The Better League, 3

June 8, 2022
                                                            The Better League, 3

The Integration Adjustment

           

            OK, let’s take on the issue of integration, and its effects on league quality.  We have two problems here:

1)     How do we calibrate the progress of integration in baseball, one league vs. the other, and

2)      How do we integrate (no pun intended). . .how do we integrate that data into our study?

 

I decided to chart the progress of integration by counting the number of Black REGULARS in each league each year.  Of course, that means that we have to decide Who is black and Who is a regular.  My rules were:

1)     The only issues about who is black have to do with Latin players, of course.  My rule was that a player was only counted as Black if he was dark enough that he HAD to be counted.   Ruben Gomez and Chico Carrasquel. . .not black.  They could have passed and played before integration, at least possibly.  Roberto Clemente. . . .no way.

There are many other people who have documented the progress of integration in baseball, and I used their lists to help sort it out.  The only player that I departed from the other experts on is Vic Power.  For some reason, the lists of players who broke the color line never include Vic Power, but I don’t see how anyone can possibly exclude Power.  The Yankees had Power and Moose Skowron.  Both had great seasons in the American Association in 1953.  The Yankees pretty clearly chose Skowron over Power for 1954 because they weren’t ready to break the color line—or, I guess, that Power wasn’t the right player to do it with.  Power wasn’t Yankee image.  He was a showman, a center-of-attention guy.  The Yankees didn’t really like guys like that no matter what color they were.    Power talked frequently about being disadvantaged by racial discrimination in the US—thrown out of restaurants, etc.—and it was a frequent topic of conversation that Power liked to date white women.   I don’t see how anyone could not consider him to be black athlete.  Also, the business of deciding who was black enough to be considered black in the old days is kind of creepy.

 

2) As to who was considered a regular, my rules were that as long it was reasonable to list a player as a regular, I would list him as a regular.   I decided that, for purposes of my study, there were 8 regular position players and 5 pitchers on each team who should be considered the regulars.  If a player COULD be put on that list without creating a situation in which the team had 14 regulars, then I would list him as a regular.   In other words, if there’s a call that could go either way reasonably, this guy or that guy, then I’d count him as a regular.  But if including the player would force you to count 14 men as regulars, then I wouldn’t count him. 

 

So then, we have counts of the number of Black Regulars in each league in each year beginning in 1947 until the league is considered to be fully integrated.  The National League has:

1 Black Regular in 1947

2 in 1948

4 in 1949

6 in 1950

7 in 1951

5 in 1952

6 in 1953

12 in 1954

14 in 1955

16 in 1956

14 in 1957

20 in 1958, and

23 in 1959.

 

The standard that I decided to use for "full integration" was 20% of the league’s regulars or more being players of color.   In an eight-team league, that means 21 regulars.  13 regulars per team, 8 teams, 104 regulars in the league, 20% is 21.   21 regulars, we consider the league to be integrated, so in 1959—and for every year after 1959—the National League is considered to have completed the transition from segregation to integration. 

Here's a fact that surprised me:  in 1952 there are as many black regulars in the American League as in the National, five in each league.   An out-of-line fact, but. . .that’s the count. 

In the early years of integration, the American League was really just one year behind the National, or one year and one step, let’s say.  It was after 1953 that the gap really opened up.  The American League had

No Black Regulars in 1947, but

1 in 1948,

1 in 1949,

2 in 1950,

4 in 1951,

5 in 1952,

3 in 1953,

5 in 1954,

7 in 1955,

8 in 1956,

7 in 1957,

7 in 1958,

7 in 1959, and

8 in 1960. 

 

            The American League expanded to ten teams in 1961, which raises the standard for integration to be considered reasonably complete from 21 regulars to 26.  (10 X 13 is 130, times .20 is 26.) In 1960 the American League was only 38% of the way toward being considered fully integrated. 

            In 1961 the number in the American League jumped to 18, in part because of expansion, but mostly NOT because of expansion; mostly just because the resistance was crumbling.  The numbers went up in Detroit, Chicago, Cleveland, Minnesota and Kansas City, mostly just up by one per team.  So. .

