Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Sunday, August 28, 2005

Latin American Players and Pitchers Home Page

This thread was set up for Latino players that played either their best baseball or their whole careers outside of the major leagues.

Perucho Cepeda

Pancho Coimbre

Silvio Garcia

Connie Marrero

Carlos Moran

 

 

 

 

 

 

John (You Can Call Me Grandma) Murphy Posted: August 28, 2005 at 06:25 PM | 48 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. John (You Can Call Me Grandma) Murphy Posted: August 28, 2005 at 07:08 PM (#1580066)
Here are two fine players that were requested. Any others candidates that could be considered of at least borderline HOM potential and who are eligible no more than one year after the current election can be submitted here.
   2. sunnyday2 Posted: August 28, 2005 at 08:36 PM (#1580300)
Ultimate Other All-Star Team*

Some of these are beyond our scope of work and I am not implying that we should be considering Japanese players (that is, who played their entire careers in Japan).

I can't speak to the validity of these selections other than that based on what little I know of the NeLs, I certainly disagree with some of their NeL selections--where are Ray Brown and Jud Wilson?

So who knows, right now today, about their Latin selections? I know absolutely nothing--have not even heard of--some of their selections. But I especially note that they've got Francisco Coimbre in their starting OF with Charleston and Torriente, ahead of Stearnes.

Pitchers
+ Satchel Paige
"Smokey Joe" Williams
"Cannonball" Dick Redding
"Bullet Joe" Rogan
Willie Foster
Jose Mendez
Masaichi Kaneda
Adolfo Luque
Alfredo Ortiz
Eusatquio Pedroso
Luis Padron
Ramon Arano
Diomedes Olivo
Tetsuya Yoneda
Yutaka Enatsu

Catchers
+ Josh Gibson
Louis Santop
Katsuya Nomura
Koichi Tabuchi

First Basemen
+ Sadaharu Oh
Tetsuharu Kawakami
Buck Leonard
Julian Castillo
Mule Suttles

Second Basemen
+Martin Dihigo
Sammy T. Hughes
Morimichi Takagi
Bill Monroe
Manuel Cueto

Third Basemen
+ Shigeo Nagashima
Canena Marquez
Oliver Marcelle
Ray Dandridge
Judy Johnson

Shortstops
+ John Henry Lloyd
Dobie Moore
Yoshio Yoshida
Perucho Cepeda
Silvio Garcia
Willie Wells

Outfielders
+ Oscar Charleston
+ Cristobal Torriente
+ Francisco Coimbre
Bernardo Baro
Yutaka Fukumoto
Chino Smith
Koji Yamamoto
Tetelo Vargas
Willard Brown
Turkey Stearnes
Isao Harimoto
"Cool Papa" Bell
Bob Thurman
Andres Mora
Alejandro Oms
   3. Rick A. Posted: August 28, 2005 at 11:59 PM (#1580583)
Are these players eligible to be voted on for our elections? While I'm all for learning about great players that I've never heard of, how exactly is this any different from Japanese League players? I believe we agreed to not vote for Japanese League players unless they had a substantial portion of their careers in the majors (ex. Ichiro Suzuki). Why should Latin players who have never played ML baseball, or even Negro League baseball be eligible for the HOM? This can be another separate wing of the HOM when we're done voting (along with a pioneers wing, managers wing, Japanese wing, etc.)

If this thread is for more information of players like Jose Mendez, Alejandro Oms, and any other Negro Leaguer who found himself in the Carribbean to play baseball, please ignore the above paragraph.
   4. Dr. Chaleeko Posted: August 29, 2005 at 01:36 AM (#1580736)
Coimbre is worth having a look at. If we don't have a thread for him, I think one would be worth having.
   5. John (You Can Call Me Grandma) Murphy Posted: August 29, 2005 at 01:49 AM (#1580764)
Rick, this thread has a dual purpose: 1)As a place where links can be found for certain Latino stars and 2) As a place to discuss Latin player issues that would not belong on any other page.

Your first paragraph does touch on something that I have been thinking about the past few days. I think your questions are a little different for the Latin American players than for the Japanese players? But totally? Maybe not.

Of the Latino players that are not in the majors now, are there any that would be considered potential HoMers (other than Cuba, for obvious reasons)?
   6. Dr. Chaleeko Posted: August 29, 2005 at 02:08 AM (#1580827)
The main difference between Japanese and Cuban or Latino players is that the US leagues before Jackie did not recruit any Asian players (at the big league level at least), but they recruited Latino players as long as they passed the color test.

Or to put it another way, dark-skinned Latinos faced the same types of discrimination that American blacks did. Meanwhile, the Japanese leagues didn't really start in earnest until well into the 1930s or 1940s and weren't accessible to big league scouts (for a variety of reasons) until the late 1940s or 1950s.
   7. sunnyday2 Posted: August 29, 2005 at 02:27 AM (#1580883)
Doc, maybe a more important difference is that we have a frame of reference for the Latinos, because some of them did play in the NeLs and/or the MLs. Up until (what?) 1965 or so, there was very little if any interplay between Japan and the U.S., so little or no way to compare. So I think we have the tools to evaluate the Latinos and not the Japanese, at least as of 1959.
   8. karlmagnus Posted: August 29, 2005 at 02:34 AM (#1580909)
In practice, the major leagues were not available to Japanese players until Nomo in 1995, so if we include Latin players I can see no reason not to include Japanese players from 1947-95. I would suggest that Latin, Japanese and other players be given a spearate wing, but if we decide to vote for Latin players I shall certainly insist on voting for Oh when he temporally becoems eligible (1982?)
   9. Brent Posted: August 29, 2005 at 03:26 AM (#1581021)
I will repeat a recommendation I've made several times in the past. If Japanese players (or at least some Japanese players) or some Latin American players are not eligible for election, someone needs to draft a sentence or two for the Constitution that says who's eligible and who isn't. A lot of thought went into the original constitution, and we need to respect it by amending it now that it's become necessary. It's really not suitable to try to operate according to rules that begin "I believe we agree to..."
   10. Dr. Chaleeko Posted: August 29, 2005 at 01:14 PM (#1581594)
There's another aspect to the Japan vs. Latin Leagues question: contractural obligations. At this time, IIRC, Japan has loosened its contract language to allow a little more player movement between continents. I'd be curious to know whether their contractural language made it realistically impossible for top Japanese players to come over here in the first place.
   11. John (You Can Call Me Grandma) Murphy Posted: August 29, 2005 at 02:28 PM (#1581724)
Brent, I agree. When Joe comes back from his travels, I'll speak to him about it.
   12. KJOK Posted: August 30, 2005 at 02:58 AM (#1583289)
This link on Latin America Baseball Halls of Fame might be helpful:

Other Halls of Fame
   13. JoeD has the Imperial March Stuck in His Head Posted: August 30, 2005 at 07:48 PM (#1584828)
This is a tough one . . .

I say no on Japanese players at this point, not without additional spots being added. As Rick says in post 3; I thought we settled that awhile back. But I would give credit for Japanese accomplishments if the player has a significant MLB career (Ichiro!, for example).

Are Latin leagues like the Japanese league, or are they like another Negro League? I think this should be the litmus test. Does that make sense, or is that off base.

I don't know that I understand how things worked back then enough to answer that. I agree with Brent that we need to settle this as soon as possible and add language to the Constitution.
   14. Chris Cobb Posted: August 30, 2005 at 08:22 PM (#1584921)
I agree that we should specifically exclude Japanese players at this point. The quota is a problem, our general knowledge base is a problem, and getting conversion factors is a problem.

I think that all Latin American players should be eligible, whether or not they played in the North American Negro leagues or major leagues.

