Racism: The Original Performance Enhancer

The Major Leagues of Foxx, Ruth, Gehrig, and Simmons were limited by racial segregation. (via Wikimedia Commons)

One thing that continues to bother me about the historic analysis of baseball is the extent to which we take pre-integration numbers as gospel while throwing into question numbers from various other eras. One look at historic leaderboards should be enough to tell you pre-integration players are substantially over-represented at the top, given the number of players in the major leagues at the time and the fact that pre-integration baseball now accounts for slightly less than 40% of the modern era.

And yet, as I looked, it seemed like no one had addressed this. And that meant it was time for me, the lowly member of the baseball writing humanities wing, to dust off my spreadsheet knowledge and do some calculating.

Approach

The root of this article and my methodology is trying to figure out a way to account for the extent to which the statistics of pre-integration major league players were inflated by the artificial suppression of the talent pool brought on by segregation. During the decades of segregation, the African-American population averaged about 10.3% of the U.S. population. It is thus reasonable to assume at least 10.3% of the white players in the majors wouldn’t have played without integration. Instead, they would have been beaten out by better competition. There’s a good argument to be made that the number should be substantially higher given that, especially in these early decades, baseball tended to draw the majority of its players from the lower classes, and minority populations always have been disproportionately represented among the impoverished in the U.S. Still, I used 10.3% as my number.

The next problem was how to figure out the effect of diminishing the talent pool by that amount. For that, I turned to expansion. Since my interest was primarily in the extent to which all-time greats had their numbers inflated, I looked at the top 30 position players (note: I set the minimum plate appearances significantly lower than “qualified” to allow for the appearance of outstanding partial seasons in the leaderboards) in baseball using the FanGraphs version of WAR to see how it changed from the year immediately before expansion to the year immediately after expansion.

Expansion provides a unique opportunity for analysis: It quickly increases the number of spots available in the major leagues without the build-up in developmental infrastructure that allows for the existence of substantially more major-league quality players. In effect, expansion lets us know what happens when you have more spots available than you have major league players. This is very similar to what segregation accomplished by artificially creating more spots for less-talented white players.

Caveats

A few issues needed to be smoothed out in my data. The first is that the 1960 season was the last season with 154 games. To deal with this, I simply prorated the WAR total for the top 30 players to a 162-game season. Second, there were back-to-back expansion years in 1961 and 1962. For both seasons, I used 1960 as my comparison year since the major leagues hadn’t had time to adjust.

Finally, as different numbers of teams were added and the total available roster spots slowly shifted upward, the percent of the major league population who was there because of expansion changed substantially. For each year, then, I weighted the change in WAR during each expansion year to approximate the rosters being expanded by 10.3%.

Results

By the time all the math was done, I found that expansion increased the WAR generated by the top 30 players in the league by approximately 7.6%. I then applied this percentage to revise the career WAR totals of pre-integration players (using modern-era stats only). The revised top-30 career totals list for position players can be found below.

Revised Career WAR Leaders
Player Adjusted WAR Unadjusted WAR
Bonds, Barry 164.4 164.4
Ruth, Babe 157.3 168.4
Mays, Willie 149.9 149.9
Cobb,Ty 139.4 149.3
Aaron, Hank 136.3 136.3
Williams, Ted 130.4 130.4
Musial, Stan 126.8 126.8
Speaker, Tris 122.0 130.6
Hornsby, Rogers 121.7 130.3
Rodriguez, Alex 113.7 113.7
Collins, Eddie 112.5 120.5
Mantle, Mickey 112.3 112.3
Gehrig, Lou 108.6 116.3
Schmidt, Mike 106.5 106.5
Henderson, Rickey 106.3 106.3
Robinson, Frank 104.0 104
Ott, Mel 103.2 110.5
Morgan, Joe 98.8 98.8
Matthews, Eddie 96.1 96.1
Foxx, Jimmie 95.1 101.8
Yastrzemski, Carl 94.8 94.8
Ripken, Cal 92.5 92.5
Kaline, Al 88.9 88.9
Boggs, Wade 88.3 88.3
Pujols, Albert 87.7 87.7
Brett, George 84.6 84.6
Jones, Chipper 84.6 84.6
Beltre, Adrian 84.1 84.1
Clemente, Roberto 80.6 80.6
Bagwell, Jeff 80.2 80.2

