winning the Triple Crown—leading the league in batting average, home runs, and RBIs. Williams, who played in each of Boston’s 155 games, also led the league in slugging percentage (.650), runs scored (150), doubles (39), walks (162), and total bases (368).[1] He also reached base safely nearly 50 percent of the time (his on-base percentage was a league-leading .490—a distant second was Luke Appling’s .439).

Williams had won the Triple Crown twice, first in 1942 and again in 1947. He remains the only AL player to win two Triple Crowns. The National League also has only one player to win two: Rogers Hornsby (1922 and 1925). Since 1901, only 13 players have ever won a Triple Crown.[2]

Two Personal Reasons for Undertaking This Research

Bill writes: Ted Williams was my hero growing up. I had the opportunity to meet him and visit his home on a few occasions late in his life. For a number of years, I was the volunteer editor of the annual Ted Williams Magazine, published by the Ted Williams Museum. I’ve written, with Jim Prime or myself, several books on Williams.

What if someone had erred in calculating the batting averages back in 1949? The smallest of errors could result in Ted Williams being awarded an unprecedented third Triple Crown. I could be a hero—at least in my own mind! As it happens, I had written his last published words (for the aforementioned magazine early in 2002). What if I proved that the 1949 batting title was rightfully his?

It really would take very little—just one minor bookkeeping error. The official records show Williams with 194 hits in 566 at-bats. Kell had 179 hits in 522 at-bats. What if Kell had just one more at-bat that hadn’t been properly recorded in the official records? Then he’d be 179 in 523, or .3422562. Ted Williams would be the 1949 American League batting champion—and Triple Crown winner!

What if Williams had 565 at-bats, not 566? He’d be .3433628. Again, he would win the batting title—and earn the Triple Crown!

Any number of things could have been off. It wouldn’t be the first time.[3] In fact, over the past 50 years, the Elias Sports Bureau (the official statistician for the National League since 1923 and for Major League Baseball since 1987) has changed the batting average of nine batting champions—Elmer Flick (1905), Ty Cobb (1908, 1909, and 1911), Honus Wagner (1911), George Sisler (1922), Harry Heilmann (1923 and 1925), and Johnny Mize (1939). Fifty years and two leagues—100 batting titles—and nine of their averages were changed? That’s a pretty significant number. All I was looking for was a minuscule .00015 change.

Take the most recent adjustment—the Johnny Mize change for 1939. In 2009—70 years after the fact—the Elias Sports Bureau discovered a one at-bat error in Mize’s originally generated official day-by-day (DBD) record for the game played between the Cardinals and the Dodgers on August 23, 1939, in Brooklyn, which changed his NL-leading batting average from .349 to .350.[4]

Johnny Mize—MVP of the National League in 1939.

There have been changes in other statistics over the course of time, among the most famous being Hack Wilson’s 1930 runs batted in number.[5]

I didn’t have anything against George Kell. He seemed like a nice-enough guy the couple of times I met him at the Ted Williams Museum. This wasn’t about him. It was about trying to get Ted another Triple Crown. Fair and square. Through research. On my first run through, I discovered some tabulation errors in the official Day-By-Days for Ted Williams (reported below). Could this really be? Was I really starting to find something? I was a little scared by the prospect of truly finding something that would hold up. I wanted to be sure to do things right, and not to take any chances at all that I’d be crowing with success—only to find that someone spotted an error in my own calculations and thus I would have to eat crow.

After doing my own initial look at matters, I contacted Herm Krabbenhoft because I knew of his research, particularly that described in note 3(a). I then found that he harbored the same curiosity about the Kell-Williams batting race and that he had also done some preliminary research on it. So, upon learning of our mutual interest in this effort, we decided to combine our research.

Herm writes: I first became aware of the 1949 batting title race around 1958 when I was 13. I bought a baseball magazine, Baseball’s All-Stars—The Greatest Games of Today’s Top Players put out by Sport magazine; I still have the magazine. One of the articles, “Batting Champ by Inches” recounted how Kell won the 1949 AL title on the last day of the season. At the time, I was probably happy for Kell, since he was a Tigers player. However, I wonder if something stated in the article bothered me then as much as it bothers me now: with Kell scheduled to bat fourth in the last inning of the final game of the season, Johnny Lipon made the first out. Dick Wakefield, batting for relief pitcher Hal White, singled. Eddie Lake was coming up next, with Kell to follow. “Yet even as Lake was digging in at the plate, [Manager] Red Rolfe and Kell, conferring in the Detroit dugout, decided to let George sit it out the rest of the way. Joe Ginsberg, a rookie catcher up from Toledo, was told to go out on deck. He would hit for Kell and that would be that. But Ginsberg never got to hit, as indeed, Kell never would have. Lake hit a hot smash to [Cleveland Indians] shortstop Ray Boone who turned it into a speedy season-ending double play, and thereby put a lock on Kell’s batting title.”