            18 in 1961,

            20 in 1962,

            19 in 1963,

            20 in 1964,

            22 in 1965,

            24 in 1966,

            24 in 1967, and

            26 in 1968.

 

            So the National League reached the standard of complete integration in 1959, the American League not until 1968. 

 

            Now, how do we integrate this information into our analysis? 

            Let us ask this question.  Suppose that there are two teams in a team, otherwise equal, but one team limits itself by refusing to use black players, while the other team does not.   How much of an advantage does this give to the second team? 

            It is not reasonable to suggest that this is One to Nothing (1.000 to .000), because the integrated team would not have a winning percentage of 1.000, and the segregated team would not have a winning percentage of .000.  But what WOULD the difference be? 

            The numbers have to work with the analysis I am going to explain in a moment, but there are different sets of numbers that would work.   We have to choose numbers that work, and we have to choose numbers that are reasonable. 

            The numbers I chose are .450 and .55738.    This is what I am saying.  First of all, we establish to what extent each league is considered to be fully integrated.  In 1949 the National League had 4 black regulars.  We will consider that to be 19% integrated. 

 

1947

AL

0

21

.000

1947

NL

1

21

.048

1948

AL

1

21

.048

1948

NL

2

21

.095

1949

AL

1

21

.048

1949

NL

4

21

.190

1950

AL

2

21

.095

1950

NL

6

21

.286

1951

AL

4

21

.190

1951

NL

7

21

.333

1952

AL

5

21

.238

1952

NL

5

21

.238

1953

AL

3

21

.143

1953

NL

6

21

.286

1954

AL

5

21

.238

1954

NL

12

21

.571

1955

AL

7

21

.333

1955

NL

14

21

.667

1956

AL

8

21

.381

1956

NL

16

21

.762

1957

AL

7

21

.333

1957

NL

14

21

.667

1958

AL

7

21

.333

1958

NL

20

21

.952

1959

AL

7

21

.333

1959

NL

21

21

1.000

1960

AL

8

21

.381

1960

NL

21

21

1.000

1961

AL

18

26

.692

1961

NL

21

21

1.000

1962

AL

20

26

.769

1962

NL

26

26

1.000

1963

AL

19

26

.731

1963

NL

26

26

1.000

1964

AL

20

26

.769

1964

NL

26

26

1.000

1965

AL

22

26

.846

1965

NL

26

26

1.000

1966

AL

24

26

.923

1966

NL

26

26

1.000

1967

AL

24

26

.923

1967

NL

26

26

1.000

1968

AL

26

26

1.000

1968

NL

26

26

1.000

 

            The standards we are using are .450 and .55738, a gap of .10738.  We apply the 19% (.190) to the gap of .10738, and we have .0204.   Add that to the .450 base, and we have .4704.  We thus conclude that the "integration Strength Index" for the National League in 1949 is .4704. 

            What we are saying, in plain English, is that every segregated league in history, every league in which there are no black regulars, has a presumptive strength index of .450.   Every fully integrated league—the National League since 1959, and the American League since 1968—has a presumed strength index of .557.   In the transition era, each league’s strength index is determined by the extent to which the league is integrated.  And that creates the following numbers for the transition era:

 

1946

AL

 

 

.450

.450

1946

NL

 

 