Note that Ted Williams and Stan Musial both had careers straddling the pre- and post-integration eras and that, given the slow integration of all the teams, it’s difficult to properly weight their totals for adjustment. In any case, these new weights would alter their final WAR totals but not their places on the list. Correspondingly, I have left their career totals unaltered on the chart below.

While these changes don’t substantially alter the members of the list, it does alter the order and brings many pre-integration players back to the pack in a substantial way.

Limitations and Suggestions for Further Research

Expansion hasn’t happened enough times in baseball history to provide us with a substantial sample, so it’s obvious these data are imperfect. Further, while using the top 30 players made for simpler analysis, I’m aware top 30 means something different when there are 400 major league players than it does when there are 750. My hope is that, going forward, others will take this as a beginning and do a more in-depth analysis to provide a proper weighting of pre-integration numbers.

An acknowledgement needs to be made that white players from pre-integration years were playing with a form of performance enhancement built into the game, and that their numbers thus need to be taken with more than a few grains of salt. There may even be an argument as to whether we can consider pre-integration baseball to have contained a legitimate major league, though that stance certainly is up for debate.


Jason teaches high school English, writes fiction, runs a small writing program and writes about education and literature. He also writes for Redleg Nation and both writes and edits for The Hardball Times. Follow him on Twitter @JasonLinden, visit his website or email him here.
newest oldest most voted
T. Slothrop
Member
T. Slothrop

Presumably only about 5% of the population of the U.S. was comprised of black *men*, so is the 10.3 figure the best one to use?

Greg Simons
Editor

Only men have ever played in MLB, so 10.3 percent seems a reasonable number.

lilpudge
Member
lilpudge

Right, only about 50% of the population in the US was comprised of *men* so 5%/50% and 10%/100% are equivalent. And the point that baseball was a socially unacceptable career until the tail end of the dead ball era, when Cobb and other elite players began to establish the social phenomenon of star players with national fame, means that the author’s point about lower classes being overrepresented is right, and 10.3% is almost certainly a conservative benchmark.

lilpudge
Member
lilpudge

Honus Wagner has 138 fWAR, unclear why he’s not in the table. Big fan of this article though!

David Ducksworth
Member
David Ducksworth

“Further, while using the top 30 players made for simpler analysis, I’m aware top 30 means something different when there are 400 major league players than it does when there are 750.”

This is tossed out as a throwaway line, but literally undermines the entire analysis.

Richie
Member
Richie

Yup, other than “undermines” and “entire” are overstatements. You just have to then also factor in the 400/750, along with population changes. (and then move into the next dozen ignored variables)

martyvan90
Member
martyvan90

It’s a worthy effort but any model assumptions can be challenged…
-If society had allowed black players to play, would they have played baseball proportionally to the population?
-Would black players have gravitated to football and basketball the way they do now?
Obviously a separate issue but…
-Did steroid use elevate Barry Bonds, ARod numbers?
-Did greenie use affect the numbers of players in the 60’s, 70’s and 80’s?
Interesting work thanks for starting a discuss with a statistical model.

Richie
Member
Richie

Baseball historian Bill James said that a 1970s Lou Gehrig would’ve become a football player instead. Baseball preintegration got pretty much every topnotch white athlete. That lifts the pool substantially. Wouldn’t think it equals the ‘no blacks’ effect, but it’s definitely still there.

rounders
Member
rounders

Black players were over represented during segregation and gradually left the sport along with the fans under integration.

WARrior
Member
Member
WARrior

Trying to correct for integration is fine, but as noted by Marty, one also has to take into account the growth of the NFL, NBA, and numerous niche sports that would lower the effective size of the pool from which MLB players are drawn. To my knowledge, no one knows the degree to which this pool shrinking has occurred.