They were not excluded as a group from access to the major leagues, and the evidence, so far as I know it, is that Latin American players who fit majors' racial criteria and who had the talent to play in the majors generally did so. Most of the best Latin American players who did not fit those stereotypes played in the Negro Leagues at least for part of their careers. Some didn't, and U.S. racism was a factor in their decisions not to (not to mention the way it made professional baseball less remunerative for non-whites).

For this reason, I don't think that we can draw a line between Latin stars with substantial U.S. careers and those without substantial U.S. careers, esp. given the unsettled finances of the Negro leagues during the 1930s.

I'm doubtful that there are more than a handful of serious Latin candidates with little or no playing time in the top North American leagues. They'll be a challenge to evaluate, though our improving conversion factors for the CWL and the MeL are helping. I certainly don't think they are so many as to make a revision of the election quotas necessary.
   15. Dr. Chaleeko Posted: August 30, 2005 at 08:34 PM (#1584954)
I don't know if anyone else has tried to look at Pancho Coimbre's record, but last night I cracked open my CWL, MxL, Holway, Riley, and PRWL books to get a sense of him.

All I got was a mess.

Some sources have him playing 1926-1950. All sources have him in the NgLs in the forties. Some sources have him in Cuba but don't say when. Figeruero doesn't have him in the index except where the Carribean series is concerned.

By the time Coimbre hit the NgLs he was, according to Riley, about 31 years old. And he hit like the dickens. At the same time in the PRWL, he hit .400 twice in five seasons (1942, 1945) and didn't strike out AT ALL for two or three years. In Mexico in 1945, he hit in the mid .300s. But his career in the 1930s is a complete mystery.

Can anyone help? I think he could be a very serious candidate, but I don't know it because I don't have enough data to do much with him yet.
   16. KJOK Posted: August 30, 2005 at 09:08 PM (#1585054)
Are Latin leagues like the Japanese league, or are they like another Negro League? I think this should be the litmus test. Does that make sense, or is that off base.

In this case, I don't think you can use LEAGUE as your basis. For "black" Latin players, being in the Latin Leagues was very much like another Negro League, because they were not allowed in MLB. For "less-black" Latin players, being in the Latin Leagues was more like the Japanese League - theoretically they COULD play in MLB, but rarely did.
   17. Dr. Chaleeko Posted: August 30, 2005 at 09:30 PM (#1585146)
It's important to remember that blacks and Latinos also integrated the minors after 1947. While quotas at the big league and possibly high-minor league levels kept dark-skinned players out of the highest reaches of the baseball world for a little while, they flocked to the peripheral leagues and made those leagues better until the quota system collapsed under the weight of competitive pressure, and black players finally got the same kind of upward mobility as white guys.
   18. sunnyday2 Posted: August 31, 2005 at 01:11 AM (#1585981)
To me the difference is that enough NeLs and Latinos crossed paths that we have a frame of reference--that is, conversion rates that we have some degree of confidence in. For Japanese players before 1965 or so, or maybe really before Ichiro, or maybe even now what with a still-small sample, I don't see where we have a fame of reference.
   19. Brent Posted: May 09, 2007 at 02:43 AM (#2359012)
Over on the Carlos Morán thread I said that I would be posting information on some of his Cuban League contemporaries. Julián Castillo, a first baseman, was the leading power hitter of the Cuban League during the “aughts.” He was a type of hitter—a big, strong power hitter—that has been common in the major leagues throughout baseball history except for the deadball era, when that type of player essentially disappeared from the majors; Castillo and Negro Leaguer Bill Pettus were perhaps the only players of that type active in the deadball era.

Riley says, “The big, heavy-hitting Cuban was hard as nails and had an equally big reputation for hitting the long ball. While he had good power, he did not distinguish himself afield and was not a good base runner.” It gives his height and weight as 6’2” and 240; I don’t know the origins of these figures, but based on a picture shown on p. 85 of Figueredo, I think the figure for his weight, at least, is exaggerated -- assuming his height is accurate, I'd guess his weight at 210 to 220. Based on the age shown in a passenger list, Castillo was probably born in 1880 or early 1881, which would have made him 20 years old in his first Cuban League seasons in 1900-01. Here are his Cuban League statistics:

Year    Tm              G   AB  R   H 2B 3B HR BB HBP SB  Avg  OBP  Slg
1900
-01 San Francisco  16   66 --  30  5  1  0 --  -- -- .455   -- .561
1901
-02 Fe             17   67 --  20  1  3  0 --  -- -- .299   -- .403
1902
-03 Habana-p       29  112 22  37  1  4  2 --  -- -- .330   -- .464
1903
-04 Habana-p       17   67 17  24  2  5  0 --  -- -- .358   -- .537
1904
-05 Habana         27  104 17  32  2  4  1  9   1  8 .308 .368 .433
1905
-06 Fe-p           24   84 11  20  5  1  0 11   4  2 .238 .354 .321
1906
-07 Fe             30  114 10  29  5  4  0 11   4  5 .254 .341 .368
1907
-08 Habana         39  147 31  47  6  4  0 --  -- -- .320   -- .415
1908
-09 Fe             40  146 19  46 11  1  0 --  -- -- .315   -- .404
1909
-10 Almendares-p   12   49 12  20  3  1  0 --  -- -- .408   -- .510
1910
-11 Almendares-p   18   57  7  12  1  0  0 --  -- -- .211   -- .228
1911
-12 Fe             26   99 18  26  8  3  5 --  --  3 .263   -- .556
1912
-13 Almendares     --   94 13  30  4  1  1 --  --  7 .319   -- .415
Total                  
-- 1206 -- 373 54 32  9 --  -- -- .309   -- .430
pennant 


Black ink (note that leaders not available for all categories for all years):
Average – 1900-01, 1902-03, 1904-05, 1908-09, 1909-10
OBP – 1904-05
SLG – 1902-03, 1904-05, 1909-10, 1911-12
OPS – 1904-05
H – 1900-01, 1902-03, 1904-05, 1908-09, 1909-10
2B – 1900-01, 1905-06, 1906-07, 1907-08, 1908-09, 1911-12
3B – 1902-03, 1903-04, 1904-05, 1906-07
HR – 1902-03, 1904-05, 1911-12
TB – 1902-03, 1904-05, 1906-07, 1911-12

Statistics for Cuban League offensive context are shown on the <a
href=" http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/carlos_moran"> thread.

Castillo did not perform very well in exhibition series again Negro League opponents. Here are his statistics from Gary's agatetype.typepad.com:

Year Tm                         G  AB  R  H 2B 3B HR BB HBP SB  Avg  OBP  Slg
1904 Habana
/Cuban X-Giants      3  12  2  5  1  0  0  2   1  0 .417 .533 .500
1905 Fe
/Cuban X-Giants          3  11  2  3  1  0  0  2   0  1 .273 .385 .364
1906 Almendares
/Cuban X-Giants  4  16  2  4  0  0  0  0   3  3 .250 .368 .250
1907 Almendares
/Phil Giants     5  17  1  5  0  0  0  2   0  2 .294 .368 .294
1907 Habana
/Phil Giants         4  12  1  3  0  0  0  3   0  2 .250 .400 .250
1908 Habana
/Brklyn Royal Giants 7  27  2  5  2  0  0  1   0  0 .185 .214 .259
1910 Almendares
/Leland Giants   8  27  1  6  1  0  0  3   0  1 .222 .300 .259
1912 Almendares
/Lincoln Giants  6  22  2  2  0  0  0  0   0  0 .091 .091 .091
1914 Almendares
/Lincoln Stars   1   1  0  0  0  0  0  2   0  0 .000 .667 .000
Total                          41 145 13 33  5  0  0 15   4  9 .228 .317 .262 