.450

.450

1947

AL

0

21

.000

.450

1947

NL

1

21

.048

.455

1948

AL

1

21

.048

.455

1948

NL

2

21

.095

.460

1949

AL

1

21

.048

.455

1949

NL

4

21

.190

.470

1950

AL

2

21

.095

.460

1950

NL

6

21

.286

.481

1951

AL

4

21

.190

.470

1951

NL

7

21

.333

.486

1952

AL

5

21

.238

.476

1952

NL

5

21

.238

.476

1953

AL

3

21

.143

.465

1953

NL

6

21

.286

.481

1954

AL

5

21

.238

.476

1954

NL

12

21

.571

.511

1955

AL

7

21

.333

.486

1955

NL

14

21

.667

.522

1956

AL

8

21

.381

.491

1956

NL

16

21

.762

.532

1957

AL

7

21

.333

.486

1957

NL

14

21

.667

.522

1958

AL

7

21

.333

.486

1958

NL

20

21

.952

.552

1959

AL

7

21

.333

.486

1959

NL

21

21

1.000

.557

1960

AL

8

21

.381

.491

1960

NL

21

21

1.000

.557

1961

AL

18

26

.692

.524

1961

NL

21

21

1.000

.557

1962

AL

20

26

.769

.533

1962

NL

26

26

1.000

.557

1963

AL

19

26

.731

.528

1963

NL

26

26

1.000

.557

1964

AL

20

26

.769

.533

1964

NL

26

26

1.000

.557

1965

AL

22

26

.846

.541

1965

NL

26

26

1.000

.557

1966

AL

24

26

.923

.549

1966

NL

26

26

1.000

.557

1967

AL

24

26

.923

.549

1967

NL

26

26

1.000

.557

1968

AL

26

26

1.000

.557

1968

NL

26

26

1.000

.557

1969

AL

 

 

1.000

.557

1969

NL

 

 

1.000

.557

 

            In 1958, when the National League had 20 black regulars and the American League 7, this creates a presumptive strength index of .552 for the National League, .486 for the American.  In all other years, the margin is less than that.    The average for all leagues over time is .50000.

            The key question is, are those numbers reasonable?  They seem reasonable to me; that’s the only defense for them.  They seem reasonable to me.   We are saying that any league post-1968 is presumed superior to any league up to 1946 by a margin of .557 to .450.   That seems reasonable to me.  It’s 22 years; the league gets stronger over time. 

            There is an argument that the numbers should be one and zero (1.000 and .000).   The argument is "Yes, no team is going to play 1.000 baseball against any other team, but this is merely one indicator out of. . ..let us say 35.  We’re not GOING to get to 35, but in theory, that’s the system. . . .a large number of small indicators.  The other indicators will dilute the impact of the segregation index.  But limiting it to .107, rather 1.000, is twice-diluting them, thus reducing the impact to near-zero.

            But I don’t believe that that’s the right logic.   IF the integrated leagues are in fact superior to the segregated leagues. . . .I have no doubt that they are. . .but IF they are, then the other indicators should be able to recognize and give credit for that fact.   If the other indicators don’t show the post-1968 leagues to be superior to the pre-1947 leagues, then we’re claiming an advantage for the integrated leagues for which the only evidence is the presumption that an integrated league must be stronger than a segregated league.  I don’t think we should rely on that assumption more heavily than we need to.  

            These are the league comparisons for the years 1947-1968, using all three of the indicators that we have created so far, and weighting them all equally:

Season

League

BL1

BL2

BL3

Total

1946

NL

.421

.503

.450

.458

1947

AL

.602

.512

.450

.521

1947

NL

.398

.505

.455

.453

1948

AL

.614

.511

.455

.527

1948

NL

.386

.493

.460

.446

1949

AL

.614

.496

.455

.522

1949

NL

.386

.505

.470

.454

1950

AL

.582

.506

.460

.516

1950

NL

.418

.505

.481

.468

1951

AL

.540

.507

.470

.506

1951

NL

.460

.502

.486

.483

1952

AL

.508

.494

.476

.492

1952

NL

.492

.504

.476

.491

1953

NL

.506

.508

.465

.493

1953

AL

.494

.496

.481

.490

1954

NL

.534

.511

.476

.507

1954

AL

.466

.500

.511

.492

1955

NL

.550

.511

.486

.516

1955

AL

.450

.496

.522

.489

1956

NL

.538

.499

.491

.509

1956

AL

.462

.502

.532

.499

1957

NL

.522

.492

.486

.500

1957

AL

.478

.505

.522

.502

1958

NL

.515

.499

.486

.500

1958

AL

.485

.501

.552

.513

1959

NL

.550

.502

.486

.513

1959

AL

.450

.499

.557

.502

1960

NL

.575

.495

.491

.520

1960

AL

.425

.499

.557

.494

1961

NL

.552

.487

.524

.521

1961

AL

.448

.508

.557

.504

1962

NL

.561

.497

.533

.530

1962

AL

.439

.512

.557

.503

1963

NL

.616

.489

.528

.544

1963

AL

.384

.507

.557

.483

1964

NL

.605

.498

.533

.545

1964

AL

.395

.495

.557

.483

1965

NL

.601

.497

.541

.546

1965

AL

.399

.485

.557

.481

1966

NL

.567

.503

.549

.540

1966

AL

.433

.494

.557

.494

1967

NL

.590

.507

.549

.549

1967

AL

.410

.502

.557

.490

 

            So we still show the National League as becoming the superior league in 1953, and remaining the superior league throughout most of this era, with a couple of out-of-line seasons.  The margin peaks in 1965, at .546 to .481, or 65 points. 

            Thanks for reading.  

 
 

COMMENTS (12 Comments, most recent shown first)

Manushfan
As to Mantle vs. Mays-the obvious answer is Musial.
11:48 AM Jun 10th
 
shthar
oh yeah.

Just ask any Strat-O-Matic player about east coast bias.


11:33 PM Jun 9th
 
DefenseHawk
I think to both Marc and Gfletch's points, once Mays and the Giants moved to the West Coast the amount of favorable press Mantle received in comparison (being in baseball's capital) was enormous. The 1961 home run race only added to that.

I do think the "Mantle cult" Marc referred to probably is an offshoot of a Yankee cult. There are a lot of Yankee fans out there.

Was at an LA Kings game with my niece in around Dec 2013 and saw an older gentleman after the game walking to his car. He had on a Yankee jacket with all the championship years listed. I yelled out, "Hey, nice jacket!"

He turned, started to say thanks and then he covered his face in horror and disgust, mumbling, when he saw my Red Sox cap. I told my niece afterwards he's probably going to tell his friends what happened and each time he does he'll be reminding himself who the World Champions are.

To Gfletch's point about the World Series, I think that also points to the weakeness of the American League. The Yankees had the money in those days. They were able to sign the players, pay the bonuses and even stock a farm team in the American Leage (Kansas City).

Whereas Mays and the Giants had to compete against some very strong Dodgers and Braves clubs.
1:55 PM Jun 9th
 
Gfletch
It's always fun to follow your logic, Bill. This exercise is more heavily dependent on assumptions than others, but they are very reasonable assumptions (seems to me).

I also liked DefenseHawk's thoughts as he drilled down to the ancient but still strong debate about Mantle vs Mays. Using the reasonable conclusion that the NL was quite a bit stronger than the AL, you'd think that Mickey was clearly a cut below Willie. But on the other hand, Mickey's Yankees did win the World Series vs the best the NL had to offer. Proving only that dragging the logic of league strength down to individual player comparisons is...tenuous.
11:01 AM Jun 9th
 
Marc Schneider
"I suggest we should all take with a grain of salt comparisons of, for example, Willie Mays and Mickey Mantle, which do not include a huge adjustment for the quality of the two leagues. According to Win Shares, Mickey Mantle deserved his third MVP Award in 1962, beating out Norm Siebern and Floyd Robinson. Meanwhile, Willie Mays competed with a large crop of superstars to win just two MVPs. He might have won every year had he changed leagues."

That's a really interesting point considering how often you see people on Facebook groups talking about how Mantle was the GOAT or would have been without the injuries and so forth. It always seemed to me that Mays was the better player (in part because he was a better outfielder), but there is such a Mantle cult out there. In fact, saying anything even critical of Mantle can get people riled up. I'm not saying it's because he was white (he was a great player after all), but he really struck a nerve in people who were raised in the 50s that Mays probably didn't.
8:47 AM Jun 9th
 
DefenseHawk
The National League didn't just have the best black and Latino players from the late '50 and throughout the 1960s. They arguably had many of the best white players, too.