Shirtless George Brett
Member
Shirtless George Brett

I’m a bit skeptical that the NFL/NBA are really drawing a ton of talent from modern MLB. Sure there is definitely some competition for athletes but ALOT of the players in those other sports were never going to be viable baseball players just based on their physical attributes. 350 lb lineman (which are 50% of the players on the field) probably were never going to be baseball players even if they tried. And, Aaron Judge aside, 6’8″+ guys playing in the NBA are probably not suited very well for baseball. Thats ALOT of people ruled out right off the hop.… Read more »

oddsox1919
Member
oddsox1919

Sure — and we could list dozens more, as each assumption raises multiple questions.
I’ll add just one:

Would these Top 30 players have felt a full 10.3% impact on their performance?
I think Joe DiMaggio would not have felt the pain so much as Vince or Dom.
Paul Waner not as much as Lloyd.

Do you agree?

Ball in Glove
Member
Ball in Glove

Great idea. I don’t think the actual mathematical calculation of the WAR change is as important as the acknowledgement that there would have been an effect to player performance based on a deeper talent pool.

It is also possible that some players could have lost years of accumulating stats because their roster spot would have been filled by another talented player of color.

Paul G.
Member
Member
Paul G.

Quality of play is certainly an interesting topic, and the exclusion of African-Americans would most certainly have played a significant role. However, it is something that is very difficult to quantify. If you look at integration the Negro Leagues produced a number of stars in quick succession, but not every player ended up being a star. Presumably the MLB was selecting the best players available and like with prospects today it can be hit and miss. It was more hit than miss, which is to be expected when artificially blocking talent, but there were misses nonetheless. There are so many… Read more »

Richie
Member
Richie

In Honus’ particular case, everyone else would’ve caught up with him. He was a training tiger for the time.

tz
Member

I tried to put together some “era adjustments” for MLB and came to the conclusion that the sheer number of variables would make it super imprecise. However, I did come to the conclusion that, during the expansion era, the net impact of these items broadly kept up with the number of teams: Early expansion era (1961-1976): explosive growth in US pool (due to baby boom) – early stages of growth in NFL, NBA etc + rapid growth in use of Latin American players = support for growth from 16 teams to 26. Later expansion era (1977-1998): US pool growth –… Read more »

BillCubbieBlue
Member
Member
BillCubbieBlue

Unless you’re including results from the top Negro League seasons this is really just a half-hearted exercise. The data is out there, it can be used, but most never do.

oddsox1919
Member
oddsox1919

But more discoveries almost weekly. Check out Seamheads and watch the data grow!

Patrick Bowe
Member

Would the fact that there were far less team then make up for the smaller talent pool.

Jetsy Extrano
Member
Jetsy Extrano

Expansion is a lowering of the replacement level, so the effect on WAR/162 I figure should be additive (everybody gets +0.3) rather than multiplication (everybody goes up 11%). The old 0 WAR player doesn’t stay 0, he goes up because we brought in a bunch of worse players.

You are pretty robust to this since you narrow to the top 30, but calculating the additive effect might increase the data size and avoid trouble with “top 30” being variable quality over time.

Jetsy Extrano
Member
Jetsy Extrano

Though you might also have found that I’m wrong about it being additive. If the top 30 are a 5 WAR bunch, your 7.5% is about +0.4 wins. Can a 10% expansion mean you dig down to what were -0.4 WAR players? Hard to buy given the shape of the talent curve: the slew of just-under-0 players in the minors at any given time.

Maybe you found a “good hitters feast on the new bad pitchers” effect or something?