And here are data for the pitcher-excluded series context:

Year Tm SerAvg SerOBP SerSlg
1904 Habana/Cuban X-Giants .205 .280 .231
1905 Fe/Cuban X-Giants .218 .280 .230
1906 Almendares/Cuban X-Giants .220 .299 .251
1907 Almendares/Phil Giants .229 .304 .256
1907 Habana/Phil Giants .229 .304 .256
1908 Habana/Brklyn Royal Giants .204 .276 .228
1910 Almendares/Leland Giants .219 .290 .247
1912 Almendares/Lincoln Giants .252 .344 .296
1914 Almendares/Lincoln Stars .203 .297 .230
Total .222 .297 .250


Here are his statistics against major league opponents:
Year Tm                         G  AB  R  H 2B 3B HR BB HBP SB  Avg  OBP   Slg
1908 Habana
/Cincinnati Reds     2   6  0  0  0  0  0  2   0  0 .000 .250  .000
1909 Almendares
/Detroit Tigers  6  20  4  5  2  0  0  2   1  1 .250 .348  .350
1909 Almendares
/All Stars       2   6  0  4  1  1  0  2   0  0 .667 .750 1.167
1910 Almendares
/Detroit Tigers  5  20  1  0  0  0  0  1   0  0 .000 .048  .000
1910 Almendares
/Phil Athletics  2   7  1  2  0  0  0  0   0  0 .286 .286  .286
1911 Almendares
/Phil Phillies   4  13  1  1  0  0  0  3   0  1 .077 .250  .077
1911 Almendares
/NY Giants       6  23  4 10  2  1  0  0   0  0 .435 .435  .609
1912 Almendares
/Phil Athletics  5  18  0  5  1  0  0  3   0  0 .278 .381  .333
1908
-13 vsMajor League teams 32 113 11 27  6  2  0 13   1  2 .239 .323  .327 
   20. Brent Posted: May 09, 2007 at 02:50 AM (#2359025)
And here are statistics for context:
Year Tm                       SerAvg SerOBP SerSlg
1908 Habana
/Cincinnati Reds    .212   .287   .260
1909 Almendares
/Detroit Tigers .233   .301   .286
1909 Almendares
/All Stars      .227   .278   .291
1910 Almendares
/Detroit Tigers .241   .302   .272
1910 Almendares
/Phil Athletics .220   .295   .276
1911 Almendares
/Phil Phillies  .221   .306   .283
1911 Almendares
/NY Giants      .232   .291   .312
1912 Almendares
/Phil Athletics .281   .359   .350
1908
-13 vsMajor League teams .238   .307   .297 


Castillo MLEs

My calculation procedure is exactly the same as the one described in detail on the Morán thread. I'll add the data I've used to calculate playing time. Similarly to Morán, it is based on the share of team games played, prorated to 154-game schedules:

1900-01 - .89, 1901-02 - .94, 1902-03 - .85, 1903-04 - .85, 1904-05 - .90, 1905-06 - .96, 1906-07 – 1.00, 1907-08 - .87, 1908-09 - .93, 1909-10 - .71, 1910-11 - .62, 1911-12 - .81, 1912-13 - .89, total = 11.2 seasons or 1728 G.

For plate appearances (excluding SH), I compared him to Flick (4.24 PA/G), Donlin (3.99), and Stone (4.21) – I decided to set his PA/G to 4.15, implying 7166 MLE PA.

For context, I used the NL for 1901-13. As was the case for Morán, the effects of adjusting for the low offensive context in Cuba more than offset the effects of adjusting for league quality (again, the quality factors used were .90 for average and .81 for walk rate and for isolated power).

G      PA   AB    H   TB BB+HBP  Avg  OBP  SLG  OPS
1728 7166 6496 2008 3026    670 .309 .374 .466 .839

Avg   OBP  SLG  OPS 
LgAvg LgOBP LgSLG LgOPS BAOBPSLGOPS
.309 .374 .466 .839 .264  .328  .345  .673  117  114  135  149 


Comparisons: I'll compare Castillo to the other sluggers with high OPS+ and poor defensive reputations: Cravath, Frank Howard, and Pete Browning. I rank them in the following order. Cravath (with minor league credit for 1906-07 and 1909-11) has MLE OPS+ of 147 and (converting to 162-game schedule) 2041 G, which IMO places him first in this group. Castillo, with MLE OPS+ of 149 in 1818 G (at 162-game schedule) is second, Howard (OPS+ of 142, 1910 G at 162-game schedule) is third, and Browning (National League equivalent OPS+ of 147, 1531 G at 162-game schedule) is fourth. I'm not a huge fan of this type of player; last election I had Cravath 13th and Howard 22nd (and Browning about 60th). I'll probably slot Castillo in just off the end of the 15-player ballot.
   21. Howie Menckel Posted: May 09, 2007 at 03:04 AM (#2359037)
Excellent work, Brent.
Amazing how many dozens of people make this project better....
   22. Brent Posted: September 07, 2007 at 04:07 AM (#2514757)
A few days ago, the 2004 ballot discussion thread featured a debate about whether Al Cabrera (born in the Canary Islands; longtime Cuban League player) could claim to be the first “Latin American” player with the St. Louis Cardinals. Although the social scientist in me may wish to keep that debate going, I decided instead to post some information about a player who is probably unfamiliar to most readers of this blog.

Alfredo Cabrera (nickname “Pájaro,” which is “Bird” in English) was a shortstop and first baseman who played 19 seasons in the Cuban League from 1901 to 1920. He first played in the United States in 1903-05 with the All Cubans, an integrated barnstorming team that provided the U.S. its first widespread exposure to Cuban League players. He played for minor league teams in New Britain, Waterbury, and Springfield from 1908 to 1915 and made it into the major league record books with one game for the 1913 St. Louis Cardinals at the age of 32.

Here’s his actual Cuban League batting record:

Year Tm  Pos G   AB  R   H  2B 3B HR BB HBP SB Avg  OBP  Slg
1900
-01 Almendares   --  18   64 --  12  1  0  0 --  -- -- .188  --  .203
1901
-02 Almendares   --  17   69 --  13  0  0  0 --  -- -- .188  --  .188
1902
-03 Almendares   1B  19   68  5  12  0  1  0 --  -- -- .176  --  .206
1903
-04 Almendares   1B  19   70 10  13  0  1  0 --  -- -- .186  --  .214
1904
-05 Almendares-p SS  29  106 20  30  3  2  0 12   0  8 .283 .356 .349
1905
-06 Almendares   SS  25   92  9  12  0  1  0  9   1 15 .130 .216 .152
1906
-07 Almendares-p SS  31  111 11  27  2  1  0  8   1 10 .243 .300 .279
1907
-08 Almendares-p SS  44  160 25  43  4  4  2 --  -- -- .269  --  .381
1908
-09 Almendares   SS  35  130 16  23  0  3  0 --  -- -- .177  --  .223
1909
-10 Almendares-p SS  17   64 15  18  5  2  0 --  -- -- .281  --  .422
1910
-11 Almendares-p SS  27   98 11  22  2  0  0 --  -- -- .224  --  .245
1911
-12 Fe  SS  21   76  7  26  2  0  0 --  --  8 .342  --  .368
1912
-13 Almendares   2B  --   73 11  22  0  0  0 --  --  7 .301  --  .301
1913
-14 Almendares-p 1B  33  112 10  30  1  2  0 --  -- 16 .268  --  .313
1914
-15 Almendares   1B  29   94  9  16  2  0  1 --  --  4 .170  --  .223
1915
-16 Almendares-IF   5   16  0   1  0  0  0 --  --  0 .063  --  .063
1916
-17 Red Sox   1B  11   39  2   7  3  0  0 --  --  1 .179  --  .256
1918
-19 Almendares   IF  -—   -— -—  -—  —  —  — -—  -— --  --   --   --
1919-20 América   IF  --   20  2   2  1  0  0 --  --  0 .100  --  .150
Total  
-- 1462 -- 329 26 17  3 --  -- -- .225  --  .272 

p = pennant
Sources: 1904-05, 1905-06, and 1906-07 from data compiled by Gary Ashwill posted on agatetype.typepad.com. Other seasons from Jorge S. Figueredo, Cuban Baseball: A Statistical History, 1878-1961.