As a kid, I also liked that Williams won the batting title in 1957 and 1958. I had hoped that he would win it in 1956 and that Al Kaline would win the RBI crown. I liked that in 1941 he didn’t sit out the final two games of the season with a rounded-up .400 batting average (.39955) and ended up with a bona fide .406 instead. I couldn’t understand why Williams didn’t win the batting crown in 1954 (because my baseball cards showed he had a .345 BA, while the champion Bobby Avila had a .341 average). And I felt that he would also have won it in 1955 if he had played the whole season—because a 1956 baseball pamphlet I picked up in Kansas while on our family vacation listed each team’s leader in a variety of statistics. While Kaline, my favorite player, won the batting title with a .340 mark, Williams batted .356. So, I had developed a liking for Williams because he was really a fantastic hitter, but he seemed to have been the victim of circumstances. That feeling was furthered as I learned of his losing five prime seasons to both World War II and the Korean War. And then decades later, while doing the research described in note 3(a), I determined that I had to revisit the Kell-Williams batting race.

Ted Williams served in the Navy as a pilot and later in the Marine Corps.  Although he did not see combat in WWII, he did fly 39 combat missions during the Korean War.
Courtesy of: Frank Ceresi Collection

Research Procedure

Each of us decided to utilize the game accounts provided in the various newspapers to try to ascertain what Kell and Williams did in each of their plate appearances. Since Bill lives in Boston and Herm is frequently in the Detroit area, we had ready access to the principal daily newspapers in Boston (Globe, Herald, and Post) and Detroit (Free Press, News, and Times). Newspapers from the cities of the other AL teams were also examined. As it developed, there were just too many plate appearances for which we could not ascertain precisely what Kell and Williams did. Fortunately, while we were working on this project in a now-and-then from time-to-time manner, Retrosheet was diligently working on getting play-by-play (PBP) accounts for every game for every team from the 1949 season, as well as many other seasons. So, in 2016, not long after Retrosheet updated its website to include the “complete” PBPs for the 1949 season, we returned to our quest and took a close look at the Retrosheet PBP information for the 1949 Tigers and Red Sox. While Retrosheet does provide PBP information for each plate appearance for Kell and for Williams, some of the plate-appearance results are deduced. For example, for Kell in the game on April 27 between the Tigers and the Browns at Briggs Stadium in Detroit, Retrosheet has the following for his plate appearances (shown here in boldface):

Tigers 1st: Berry grounded out (second to first); Lipon grounded out (third to first); Kell tripled to center; Kell stole home; Wertz flied out to center; 1 R, 1 H, 0 E, 0 LOB.

Tigers 3rd: Newhouser out on an unknown play; Berry out on an unknown play; Lipon singled; Kell singled [Lipon to second]; Wertz singled [Lipon scored, Kell to second]; Evers struck out; 1 R, 3 H, 0 E, 2 LOB. Browns 3, Tigers 2.

Tigers 5th: Berry out on an unknown play; Lipon out on an unknown play; Kell out on an unknown play; 0 R, 0 H, 0 E, 0 LOB. Browns 3, Tigers 3.

Tigers 8th: Kell out on an unknown play; Wertz out on an unknown play; Evers out on an unknown play; 0 R, 0 H, 0 E, 0 LOB. Browns 3, Tigers 3.

Tigers 11th: Kell out on an unknown play; Wertz out on an unknown play; Evers out on an unknown play; 0 R, 0 H, 0 E, 0 LOB. Browns 3, Tigers 3.

Thus, Retrosheet deduced—from its evaluation of the information given in the box scores presented in the various newspaper accounts and in the originally generated official DBD records—that Kell was “out on an unknown play” in each of the fifth, eighth, and 11th innings. Our own examination of the information provided in the Detroit Free Press, News, and Times and the St. Louis Globe-Democrat and Post-Dispatch, not surprisingly, arrived at the same conclusions—Kell was retired in each of his plate appearances in the fifth, eighth, and 11th frames.

Kell played ball for 15 years. He was a 10-time All-Star and was inducted into the Baseball Hall of Fame in 1983.