Pirates Hurdles
Member
Pirates Hurdles

Wait, is it lowering replacement level? The US population has more than quadrupled since 1900 as the pool for MLB players is actually larger now than back then even when more than doubling the number of MLB jobs.

nathanj
Member
nathanj

While I understand the enticement of social signally premises/articles such as this, I caution restraint with incendiary titles & topics. We may alienate people from our cause (I fear another 4 years of Trump) by re-writing history in this fashion. Of course, African Americans were one of the groups discriminated against in the US/MLB over the course of it’s history. But of course, this narrative ignores all other groups that were discriminated against. For example, Italians were discriminated in baseball/America for a long time. General acceptance sites an inflection point as WWII due to over 1mm Italian Americans fighting for… Read more »

Ryan DC
Member
Member
Ryan DC

If you think the de jure and de facto discrimination against African-Americans was comparable to that of ethnic white immigrants, you may want to read literally any history textbook. Also what “narrative” are you referring to? And what is “our cause”? No offense but this comment is gobbledygook.

fxb
Member
fxb

I have no interest in jumping into the main debate, but I would like to point out that Italians weren’t always considered “white” and were also subject to lynch violence. I’d suggest the recent opinion piece in the Times — “How Italians Became ‘White'” by Brent Staples, Oct 12, 2019.

oddsox1919
Member
oddsox1919

fxb, my heritage includes Italian immigrants.
Interesting about “How Italians Became White,” I’ll find that piece, thank you.

My grandfather’s immigration papers indicated his race as “Italian.”
His death certificate 7 decades later read: “White.”

oddsox1919
Member
oddsox1919

RyanDC, you’re right about discrimination of American Blacks being more intense than that of Jews, Italians, Asians and Latins.
But nathanj’s comment is hardly “gobbledygook.”
With an election year coming up, his call for “restraint with incendiary titles & topics” is appropriate, timely and prescient.

We all need to be a little kinder to eachother — especially when we disagree!

Best regards & Enjoy the rest of the World Series!

ckruschke
Member
ckruschke

It’s “goggledygook” for people who haven’t read “literally any history textbook”. Racisms / Classism was significant for just about every non-WASP ethnic group. Italians were hated, the Irish were hated, the Jews were hated, the Native Americans were hated. Asians were hated. How many good Native Americans were kept out of baseball besides Chief Bender, Chief Meyers, and Jim Thorpe?

oddsox1919
Member
oddsox1919

Stereotypes of Native Americans were a better fit for baseball in the deadball era. “Swift, keen, cunning, etc.” Yes, there was still discrimination, especially by today’s standards. Almost without exception they were all called “Chief. ” The few exceptions included Thorpe, the Olympic champion. But the discrimination was more along the lines of treatment, not exclusion. Moses Whitehorse had his chance (fascinating individual, look him up if you don’t know him.) There was also an element of acceptance, even affection for Native American players And a couple team nicknames were inspired by Native Americans, as you know. Braves and Indians… Read more »

CC AFC
Member
Member
CC AFC

Like, almost all of them since they were subject to mass genocide

Antonio Bananas
Member
Member
Antonio Bananas

Not only did they have the advantage of an artificially small talent pool to compete against, but also think of the pay disparity. Most, if not all of the non Superstar players had other jobs. And had vastly different training and dietary resources as the superstars.

Basically, imagine if Mike Trout only had to play against white dudes, and the vast majority of them were also plumbers, carpenters, and electricians in the off-season.

oddsox1919
Member
oddsox1919

Yes, to get true and fair comparisons you’d have to do away with the Reserve Clause too.
And even then…..

Rich Dunstan
Member
Rich Dunstan

Thanks for this. While acknowledging the validity of many of the reservations expressed here, I think any reasonable attempt to quantify such an elusive point is valuable and praiseworthy. FWIW, I made a somewhat similar attempt a while ago using win share totals from the 60s, the first decade consisting entirely of integrated teams. My very loose calculation suggested the quality of play in the 60s would have been about 20 % worse without the black players.

rmctigerfan
Member
rmctigerfan

Why doesn’t, say, Oscar Charleston get downgraded because he never had to face the top white pitchers?

ckruschke
Member
ckruschke

That was going to be my point. You can’t take the stats from the top Negro League players and transfer them over to MLB. Although there were much less teams, when you have a league made up of only app. 10% of the population, you’ll have to fill rosters with players who don’t really have any reason to be there. So a direct correlation is difficult. Although players like Ruth and Cobb would face better pitching so thus their BA/OBP/SLG would decrease, their R and RBI might actually go up since they may have better players hitting around them. So… Read more »

oddsox1919
Member
oddsox1919

Right — the best players would still have the best numbers. Just as they do today.