The average league hitting context (excluding pitchers) over his career was .233 (batting) and .281 (slugging).

Gary’s site also shows complete data for eight series that he participated in during 1904-14 between American Negro League and Cuban League teams.

G   AB  R  H 2B 3B HR BB HBP SB  Avg  OBP  Slg OPS+
50 169 18 36  1  2  0 15   1  9 .213 .281 .243  95 


Gary also presents statistics from series played in Cuba between Cuban League and major league teams:

G   AB  R  H 2B 3B HR BB HBP SB  Avg  OBP  Slg OPS+
45 153 11 36  4  1  0  9   0 10 .235 .278 .275  78 


Unfortunately, I don’t have Cabrera’s minor league statistics. However, I tried using the same methods for calculating MLEs that I used for Carlos Morán. It’s clear that Cabrera didn’t hit enough to have been a major league first baseman, so I limited the conversion to his 1904-05 thru 1912-13 seasons (converted to the 1905-13 NL). During those seasons, Cabrera played about 90% of his team’s games, so I show him with 1246 games, and assumed 3.65 PA (excluding SH) per game.

Here are Cabrera’s career MLEs:

G   PA   AB H   TB BB+HBP  Avg  OBP  SLG  OPS
1246 4555 4247 1059 1371 308 .249 .300 .323 .623

Avg   OBP  SLG  OPS 
LgAvg LgOBP LgSLG LgOPS BAOBPSLGOPS
.249 .300 .323 .623 .261  .329  .345  .674  /  95  91   94   85 


Of course, throughout baseball history there have been shortstops who’ve been able to put together major league careers with similar (or worse) batting statistics. During Cabrera’s time, examples include Mickey Doolan (1728 G, 72 OPS+) and George McBride (1659 G, 65 OPS+). On the other hand, Cabrera’s hitting was close enough to replacement level that, even if he’d been given a major league opportunity at age 24, a sustained slump could easily have sent him back to the minors. And during that era, players were seldom given second chances.

How was Cabrera as a fielder? Roberto González Echevarría, author of The Pride of Havana, thinks highly of him. He includes Cabrera with José Méndez and Julián Castillo as the three greatest Cuban League stars of the aughts, and lists Cabrera with Luis Bustamante (a contemporary black player) and Silvio García as the greatest defensive shortstops in Cuban League history. It also can be noted that during Cabrera’s 7 seasons as shortstop with the Almendares Blues, they won 5 pennants, fielding a team that was weak with the bat and strong on pitching and defense. (And the pennants weren’t entirely attributable to Méndez—the first two pennants came before he joined the team.) Gary’s posted defensive statistics for the series and seasons; I haven’t systematically analyzed them, but they seem to support the view of Cabrera as a good fielder.

As has been discussed on the Dolf Luque thread, early white Cuban players endured considerable prejudice. In a more just world, Cabrera would have gotten a chance in the majors several years earlier (as would Home Run Johnson, Pop Lloyd, and Luis Bustamante). I can’t blame the Cardinals for not taking him in 1913, though—at age 32, his age must have been showing.

In Cuba, Cabrera went on to manage for at least five seasons, winning at least two pennants (in 1915-16 and 1925-26). He also served as an umpire, and by the 1940s and 50s was head groundskeeper of Gran Stadium in Havana. In 1942 he was elected to the Cuban Hall of Fame.

Cabrera obviously wasn't a HoMer, or even a candidate for the HOVG. I do think, however, that Cabrera deserves to be better known than just as the answer to a trivia question. On the other hand, he wasn’t the only deadball era shortstop who didn’t get a fair trial in the majors—besides the obvious NeLgers, PCL slugger Truck Eagan was another shortstop who would have been an outstanding major league hitter if he'd had the chance.
   23. Dr. Chaleeko Posted: September 07, 2007 at 07:12 PM (#2515470)
Brent,

Do you know what MiLs Cabrera was in?
   24. Dr. Chaleeko Posted: September 07, 2007 at 07:46 PM (#2515532)
New Britain, Waterbury, and Springfield from 1908 to 1915

Working toward an answer of my own question I went to Mike McCann's page and looked quickly for candidate leagues.

Central League
-Springfield, IL (I think), had a team in the 1912-1914 Central League

Connecticut Association/Eastern Association
-Waterbury, CT, had a team in the 1908-1914
-New Britain, CT, had a team from 1908-1912.
-Springfield, MA (I think), had a team from 1908-1914.

3-I League
-Springfield had a team from 1908-1914

Twin States League
-There is this entry for this short-lived league: Springfield-Charlestown
Hyphens. A typhography dream team?

Eastern Illinois League
-From 1907-1908, Springfield did not have a team...but Shelbyville did!
   25. Brent Posted: September 08, 2007 at 02:55 AM (#2516254)
He played in the Connecticut League/Eastern Association. You can see several references to Cabrera in various editions of the Library of Congress Spalding Base Ball Guides; unfortunately, during those years Spalding published the minor league statistics in a separate volume, (which cost an addition 10 cents--you can look it up!)

Another interesting place to look for information on the early years of Cuban players in organized baseball is a new biography of Armando Marsans by Peter Toot. If you go to Google Books and search for "alfredo cabrera" and "new britain," it should pop up. Of course you can only read a few pages, but the pages I read sure painted an interesting picture of the prejudices that early Latino players faced playing in America. This book definitely goes on my wish list.
   26. Brent Posted: September 08, 2007 at 03:55 AM (#2516290)
Scratch "new"--I just looked at the Marsans book on Amazon and noticed that it was published in 2003. Please replace with "new to me" :-)

I guess a great thing about Google Books is that you come across books you otherwise would've missed.
   27. Paul Wendt Posted: September 08, 2007 at 02:38 PM (#2516443)
Central League
-Springfield, IL (I think), had a team in the 1912-1914 Central League


I would have guessed Ohio and that fits Springfield IL in the III League.

Mike McCann now chairs the SABR Minor Leagues Cmte. Probably they should compile all this sharing a "city" code with the Biographical database.
   28. Paul Wendt Posted: December 21, 2008 at 12:47 AM (#3034885)
With this revision the directory will cover all of so-called discussion pages that are noted in my desktop database.

add the line with the link
[Juan] "Tetelo Vargas"
http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/tetelo_vargas
   29. Paul Wendt Posted: December 21, 2008 at 01:04 AM (#3034892)
Evidently I don't have the same authorization to comment on the directory of Negro Leagues player pages.
http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/the_negro_league_home_page/

There all of the entries seem to be correct.
Do they cover all of the NeL players who have discussion pages? I don't know.

The listed player pages cover the 21 old Hall of Fame members including Campanella and Irvin but not Robinson and Doby

The listed player pages cover the 13 of the 17 new Hall of Fame members as expected, including Sol White but not Manley, Pompez, Posey, Wilkinson

The listed player pages cover 15 of the 22 finalists who were not elected. For one or two Taylors that is "as expected". Do any of the seven have discussion pages that should be listed?

Bell William
Dixon Rap
Jenkins Fats
Minoso Minnie
Parnell Red
Taylor Jim
Taylor C.I.

The listed player pages cover 18 of the 55 semifinalists who did not advance to the final balot. Do any of the other 37 have discussion pages that should be listed?

Ball Walter
Baro Bernardo
Bolden Ed
Brooks Chester
Brown Larry
Cannady Rev
Cash Bill
Cockrell Phil
Duncan Frank
Fernandez Jose
Fowler Bud
Gardner Jelly
Greenlee Gus
Harris Vic
Holland Bill
Kimbro Henry
Leland Frank
Manley Abe
Martin J.B.
Martinez Horacio
Mathis Verdell
McClellan Dan
McNair Hurley
Patterson John
Payne Jap
Radcliffe Alex
Robinson Neal
Rogers Nat
Smith Clarence
Stovey George
Walker Moses
Warfield Frank
Wickware Frank
Wiley Wabishaw
Williams Clarence
Williams George
Wilson George

The standard URL for a player page is
: http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/sol_white
where the final component is the player name punctuated with underscore. All you need to do is guess the version of the player name.
   30. Crispix reaches boiling point with lackluster play Posted: December 21, 2008 at 01:21 AM (#3034905)
Pitchers
+ Satchel Paige
"Smokey Joe" Williams
"Cannonball" Dick Redding
"Bullet Joe" Rogan
Willie Foster
Jose Mendez
Masaichi Kaneda
Adolfo Luque
Alfredo Ortiz
Eusatquio Pedroso
Luis Padron
Ramon Arano
Diomedes Olivo
Tetsuya Yoneda
Yutaka Enatsu

Viktor Starffin?
   31. Crispix reaches boiling point with lackluster play Posted: December 21, 2008 at 02:33 AM (#3034961)
did not notice that this is a super old thread -- my suggestion must have been considered and rejected, long ago.
   32. Paul Wendt Posted: December 21, 2008 at 03:33 AM (#3035009)
29. Paul Wendt Posted: December 20, 2008 at 09:04 PM (#3034892)
Evidently I don't have the same authorization to comment on the directory of Negro Leagues player pages.
http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/the_negro_league_home_page/


Let me be clear.
#29 is my comment on the directory of "Negro Leagues" player pages, as #28 is my comment on this directory of "Latin American" player pages. They are both posted here because I am permitted to post here.
   33. Bleed the Freak Posted: January 08, 2011 at 07:34 PM (#3726517)
I wanted to add a comment about the greatest Cuban league player of the last 25 years, and possibly of all-time - although, as I understand it, he is not eligible for the Hall of Merit, since he had legitimate opportunities to play baseball in MLB, but chose not too, or if the criteria is North/Central American players, than Linares is a player I will consider a contender for a ballot slot:

Omar Linares:

http://mopupduty.com/index.php/omar-linares/

"It has been said that Omar Linares is the greatest ballplayer from Cuba to have ever played the game. If you ask the fanaticos down at the Esquina Caliente, they will tell you that he is the best in the world.

“Omar Linares is the best amateur player in the world. He has all the tools. He is fast, a strong arm, power, and above all has a great temperament for the game” – Pinar del Rio manager Jorge Fuentes, 1991.
Omar Linares se encuentra a la sombra de ningún jugador de béisbol.
Translastion: Omar Linares stands in the shadow of no baseball player.
This is a phrase commonly used in Cuba when discussing Omar Linares Izquierdo (born October 23, 1967 in Pinar del Río, Cuba), the greatest ballplayer to ever play on the international stage. Linares, son of former national team fixture and all-star Fidel Linares, began to play baseball when he was still in diapers. He burst onto the Cuban baseball scene at the age of 15, playing for the now-defunct Vegueros, based in Pinar del Rio. Because of his young age he was quickly given the nickname “El Niño” or “The Kid.”

His first foray into international play came 2 seasons later at the Intercontinental Cup in Edmonton, Alberta, Canada in 1985. There he announced his presence by hitting at a .467 clip. Following the tournament, the visionary GM of the Toronto Blue Jays, Pat Gillick, reportedly tendered Linares a contract to play all of the Blue Jays’ home games in Canada (thus sidestepping the US trade embargo). Linares declined, saying “Baseball is so important in Cuba. The most important thing for a player is to be chosen for Team Cuba, to represent our Cuba.”
The 6? 1? 225 lbs. Linares was a third baseman and a bona-fide 5-tool player. He had exceptional power to all fields, quick feet and lightning fast reflexes at third base to go with a cannon of an arm and the ability to hit for a high average. He was fast on the basepaths until late in his career when a serious leg injury and bad knees limited his speed.
Linares led the Cuban team to gold medals in the 1992 Olympics and 1996 Olympics and a silver in 2000. In 1992 at the Olympics in Barcelona, Linares batted .500 (20 hits in 40 at bats) with 4 HR and 8 RBI. At the 1996 Atlanta Olympics, he recorded another 20 hits (20/42) for a batting average of .487 to go with 8HR with 16 RBI. In the gold medal game vs. Japan, Linares hit 3HR - 1 to each field - as he led Cuba to a 13-9 victory. Linares’ flawless defense in Atlanta drew comparisons to Hall of Famer Brooks Robinson. His 2 towering home runs vs. Team USA led him to be compared to Albert Belle, only more “politically correct.”
“He could definitely play here,” White Sox third baseman Robin Ventura said in 1992, four years after he had faced Linares in the ’88 Olympics.

Domestically, Linares spent his entire career with the Pinar del Rio team of the Cuban National Series. Over the course of 20 seasons, Linares collected 2,195 hits and 404 home runs in fewer than 6,000 (5,962) at-bats (the Cuban season is much shorter than Major League Baseball). He hit .368/.491/.644 to go along with a solid 204 steals. He hit an average of 1 home run for every 14.8 at bats. That being said, much of his career was spent using aluminum bats until the Cuban National Series switched to wood in the mid 1990?s.
4 times in his career Linares hit over .400 in a season: .409 in 1985, .426 in 1986, .442 in 1990 & .446 in 1993.

Again, Major League teams attempted to lure Linares to the United States.
Linares is a gifted, solid, everyday, front-line major league player. – 1993 scouting report.
The New York Yankees reportedly offered Linares $40 million to play for them. Not surprisingly, Linares declined. Fidel Linares, the father of Omar, was a 3rd grade dropout from rural Pinar del Río. He owed his own ballplaying career to the Revolution, which developed baseball (and literacy, and medicine) in Pinar Del Rio, the most neglected of all of Cuba’s provinces.
Linares described his decision to remain in Cuba, making the equivalent of $20 per month, as based upon the gains he and his family have made through the Revolution:
“I come from a humble family. I think I owe this to the Revolution. It’s helped me. I’ve been world champion, Olympic champion, Central American champion, thanks to the Revolution.”
”I have all of my family here and the Revolution has given me everything. It has permitted me to study, to practice sports, and to reach the level I have reached. I don’t need to leave my country or the Revolution.” – CNN, 1996.
While priding itself on being an egalitarian society, Cuba goes to great lengths to keep its top players. Linares’ modern car – a burgundy sedan - is in stark contrast to the melange of Chinese bicycles and 1950s relic automobiles that pass by his home along the main street of Havana. The kind of clothes Linares wears are of the quality found in a local North American mall, stylish and comfortable. Linares is paid a special stipend for incidentals - 600 pesos per month – which works out to twice that of an average worker, whose pay is the equivalent of $144 a year.