As it turns out, there are 71 “Kell out on an unknown play” deductions as well as 27 other plate appearances where Kell was retired, but some information is lacking, such as in the game on August 25 between Detroit and Washington in the Motor City—for each of Kell’s plate appearances (in the second, fourth, sixth, and eighth innings) the Retrosheet PBP is “Kell flied out on an unknown play.”

For Williams, the Retrosheet PBPs include 11 “Williams out on an unknown play” deductions and 23 plate appearances where Williams was retired, but some information was lacking, such as in the game on June 20 between the Red Sox and Tigers in Boston—for his third-inning plate appearance the Retrosheet PBP states, “Williams made an out to shortstop” and for his eighth-inning plate appearance the Retrosheet PBP states, “Williams made an out to third.”

Results

Comparison of Kell’s official DBD records with his Retrosheet Player Daily file reveals that there are no discrepancies for the at-bats Kell had; likewise for the hits Kell collected. The only significant discrepancy for Kell was for the first game of the double header on August 19 (Tigers vs. Browns in St. Louis)—the official DBD records for Kell show him with four (4) at-bats, one (1) hit, two (2) total bases, one (1) double, and one (1) grounded into double play; the Retrosheet PBP states he grounded into a 4–6–3 double play in the first inning, was out on an unknown play in the third inning, was safe on an error by the shortstop in the sixth inning, hit a double in the seventh inning, and was walked intentionally in the ninth inning. The description in the Detroit Free Press states, “. . . after Kell was intentionally passed, Wertz popped up and Evers struck out.” Similarly, the Detroit News account gives, “An intentional pass to Kell refilled the bases and brought Wertz to the plate.” The Retrosheet Discrepancy File shows that Eddie Lake was erroneously credited with a walk in the official DBD records. Thus, in summary, there appear to be no errors in Kell’s originally generated official DBD records impacting his batting average—he did, in fact, collect 179 hits in 522 at-bats yielding a .3429 batting average.

Comparison of Williams’s official DBD records with his Retrosheet Player Daily file reveals that there are no discrepancies for the at-bats (ABs) Williams had; likewise for the hits Williams collected. There were, however, some “mathematics” errors, which were later corrected during the season. Thus, the at-bat entry for the game on July 6 is two (2) with a to-date total of 275. Then, for the game on July 7, the AB entry is three (3) with a to-date total of 277—not 278. The “missing” at-bat was carried through the game on August 17, which showed an AB entry of five (5) and a to-date total of 422—not the correct value of 423. Then, for the game on August 19, the AB entry is three (3) and a to-date total of 426—the correct value. Similarly, for the game on August 9, the hits entry is two (2) and the to-date value is 135. Then, for the game on August 10, the hit entry is two (2) with a to-date total of 136—not the correct value of 137. The “missing” hit was carried through the game on September 20, which showed a hit entry of zero (blank cell) with a to-date total of 187. Then, for the game on September 21, the hit entry is one (1) and the to-date value of 189—the correct value. There was also another AB math error [three (3) extra ABs] in the to-date entry for the game on August 27; this plus-three AB mistake was carried through the game on September 13 and then corrected with the to-date entry for the game on September 14. Thus, in summary, there appear to be no errors in Williams’s originally generated official DBD records impacting his batting average—he, in fact, did collect 194 hits in 560 at-bats yielding a .3427 batting average.

CONCLUSION

With heavy reliance on the Retrosheet PBP information, our re-examination of the 1949 batting race in the American League reveals that the originally generated official DBD records appear to ultimately be correct—Kell achieved a batting average of .3429 and Williams accomplished a batting average of .3427. Thus, Kell did indeed win the 1949 AL batting crown. History has not been overturned. Williams did not earn a third Triple Crown.[6]

Though Ted Williams did not earn a third Triple Crown, he is still considered one of the best hitters in baseball history.
Original artwork by
Peter Chen

 

Thanks to all of the Retrosheet Volunteers for all the great work they do to generate PBP information, of which we relied significantly for this article. Retrosheet is truly a baseball research enabler!