Joseph Luchok
Member
Member
Joseph Luchok

While a good philosophical exercise the bottom line is will never know. Players play to the level of their competition and make adjustments as the competition changes. Also the way the game is played has changed. Ty Cobb would have a terror in any era but his stolen base total and batting average would change because of how the game is played. His home run total would be higher.

sbf21
Member
Member
sbf21

I may have missed the WAR adjustments for Bonds and ARod that takes their PED use into account.

johnforthegiants
Member
johnforthegiants

Might be worth mentioning that throughout Babe Ruth’s life there were rumors that he had some black ancestry (based mainly on his physical appearance), and was therefore according to the ‘one drop’ rule black.

TKDC
Member
Member
TKDC

I simply don’t believe this methodology has been even effectively argued as possibly true. At the same time, it seems incontrovertible that the top pre-integration players would have performed worse relative to their peers if more of their peers were eligible to play. I just don’t believe at all you can pick out one number based on general facts, and slap an adjustment on every player that would equate to what would have actually occurred if baseball were always integrated. I think it is enough to know and acknowledge the advantage in WAR that pre-integration players enjoyed as a general… Read more »

sabr16
Member
sabr16

Lazy man’s approach in attempt to try and prove a narrative. Do Josh Gibson and Oscar Charleston get adjusted because they never faced major league pitching?

oddsox1919
Member
oddsox1919

No, and it’s not worth the effort. Just remember them as worthy Hall of Famers and two of the all-time greats.

cwendt
Member
cwendt

Here is the problem with using a 1947 cutoff: Jackie Robinson did not play in an integrated era. Neither did Willie Mays. Both played in an era where there were only a handful of African American players.

Integration was a wave, not a light switch. It would be worth it to think about how that slow integration (which peaked in the 70s/early 80s), and the integration of Latin players, affected the talent pool.

https://sabr.org/bioproj/topic/baseball-demographics-1947-2012

oddsox1919
Member
oddsox1919

Mentioned this upthread, but it bears repeating… we’ll never get the “right numbers” through statistical manipulations.
Just let the records stand until they are broken.

oddsox1919
Member
oddsox1919

Yes, and 5 thumbs up cuz more than a few all-time greats played in both the Negro Leagues and MLB.
Another name that might have appeared on that Top 30 list if baseball had been integrated sooner: Roy Campanella, whose career was split nearly in half.

oddsox1919
Member
oddsox1919

Just an observation: The words “racism” and “race” appear only in the clickbait headline and photo caption.

Sometimes the story’s author and the headline/caption writers are not the same people.

oddsox1919
Member
oddsox1919

self correct: “racism” and “racial.”

oddsox1919
Member
oddsox1919

Were MLB integrated from the get-go, we’d see a few more black names on that Top 30 list. Maybe a half dozen.
My picks (5 in no particular order): Martin Dihigo, Oscar Charleston, Josh Gibson, Pop Lloyd and Cool Papa Bell.
How about you?

oddsox1919
Member
oddsox1919

Some people love crunchy numbers, but I prefer the peanut butter.
Truth is, we’ll never get the right ones through manipulation and revision.

Wanna see and feel difference between today’s game and pre-integrated baseball?
Just watch Game 6 tomorrow night and ask yourself as each batter steps into the batters box or toes the rubber:
“Would he have been allowed to play in 1946?”

Rallymonkey5
Member
Member
Rallymonkey5

“Note that Ted Williams and Stan Musial both had careers straddling the pre- and post-integration eras and that, given the slow integration of all the teams, it’s difficult to properly weight their totals for adjustment.”

Don’t see why that is so hard, we have WAR at the seasonal level. Just apply the adjustment to their pre-1947 WAR totals and add that to their actual 1947+ totals.