As an added incentive to keep the top players tethered to the island, Cuba’s sports ministry created the Athlete Care Commission. This commission was designed to provide special benefits to athletes unavailable to the general public:
“Not only are they taking care of the ballplayers but also the ballplayer’s family,” said Martin Hacthoun, a Cuban sportswriter. “If the mother is sick, she is rushed to the hospital. Or if they need cement for repairing their houses, they get the repair.”
“There are people who can’t be bought with money,” said Linares, “and I’m one of them.”
“The player who speaks here is not Omar Linares but the son of a country who rejected a contract of $40 million to make himself a professional.”
- Fidel Castro, April 5, 1999

Americans’ first and only opportunity to see Linares play against Major League-calibre competition was when the Baltimore Orioles and the Cuban National Team played a 2-game exhibition series. Unfortunately, at this point in time Linares’ career was beginning to wind down. Regardless, in Cuba’s game 2 12-6 win, Linares reached base in every plate appearance with 3 singles, a double and 2 walks.
“Imagine what he was in his prime,” current Mets general manager Omar Minaya said at the time. “He could have been Jose Canseco at (shortstop) in his prime.”
Following Cuba’s loss in the gold medal game of the 200o Olympic in Sydney, the Cuban government instituted a program to allow players to play abroad professionally. It was partly to showcase Cuban talent to the world as well as educate potential future managers as to how the game is played internationally. Yet it was also a revenue generator for a cash-strapped government that garnished 80% of the player’s salary."
   34. Bleed the Freak Posted: January 08, 2011 at 07:41 PM (#3726520)
Additional information regarding Linares:

http://baseballguru.com/jalbright/analysisjalbright_c01.html

Top Players in Cuba during the Fidel Castro Era:

Player Career Score
Linares, Omar 618
Vinent, Braudilio 456
Munoz, Antonio 436
Garcia, Rogelio 435
Kindelan, Orestes 425

The number one guy on the list is third baseman Omar Linares at 618 points. Linares is first in career runs scored and slugging percentage, second in career walks and average, third in career hits and homers, ninth in career doubles, and tenth in career steals. In the National Series, he won in batting average five times, runs six times, walks seven times, and triples once. In other series, his wins all came in Select Series. He won batting average, homers, and hits once each, RBI and runs twice each, and walks three times. I'm quite impressed by the broad range of his skills demonstrated by these marks: hitting for average, hitting for power, speed, and plate discipline.

Linares acquired almost mythological status for his ability to supplement these superior marks with a penchant for delivering in key situations on big stages. As with many myths based on historical fact, it can be difficult to separate the myth from fact.

Linares was born in October, 1967, the son of a baseball player good enough to have made the Cuban National team in his own right as an outfielder. Omar broke on to the scene at the tender age of 14, in 1982. This is a major reason he earned the nickname "El Nino" ("The Kid").

At his peak, Linares was just over 6 feet tall and 200 pounds. He was professionally courteous though soft spoken. He was also a staunch supporter of the Castro government. He had a quick bat, foot speed, nimble feet on defense, and a powerful throwing arm.

However, in his final five seasons in Cuban ball, Linares gained weight and appeared to have lost some enthusiasm for the game. Some of this may have been due to nagging injuries exacerbated by an already lengthy playing career. Chief among the injuries were his shoulders and knees. Other issues may have been the general lack of financial incentives for a living legend like Linares to continue playing, but also conditions somewhere between those experienced by Negro Leaguers and modern minor leaguers: travel on buses lacking air conditioning over poor roads, then sleeping in sparse dormitory-style accomodations under ballpark grandstands.

Here's the season-by-season record of Linares' career inside Cuba, which includes the National Series, the Revolutionary Cup, and the Selective Series as given on page 355 of Bjarkman's book, though I will note I have had to calculate AB from H/avg and TB/slg and OBP from (H + BB)/(AB + BB). My sources give the career AB, but my calculations may not add up to that figure.

Played in Cuba from 1982-2002, with career totals as follows:
1700 G
5962 AB
1547 R
2195 H
327 2B
54 3B
404 HR
1221 RBI
1327 BB
368 AVG
483 OBP
644 SLG
246 SB
95 CS
1581 PO
3633 A
285 E
948 %
476 DP
   35. Bleed the Freak Posted: January 08, 2011 at 07:44 PM (#3726522)
Are MLE's possible for the Cuban leagues of 1980s and 1990s/is anyone familar with the league strength of the Castro era?

Esteban Rivera or Brent, what are your thoughts on Linares, as you two are more knowledgeable than most with regard to the Cuban leagues.

If he is a potential candidate, I would suggest a thread is set up in his honor.
   36. Dr. Chaleeko Posted: June 29, 2017 at 09:01 PM (#5485741)
Since the Negro Leagues Stars thread no longer accepts comments, I'm going to do this here....

It's been about a decade since most of the MLEs I worked on were completed, and it feels like it's about time to update the methodology on them. So I hope no one minds me using this spot to do so.

There are several reasons why now is a good time for this work.
1) We have boatloads more data on the Negro Leagues at the level of players, leagues, and teams, and all of it meticulously curated by Gary A. and others at Seamheads.com's Negro Leagues Database.
2) We have more minor league data available at BBREF than ever.
3) We have more splits and PBP for MLB available at BBREF than ever, which can be used to support MLEs.
4) We basically didn't have WAR when I first started doing these, and it's a far more flexible and useful framework for translation purposes.

In terms of the work I did back then, the process I used was a little more complicated than I'd have liked it to be, and I probably didn't listen to people who suggested streamlining ideas back then. The complexity can cause certain annoying knock-on effects along the chain of calculations if the translator isn't very careful. I think I may have a way to cut through a lot of the complexity and possibly resolve several other issues simultaneously. It will necessarily result in losing certain kinds of translated information that many find useful. I'm willing to accept that loss, myself, because I don't find the lost information useful, but there are ways that this information can be restored, though I won't be willing to do that restoration. You'll see what I mean.

OK, so this is the first of a series of posts. Not all of these posts, however, will be coming today, tomorrow, or even soon. I'm developing a new system piece by piece, and I'm hoping that the many wise people in this group are willing to provide feedback that will support a better result. I'll be posting about the primary conversion engine for hitters in the next post this evening.

Before I do that, I want to be upfront about my motivations. There are three:
1) As many of you know, I've got a side project like this one going on elsewhere on the internets, and I'll be including these new translations as part of it.
2) With just a few Negro Leaguers players left on the ballot for the HOM, I want to be sure we have the latest and greatest translations for them.
3) These players deserve that their cause not be left unexamined in general.

OK, so next post is about the conversion engine for hitters. Thanks in advance to anyone who is willing/able to think along these lines with me.
   37. Dr. Chaleeko Posted: June 29, 2017 at 10:21 PM (#5485776)
CONVERTING PERFORMANCE FROM AN ORIGINATING LEAGUE TO MLB
Now let's talk about the conversion of hitters performance from one context to an MLB context. Because we have soooooo much more and better data than we did a decade or more ago, we can now do many things with greater ease and accuracy than we could in the 2000s. These are the broad steps that I'm looking as the framework. Just remember this: I am not a trained statistician, and I am keenly aware that this is a potential sore spot. That's why I'm asking for feedback.

Player = the player being converted
League = the player's originating league
Team = the player's originating team
Season = the season of the player's performance
Qualified = Qualified for batting title by either 3.1 PA/scheduled game or because the NLDB or BBREF says so

1) Find the runs created (RC) by the player in his league.
I'm using Bill James' 2002 RC calculation as expressed here. I'm certainly open to other formulae, but I find base runs to be kind of challenging and strange for individual hitters.

2) Adjust the RC for the player's home park.
This information is currently not on the Negro Leagues database, though there's a rumor it might be coming sometime later this year. If not, there are ways to home cook park factors, and I would use those to get started.

3) Adjust #2 for the quality of the player's opposition as necessary.

Because Negro Leagues teams didn't all play the same schedule and the gap in team quality was much wider than in MLB, if this information were possible to get, we would adjust for it here.

4) Divide #3 by the player's known PAs.


5) Find the standard deviation of all qualifying batters in the league for the season in the category of RC/PA
This is where the new stuff really kicks off.

6) Find the mean of all qualifying batters in the league for the season in the category of RC/PA


7) Compute the player's z-score (or standard score) for the season in RC/PA
The formula being (#4 - #6) / #5

8) Find MLB's standard deviation in RC/PA for all qualified hitters.

9) Find MLB's mean in RC/PA for all qualified hitters.

10) Recontextualize the player into MLB via z-scores.
This means (#7 x #8) + #9

11) Adjust #10 for the league's quality of play.
Over the years, we've talked discount rates many times.

That's the process. I'm not 100% sure I'm always adjusting at the exact right spot, which is one of many places I'm hoping for feedback. Here's a quick example with Josh Gibson from the 1933 NNL. We'll call it a AAA level league and a 0.80 conversion rate, and we'll give it a neutral park value for now. I don't know what the SOS ratio would be so we'll call it 1.0 for now as well.
Josh Gibson
Catcher
Age 21
1933 NNL with Pittsburgh Crawfords
1) Using Bill's 2002 RC and assuming a SB% of 55% (the MLB average), Josh created 61.1 runs.
2) As I said, we'll use 1.0 as the PF here, so he's still at 61.1 runs.
3) Don't know his strength of sked, so he's still at 61.1 runs.
4) Dividing #4 by his 239 PA yields 0.2555 RC/PA
5) The STDEV for RC/PA for qualifying hitters in his league was 0.0402
6) The mean for RC/PA for qualifying hitters in his league was 0.1373
7) So his z-score is (0.2555- 0.1373) / 0.0402 = 2.941
8) MLB's STDEV among qualifying hitters for RC/PA in 1933 is .0335
9 MLB's mean among qualifying hitters for RC/PA in 1933 is .1305
10) Placing Gibson in this context is (2.941 * .0335) + .1305 = .2290 RC/PA
11) Adjusting for QoP is .2290 * 0.80 = 0.1832 RC/PA

So this is saying that in 1933 in MLB if Josh had the same 239 PA, he'd have created 43.8 runs. That's a pace of 110 RC/600 PA. The MLB average in 1933 was 68 RC/600 PA. This is why the z-scores are really important. If we never use the z-scores, and we only adjust down for QoP, then Josh is at 48.8 translated RC, which is 122.7 RC/600. 122.7 RC vs 110 RC may seem like small potatoes, but here's the top-10 in MLB that year in RC:
1. Foxx: 180 RC (161 per 600 PA)
2. Klein: 154 RC (139 per 600 PA)
3. Gehrig: 147 RC (128 per 600 PA)
4. Ruth: 116 RC (121 per 600 PA)
5. Gehringer: 114 RC (97 per 600)
6. Manush: 111 RC (95 per 600)
7. Simmons: 109 RC (101 per 600)
8. Berger: 108 RC (113 per 600)
9. Kuhel: 108 RC (96 per 600)
10. Cronin: 106 RC (91 per 600).

Gibson's translated rate with the z-score factored in fits in nicely here. As a rule of thumb, I generally hesitate to translate anyone as better than Babe Ruth.... ;) But in all seriousness, Gibson was a 21 year old catcher in a short-season league with a much wider variation in the pitching talent he faced. Purely in terms of statistics, I suspect it makes a great deal more sense to use the z-score to translate him rather than dealing with a less imprecise measure such as the ratio of R/G in the respective leagues. Why would I prefer the z-score to that ratio?
1) The z-score method already adjusts for the offensive level without my having to do anything about it because it includes the mean RC/PA.
2) There were more errors per game in the Negro Leagues, so its actual R/G is a little wider than MLB's but it doesn't necessarily reflect real batting skill.
3) Using the entire league to translate the best players may not make the most sense. I suspect this might be a little controversial. But the reality is that we are only interested in the highest quality Negro Leaguers, so it doesn't seem like it makes sense to include not only MLB's onsey-twosey or replacement players, but worse the replacement and onsey-twosey Negro League players who are way way way below MLB replacement. There are whole teams that stank and didn't finish seasons in the Negro Leagues, and keeping the sample to the higher quality players made sense to me. But I could be very wrong!

A place I'm not entirely sure about is whether to adjust for QOP before or after the z-score stuff is calculated in steps 9 and 10. It probably makes no difference, but....

Actually, I'm not even sure whether I should be adjusting for park, SoS, or QoP before the z-score move because, none of the MLB players are adjusted for park or SoS when I find the MLB STDEV and mean. Would it be better to do the z-score transformation first, then adjust for park, SoS, and QoP?

I'd love feedback on that bold/under question. As well as, generally, whether this whole thing makes sense.

Now let me quickly contrast with what I did before to show you why this is less like a Rube Goldberg contraption. Once I got all the adjustments, etc, done what I did before was to use the method Bill James used in the Willie Davis and Sam Crawford comments in the BJNHBA to transform each seasons stat line. Using hits as the key denominator for all other offensive events, each stat line was individually recalculated to fit the new RC total. That's the where the information loss comes in. I'm no longer going to spit out a stat line. It seems kind of pointless in the WAR era. At least for me. If someone else want to do so, however, they could use the same method Bill used to transform the Negro League lines into MLB equivalents just like I did once they saw how many RC I had calculated for the player.

Now, the next step in all this is to create a playing time projection for the player in an MLB setting. I've done a little work on that over at my blog, which everyone can go look at if they feel they want to. I've already pubbed on very young players and on July 26th on very old players in MLB and how these affect potential translations for Negro Leaguers. But that and generally career shaping are the next items up for bids. Once that is established, then we quite easily generate the equivalent of BBREF's batting WAR component called rBat. Then we move on to protocols for estimating baserunning, fielding, DP avoidance (yes, there's even a little we can do there), and finally crib from BBREF's own definitions to create rPos and rRep, then finally WAR itself.

And then we can do pitchers.... ;)

What do you guys think?
   38. Dr. Chaleeko Posted: July 01, 2017 at 12:11 PM (#5486614)
I believe I have answered one of my own questions. Namely that the order of operations in this transformation would need to go like this:

Find RC
Adjust for QoP
Then do the z-score transformation
Then do the Park/SOS adjustments.

The reason for doing it this way goes like this:
1) Trying to keep like-to-like
2) BBREF's RC totals are not (as far as I know) park adjusted, and any adjustment for park and mix of pitchers/opponents occurs within the rBat and/or WAR framework, which would come after the transformation anyway
3) So to compare like-to-like I need to have an MLB-level figure that is not yet park adjusted or opponents adjusted.

Now onto the other 97,023 things!
   39. Dr. Chaleeko Posted: July 01, 2017 at 02:58 PM (#5486666)
Yeah, no, Doc. Have to do QoP after z-scor transformation to MLB context because otherwise, guys who are near average in Negro Leagues can't drop below average.... Duh. Wish it wasn't too late to edit away #38!
   40. Bleed the Freak Posted: July 02, 2017 at 11:26 AM (#5486882)
Looking forward to the analysis Doc, I will try to chime in if I can add value/find time.
   41. Dr. Chaleeko Posted: July 09, 2017 at 03:56 PM (#5490138)
OK, I've been working hard on this for a couple weeks, and I think I'm at a place to share out. Please comment away!

I've got five players that I've worked up. I chose them because they each represented something interesting to me about translations. I've also expanded out the process to generate a complete, mostly BBREF compatible WAR. So here goes. First the method then results for Lloyd, Charleston, Dandridge, Gibson, and Cool Papa Bell. I'll do each in a different post.