C Joe Torre v. Bill Freehan
1B Stan Musial v. Norm Cash
2B Pete Rose/Bill Mazeroski v. Bobby Richardson
3B Eddie Mathews v. Brooks Robinson
LHP Sandy Koufax v. White Ford
RHP Jim Bunning v. Denny McClain

Not that the A.L. didn't have some great white players: Mickey Mantle, Carl Yastrzemski, Harmon Killebrew, Al Kaline, etc.

But there's a reason the N.L. dominated the All-Star Game until Fred Lynn hit that grand slam in 1983. The N.L. was the clearly the superior league for much of that time.

I think the reason why the superiority of the N.L. during the period of integration has largely been avoided by much of the sabermetric community is because of just that: integration. It's a touchy subject.
2:13 AM Jun 9th
 
DefenseHawk
Frank, the National League just didn't have more black and Latinos than the American League in the '60s, they had the predominance of black and Latino star players.

I've noted this elsewhere, but Mark Amour did a study on
Baseball Integration, 1947-1986, originally published as “The Effects of Integration,” in the Baseball Research Journal 36 (SABR, 2007).

Using Win Shares, he found that "[In] 1965, black players made up 20% of team rosters, and fully 44% of the “star” players in baseball. This is not an anomalous season—these results are repeated to some extent for most every season for 20 years."

Looking at just the N.L., Armour found that "By the early 1960s, half of the stars in the league were black, and the number was over 60% by 1967."

How about Hall of Fame players from that period?

"The NL added a new Hall of Famer nearly every season, until 1965 when their gap on the Americans was 15-0. In 1966, Frank Robinson was traded to the Orioles, reducing the gap to 14-1, and, perhaps not surprisingly, he was immediately the best player in the AL, winning the Triple Crown."

Armour concluded:

"Since each league has the same number of wins, they will also have the same number of Win Shares. To the point, in order for the leagues to be of comparable strength in the 1960s, the white American Leaguers would have to have been significantly better than the white National Leaguers.

"Could this be true?

"Returning to 1965 again, who were the best players, of any color, in the American League? According to Win Shares, the best players were Tony Oliva, Zoilo Versalles (who won the league’s MVP award), and Don Buford, three fine black players. Going down the list, the best AL white players that year were Rocky Colavito, Brooks Robinson, Curt Blefary and Jim Hall. How much better could they really have been than Sandy Koufax, Don Drysdale, Pete Rose, Jim Bunning, and Ron Santo, each of whom had excellent seasons that year in the NL?

"I suggest we should all take with a grain of salt comparisons of, for example, Willie Mays and Mickey Mantle, which do not include a huge adjustment for the quality of the two leagues. According to Win Shares, Mickey Mantle deserved his third MVP Award in 1962, beating out Norm Siebern and Floyd Robinson. Meanwhile, Willie Mays competed with a large crop of superstars to win just two MVPs. He might have won every year had he changed leagues."

https://sabr.org/bioproj/topic/baseball-integration-1947-1986/​
1:31 AM Jun 9th
 
FrankD
if there is a racial component showing which league is better then this difference should show up in league comparison of statistics, not in a diversity tally. I'm not arguing that having more of everybody can't be better, I'm arguing that just having more of everybody doesn't make it a priori better. If so, then BJ had better add correction factors as the world of MLB has expanded through time brought in more Hispanic and Asian players.
12:19 AM Jun 9th
 
willibphx
Would it be possible to observe the effect of integration on BL2 as a reasonableness check? Is BL3 needed or isn't the effect of integration already part of the other metrics?
12:01 AM Jun 9th
 
CharlesSaeger
Apropos of nothing: I’ve been figuring defensive stats for the teams of the last century, and have done about 900 so far. Vic Power was the first baseman on the 1959 Indians, who were at 21 runs saved vs. league, best I’ve figured for first basemen so far. (The opposite number for this is one of the Philly teams on which Nick Etten played first.) The man was a helluva fielder.
11:44 PM Jun 8th
 
Manushfan
Great stuff Bill.
8:25 PM Jun 8th
 
bearbyz
The last chart is a little off as you sometimes have the American League BL3 in the National League box. For example 1959.
4:18 PM Jun 8th
 
 
©2022 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy