So how did the 1938 AL pitch Hank Greenberg?

Was the 1938 American League so poisoned with anti-Semitism that its pitchers prevented Hank Greenberg from breaking Babe Ruth’s single-season home run record by pitching around him? This has been a long-standing belief for some, and Howard Megdal’s Religion Aided a Home Run Chase, and May Have Led to Its Failure article (New York Times, 3/19/2010) claims that Retrosheet data proves this was the case.

As I think you will soon agree, the data actually indicates the opposite. But before presenting the data, let’s flesh out the historical context, as the racial and ethnic prejudices of 1938 America were mainstream to a degree that is inconceivable in 2010.

Prejudice against African-Americans was of course the most overt example. The fact that Josh Gibson and Satchel Paige were not allowed to play with or against Hank Greenberg speaks for itself. Another infamous example occurred when the Daughter of the American Revolution barred black singer Marian Anderson from performing at the group’s Constitutional Hall venue in 1939; this led to the Roosevelts arranging for Anderson to sing on Easter at the Lincoln Memorial. Prejudice was mainstream.

But unashamed prejudice was not confined to people of color. For example, prejudice against Italians is reflected in Life Magazine’s May 1, 1939 profile of Joe DiMaggio:

Although he learned Italian first Joe, now 24, speaks English without an accent and is otherwise well adapted to most U.S. mores. Instead of olive oil or smelly bear grease he keeps his hair slick with water. He never reeks of garlic.

Life Magazine was hardly a fringe publication! Prejudice was mainstream.

And this prejudice certainly extended to Jewish folks. Father Charles Coughlin provides a perfect touchstone for this, since he was an immensely influential figure, reaching millions of devotees through his radio show. A couple of quotes give a taste of what was in the minds of Coughlin and followers:
{exp:list_maker}”… the word ‘fascist’ is merely bandied about as part of Communism’s offense mechanism…. the term ‘anti-Semitic’ is only another pet phrase of castigation in Communism’s glossary of attack.” (
“If Jews persist in supporting communism directly or indirectly, that will be regrettable. By their failure to use the press, the radio and the banking house, where they stand so prominently, to fight communism as vigorously as they fight Nazism, the Jews invite the charge of being supporters of communism.” ({/exp:list_maker}The Catholic church tolerated Father Coughlin’s bile; it did not feel pressured to order Coughlin to either desist or be defrocked. Prejudice was mainstream.

Hank Greenberg certainly experienced this anti-Semitism. Greenberg’s career was full of incidents where he got into arguments and fights when standing up to players who called him names because of his religion. His own words show the flavor of the times:

How the hell could you get up to home plate every day and have some son-of-a-bitch call you a Jew bastard and a kike and a sheenie and get on your ass without feeling the pressure? If the ballplayers weren’t doing it, the fans were. I used to get frustrated as hell. Sometimes I wanted to go into the stands and beat the #### out of them.

Prejudice was mainstream, even in the stands of the National Pastime. So it is quite understandable for people to wonder whether the prejudices of 1930s America extended to American League managers and pitchers to the point that they deliberately kept Greenberg from breaking the Babe’s record.

But now that we have the Retrosheet data, what does it tell? Megdal says it shows that prejudice was a factor. In his own words:

[I]t is also impossible to ignore the statistical record. In short, the American League didn’t seem exactly thrilled with Greenberg’s pursuit. Until the Web site recently published game logs for the 1938 season, the subject of anti-Semitism during Greenberg’s record chase was a matter of opinion…. Almost no other hitter going after the home run record had anything like Greenberg’s late-season spike in bases on balls.

Megdal cites the following stats mined from the Retrosheet data as evidence that this spike existed:
{exp:list_maker}”Greenberg walked in 15.9 percent of his plate appearances through the end of August 1938. In September, that rate jumped to 20.4 percent.”
“…the way pitchers handled Greenberg early in the season was clearly different than the way they approached him as Ruth’s record came into view. Greenberg had four three-walk games in the final two months of the 1938 season, three in September.”{/exp:list_maker}As presented, a 5 percent increase in BB/PA% sounds convincing. Certainly it made me think that if such prejudice influenced performance, the Retrosheet data would also indicate whether this prejudice was league-wide or just confined to a few teams. So to take Megdal’s analysis a step further, I broke down the April-August and September-October BB/PA% for each team:

             April-August      September-October
Team       PA   BB   BB/PA    PA    BB   BB/PA    Diff
BOS        91   18    0.20     9     2    0.22    0.02
CHW        75   18    0.24    22     8    0.36    0.12
CLE        48    6    0.13    47     7    0.15    0.02
NYY        77    9    0.12    17     1    0.06   -0.06
PHA        90   12    0.13     8     0    0.00   -0.13
SLB        73   10    0.14    25     9    0.36    0.22
WSH        90   18    0.20     9     1    0.11   -0.09

The St. Louis Browns were the only team against which Greenberg had a noticeably higher BB/PA% in the period Megdal examined; every other team was in the symmetrical range of a +.12 to -.13 difference, indicating that the rest of the American League pitched Greenberg similarly both before and during September. Another way of viewing this data is to remove the Browns’ games and see how the rest of the league did cumulatively:

Period         PA    BB  BB/PA
Apr.- Aug.    471    81  0.172
Sept.- Oct    112    19  0.170

The Browns are another matter:

Period         PA    BB  BB/PA
Apr.- Aug.     73    10  0.137
Sept.          25     9  0.360

Is this difference evidence that the Browns as a team pitched like anti-Semites in September 1938? Retrosheet does indicate that the Browns were responsible for two of the three-walk games that Megdal cites:
{exp:list_maker}Sept. 5, Game 1 of a doubleheader. Browns starter Lefty Mills issued six walks.
9/27, game 1 of a doubleheader. Starter Jim Walkup issued five walks in seven innings, and Fred Johnson closed out the game by walking two in the eighth. {/exp:list_maker}However, Retrosheet also indicates the following:
{exp:list_maker}Greenberg was not walked in Game 2 of the Sept. 5 doubleheader.
Greenberg was not walked in Game 2 of the Sept. 27 doubleheader.
In Game 2 of the Sept. 27 doubleheader, Greenberg hit home runs Nos. 57 and 58! {/exp:list_maker}
This makes it obvious that the Browns as a team were not employing an anti-Semitic strategy of pitching around Hank Greenberg. Rather, the two three-walk games just seem to be a manifestation of the Browns pitchers’ horrific control. The Browns led the league with 737 walks allowed, 56 walks more than the team with the next-worst control, the Indians, and 121 walks more than the league average. And the pitchers mentioned above were not exactly control artists:

Baseball, You Make It Hard to Love You
Again, a scandal puts fans on the defensive about their game.
Pitcher        IP    BB   BB/9
Mills       210.1   116    5.0
Walkup       94.0    53    5.1
Johnson      69.0    27    3.5

Lefty Mills also comes into play because he started a Sept. 28 game in which Greenberg walked twice. Meaning that at best, the evidence in the data set the Times article made such a fuss over could be that a few Browns pitchers decided to walk Greenberg instead of giving him an opportunity to homer: If you take these three games out of the mix, the Browns actually walked Greenberg less in September than earlier:

Period         PA    BB  BB/PA
Apr.- Aug.     73    10  0.137
Sept.          12     1  0.080

However, I think that after looking at the stats of Lefty Mills and company, the most likely explanation is a few games with poor control. Indeed, if this bunch was trying to pitch around Greenberg it seems they would be just as likely to throw pitches down the middle!

However, are there other ways in which we can examine the data for evidence of anti-Semitic pitching strategies? The first item to examine is the last few games, after the Sept. 27 doubleheader put Greenberg in hailing distance of Ruth’s record. At this point he needed three HRs, with two more home games against the Browns and a three-game series in Cleveland to close the season – that is, against the staffs with the worst control in the league. The games played out as follows:
{exp:list_maker}9/28, Browns: 2 walks
9/29, Browns: 0 walks
10/1, Indians, 0 walks
10/1(1), Indians, 1 walks
10/1(2), Indians, 1 walks {/exp:list_maker}This isn’t exactly Barry Bonds treatment.

Another stat worth a glance is HBP. If the 1938 AL pitchers were so anti-Semitic, it might manifest in bean balls. However, Greenberg was only hit three times that year; third on his team, behind Charlie Gehringer (five) and Billy Rogell (four). Greenberg was not hit by a pitch in September.

Finally, another thought was that Megdal examining September alone was the wrong perspective. Rather, it might be better to look at the season, determine when Greenberg may have gotten on pace to threaten the Babe’s record, and examine the before-after data from that point.

Retrosheet indicates that going into July 6 All-Star Game Hank Greenberg had 22 HRs, one less than Jimmy Foxx. This would have put him on pace for 48 HRs. It was a 15-HR July that propelled Greenberg into a Ruthian pace; more specifically, his pace spiked from July 24 to July 30, when he hit nine HRs in eight games against the Senators and A’s. That week put him on a pace of 62 HRs.