 

Notes


[1] Williams also led the league in a number of modern metrics not employed in 1949, such as WAR, the closest to his 9.3 WAR being Eddie Joost’s 6.3 WAR.
[2] Herm Krabbenhoft, (a) “Accurate RBI Records for Players of the Deadball Era: Part 4—The Players on the 1912 Braves, Cubs, Giants, and Pirates (The Definitive Resolution of the Discrepancy for the 1912 National League Triple Crown),” The Inside Game (February 2015): 5; (b) “Seeking Resolution of the Discrepancy for the 1912 NL Triple Crown,” Baseball Research Journal (Spring 2015): 54.
[3] See, for instance, Herm Krabbenhoft, (a) “Ted Williams’ On-Base Performance in Consecutive Games,” Baseball Research Journal (2004): 41; (b) “The Authorized Correction of Errors in Runs Scored in the Official Records (1945-2007) for Detroit Tigers Players,” Baseball Research Journal (2008): 115; (c) “The Authorized Correction of Errors in Runs Scored in the Official Records (1920-1944) for Detroit Tigers Players,” Baseball Research Journal (Spring 2011): 66; (d) “Additional Corrections in the Official Records (1920-1944) of Runs Scored for Detroit Tigers Players,” Baseball Research Journal (Fall 2013): 99; (e) “Missing . . . Found . . . Phantom: The Accurate Runs-Scored Record for the 1906 Detroit Tigers,” The Inside Game (April 2014): 3; (f) “Accurate Runs-Scored Statistics for the Players on the 1912 Chicago Cubs,” The Inside Game (December 2014): 1; (g) “Accurate Runs-Scored Records for Players of the Deadball Era: The Players on the 1916 Detroit Tigers,” The Inside Game (April 2016): 1; (h) “Lou Gehrig’s RBI Record,” Baseball Research Journal (Fall 2011): 12; (i) “Lou Gehrig’s RBI Record: 1923-39,” Baseball Research Journal (Fall 2012): 10; (j) “The Accurate RBI Record of Babe Ruth,” Baseball Research Journal (Spring 2013): 37; (k) “Hank Greenberg’s American League RBI Record,” Baseball Research Journal (Spring 2012).
[4] The officially generated DBD record shows Mize having played in that August 23, 1939, game, appearing as a pinch hitter and going hitless in one at-bat. However, as reported in the box scores in the various newspaper accounts that covered the game, Mize did not play in that game. But, Lynn Myers did play in that game, batting for Tom Sunkel in the ninth inning. Myers, who was retired in his only at-bat, is not shown as playing in the game in the official DBD records. It seems, therefore, that someone (perhaps the game’s official scorer, who filled out the game’s official scoresheet, or a staffer at Elias, who transcribed the official scoresheet information to the official DBD sheets) entered one (1) at-bat for Mize instead of Myers. The consequence of correcting the error is that, in actuality, Mize, for the entire season, had 197 hits in 563—not 564—at-bats, which yields his true batting average of .34991 (i.e., .350)—not .34929 (i.e., .349). Accordingly, Seymour Siwoff, the publisher of The Elias Book of Baseball Records, changed Mize’s batting average from .349 to .350 in the 2010 edition.
[5] Cliff Kachline, “Hack Wilson’s 191st RBI,” The Baseball Research Journal (2001): 76.
[6] There is an interesting “what-if” curiosity associated with the 1949 Kell-Williams batting race—“What if the official scoring rules that currently exist had been in effect in 1949? Since 1954, an RBI fly out has been considered a sacrifice fly, with the batter not charged with an at-bat. However, from 1940 through 1953, an RBI fly out was just like an RBI groundout (i.e., the batter was charged with an at-bat). Scrutinizing the Retrosheet PBP information reveals that in 1949, Williams had seven RBI fly outs while Kell had six RBI fly outs. So, if the current sacrifice fly rule had been operative in 1949, Williams would have had seven fewer at-bats (559 instead of 566), which would have given him a batting average of .34705. Kell, however, would have had only six fewer at-bats (516 instead of 522), which would have given him a batting average of .34690. Thus, Williams would have been the batting champion and won the Triple Crown—if the current sacrifice fly rule would have been in effect in 1949. But, it was not. [Herm Krabbenhoft, “The Sacrifice Fly Rule Is . . . Just a Little Bit Pregnant (The Impact of the Sac-Fly Rule on Royalty . . . The Kings of the Percentage Crowns for Batters),” presentation given at the SABR Baseball Records Committee Meeting, SABR-45, Chicago (June 26, 2015).] 

 

 

 

All the images within this article can been seen by clicking on the lead image and moving/hovering your mouse over on the center right. A pointing hand will appear, you can click on it to scroll through the images.
Please feel free to contact us if you additional information or questions about our artifacts and columns. www.thenationalpastimemuseum.com/contact or info@tnpmuseum.com.