So instead of focusing on September-October as Megdal did, let’s compare Greenberg’s walk rates before and after August:

Period         PA    BB  BB/PA
April-July    404    70  0.173
August-Oct    277    49  0.177

Obviously the league as a whole did not start walking Greenberg after hit a Ruthian stride. How about specific teams?:

             April-July        August-October
Team       PA   BB   BB/PA    PA    BB   BB/PA    Diff
BOS        67   11    0.16    33     9    0.27    0.11
CHW        49   15    0.31    48    11    0.23   -0.08
CLE        48    6    0.13    47     7    0.15    0.02
NYY        54    8    0.15    40     2    0.05   -0.10
PHA        67   10    0.15    31     2    0.06   -0.09
SLB        43    6    0.14    55    13    0.24    0.10
WSH        76   14    0.18    23     5    0.22    0.04

I don’t see anything here; differences around .10 for such small sample sizes seem insignificant. And we already know that Lefty Mills and friends account for the Browns’ difference. However, since Retrosheet lets us examine on a game-to-game basis, let’s see if anything stands out regarding the Carmine Hose.

As it happens, the Red Sox were responsible for one of the three-walk games to which Megdal referred. This was an Aug. 5 game started by Lefty Grove, who walked three batters in six innings; Retrosheet lacks 1938 play-by-play data, so it’s not clear how many times he walked Greenberg vs. his relievers. However, examining Grove’s 1938 games reveals a few interesting facts:
{exp:list_maker}The two games in which Grove walked the most batters occurred against the Tigers. He walked eight on May 24, and six on May 3.
Both were complete games; Grove walked Greenberg walked twice in each game.
Grove had three other starts against Detroit; he walked no one in those. {/exp:list_maker}So a pattern that emerges is that Grove had a few games against the Tigers in which he struggled with his control, and in them he walked Greenberg a few times. With that context, there seems nothing remarkable about the three-walk Aug. 5 game. And if you remove that game, it reduces the difference in the before/after BB/PA percentage to the point of triviality:

     April-July        August-October
   PA   BB   BB/PA    PA    BB   BB/PA    Diff
   67   11    0.16    28     6    0.21    0.05

So to sum up, instead of providing evidence that the 1938 American League pitched around Greenberg for anti-Semitic reasons, Retrosheet indicates that players and managers did not bring whatever prejudices they may have had onto the diamond. This backs up Greenberg’s own words on this topic:

[S]ome people still have it fixed in their minds that the reason I didn’t break Ruth’s record was because I was Jewish, that the ballplayers did everything they could to stop me. That’s pure baloney. The fact is quite the opposite: So far as I could tell, the players were mostly rooting for me, aside from the pitchers.

newest oldest most voted
Northern Rebel
Northern Rebel

Mr Szabo is correct, when he refers to prejudice being mainstream in America, in Greenburg’s days. We take for granted how far we’ve come, because of the bleating of a few, that make money fueling the “race industry.” I’d like to think pro sports was different from mainstream society, due to the concept of competition. If a guy could help you win, it didn’t matter what his personal circumstances. It wasn’t the players that prevented blacks from playing, for the most part. And lest we forget, it was sports that took the first steps. in integration, on almost every front,… Read more »

Northern Rebel
Northern Rebel

BTW George, I am a Red Sox fan, and I play the bloody sock game over and over, and over……


I would be interested in hearing Mr. Megdel’s response to your more in-depth research.  My guess is he probably won’t even return the e-mail.

george szabo
george szabo

Northern Rebel, I just assume that Mr. personally streams the bloody sock game 24/7/365.   anyway, i think your point about competition making sports an early adopter of integration is correct.  reminds me of a funny story i came across researching this piece: >> one of his teammates (Jo-Jo White) walked slowly around Greenberg, staring at him. Greenberg asked him what he was looking at. White said he was just looking, as he’d never seen a Jew before. “The way he said it,” noted Greenberg, “he might as well have said, ‘I’ve never seen a giraffe before.’” I let… Read more »

Steve G
Steve G

great analysis although this blog already beat you to it

george szabo
george szabo

In looking over the article, the difference in its approach is worth commenting on. concludes from Greenberg’s monthly walk rates that the spike in September isn’t significant because it the September rate is similar to the April-May rate, while this article examines walk rates across the season on a team basis. Looking primarily at monthly rates is of limited value because of schedule bias. The Tigers did not play the same number of games aganast each team each month. Since team pitching varies in walk rates, the September spike in walk rates could indicate anti-Semitism was in play… Read more »