UPDATED METHOD
Player = the player being converted
League = the player's originating league
Team = the player's originating team
Season = the season of the player's performance

INITIAL PASS AT BATTING
1) Find the runs created (RC) by the player in his league.
I'm using Bill James' 2002 RC calculation as expressed here. I'm certainly open to other formulae, but I find base runs to be kind of challenging and strange for individual hitters. I've estimated Ks, GIDPs, and ROE by simply assigning league-average amounts of each. I put ROE into the A term of RC. I'm not including SB, CS, IBB, or SH to align with BBREF's policy.

2) Divide #1 by the player's known PAs.

3) Find the standard deviation of all batters in the league for the season in the category of RC/PA
This is where the new stuff really kicks off.

4) Find the mean of all batters in the league for the season in the category of RC/PA

5) Compute the player's z-score (or standard score) for the season in RC/PA
The formula being (#2 - #4) / #3

6) Find MLB's standard deviation in RC/PA for all qualified hitters.

7) Find MLB's mean in RC/PA for all qualified hitters.

8) Recontextualize the player into MLB via z-scores.
This means (#5 x #6) + #7

9) Adjust #8 for the league's quality of play.

#8 times QoP rating (see below).

OK, things get a little intricate here logic-wise. A league's STDEVs are affected by many, many things, including the homogeneity or heterogeneity of the talent top to bottom (the distance between the floor and the ceiling), R/G, expansion, general stochasticness, and, importantly, length of schedule. Some very cheeseball regressions suggested that length of sked may represent ¼ to ⅓ of variance. Some of these things are accounted for by using z-scores, such as R/G, dispersal of performance/talent, etc. Other things may or may not be. So what I'm doing is applying ½ the QoP discount and only to leagues not officially part of organized baseball.

My current discount table looks like this. Remember it's for RC/PA:
*0.90 (roughly NPB: 1923 Cuban Grand Primo; Cuban Winter League 1902-1908, 1918, 1920, and 1923---those seasons were chosen based on the fact that their STDEVs were in a much tighter range)
*0.85 (between NPB and AAA: NNL 1920-1931, NAL 1939-1943)
*0.80 (roughly AAA: Mexican League 1939-1954; NNL 1933-1944; NAL 1944; NAC 1907-1908; INT 1909; EWL 1932, ECL, CWL 1909-1916, 1922, 1927)
*0.76 (between AAA and AA: Cuban Summer League)
*0.72 (AA level: Eastern independent teams, Western independent teams, Independent teams, Mexican League 1937-1938)

These are strongly influenced by the STDEVs I observed since QoP and STDEV are related. I have chosen to exclude two-team setups like the Florida Hotel League of 1916, the various Negro League vs. MLB series, the various Negro Leagues vs. Cuban series, and the Cuba vs. MLB series.

10) Adjust the RC for the player's home park.

#9 / PF
This information is currently not on the Negro Leagues database, though there's a rumor it might be coming sometime later this year. I am using home-cooked park factors that KJOK once described here, and using max 3-year PFs, which isn't always possible due to changes to a park or new parks or new team in the league. If a team moves into a new league, the clock returns to zero.
Process looks like this:
ON YEAR FACTOR
A) (RS + RA) / (lgRS + lgRA)
B) teams in league / (teams in league - 1) + A
C) A * B
D) (75 + (25 * C)) / 100

This yields a single year PF regressed by 75% toward the league.

TWO YEAR FACTOR
Repeat steps A-C above for the subsequent year, then weight this way:
((50 * step C in current year) + 25 + (25* step C in previous year)) / 100

THREE YEAR FACTOR
Same idea: ((50 * step C in current year) + 16.67 + (16.67 * step C in previous year) + (16.67 * step C two years previous)) / 100

11) Adjust for strength of schedule
#10 * SoS adjustment (if one should become available in the future, for now, just use 1.0

PLAYING TIME AND SEASONAL RC

1) Find destination league's PA/G
2) Find the career percentage of the player's team's games in which he played
3) Subtract the player's known games from destination league's scheduled games
4) #2 * #3
5) Add player's known games that season to #4
6) #5 * #1 = PAs in destination league (initial pass)
7) #6 * weighted, rolling average of #11 ((60* Current year)+(15 * Previous year) + (15* Subsequent year) + (5 * Two years prior) + (5 * two years hence)) / 100

This will give you way too many career PAs and RC. Don't worry, it's just a first pass.

To get at a second pass version of playing time, I find 10 or 12 MLB players from the same rough era who appear pretty comparable from the perspective of fielding and hitting ability. Then I chart their annual PAs and find the median at each age to create a general trajectory and apply that trajectory to the player's initial-pass PAs. Next I adjust that second pass to ensure that he's not appearing in more than 95% of his team's games to avoid overprotecting playing time. This is simply by dividing his second-pass PAs by the destination league's PA/G. If necessary, I adjust the PAs down to represent appearing in 95% of scheduled games. I'll finalize by trimming off seasons at the beginning and end of the career where it's clear he would not be a major leaguer (generally at the beginning if he wasn't a league-average hitter, and at the end after two seasons below average) and then adjusting the seasonal PAs based on any injury or other biographical information.

If any seasons are excised, then I remove them from the translation of RC/PA way above so that they aren't dragging down what the player was doing during the years we translate him for.

TRANSLATED VALUE
Once the PAs are finalized and appear to be in line with similar players of the era, we can start creating the BBREF WAR categories.
rBAT: Find the destination league's RC/PA, multiply by the player's translated PAs, and subtract from his translated RC.
rBsr: We have no PBP for the Negro Leagues, so we have to do our best. We also don't have CS data for the most part. So what I do is to compare the player's SB/SBOPP (defined as H-2B-3B-HR+BB+HPB) to the league's. Then adjust that for his team's SB tendencies as compared to the league. Then locate MLB players in the PBP era who have similar careers in this regard, find their career rBsr value. Then once I have that, I simply apportion it yearly by PAs. In some cases, players' speed changes dramatically, and in that case, I try to account for that.
rDP: This won't apply to seasons prior to 1948 since BBREF doesn't have this available yet. For 1948+, I just do what BBREF says here.rField: The Negro Leagues Database at Seamheads.com provides DRA data for many seasons. In some cases it contradicts the oral history of the Negro Leagues. To begin, I look at what positions the guy played and when. If it's known that he switched position, I'll follow suit. In some cases, we don't have all the info, so you have to make a good guess. I see what the DRA implies for a career value. If it's wayyyy out of line with the oral history, then I might adjust upward a little bit.
rPos: Just using BBREF's published values applied to the player's PAs
RAA: Add up all those components
WAA: Do what BBREF says at that same link.
rRep: Per BBREF, (20.5/600) * translated PA
rRar: Add RAA and rRep
WAR: Do what BBREF says at that same link.

Test results coming soon!
   42. Dr. Chaleeko Posted: July 09, 2017 at 04:14 PM (#5490141)
John Henry “Pop” Lloyd
1907-1925
Ages 23-41
Translated positions: SS (1907-1921), 1B (1922-1925)
Inbound translation context: National League 1907-1925
Career totals (with ranking among all MLB players from 1893 through 1950):
G: 2293 (18, just behind Hooper, just ahead of Goslin)
PA: 9533 (22nd, just behind Willie Keeler, just ahead of Al Simmons)
Runs Created: 1249 (50th, between Berra and Kelley)
rBat: +295 (48th, between Slaughter and Fournier)
rBaser: 40
rDP: N/A
rField: +48.5 (This is half his DRA per game rate; he's got a really weird career pattern, stinking early then getting really good later; so I halved due to unsureness)
rPos: +116
RAA: +499
WAA: 54.1 (18th, between G. Davis and J. DiMaggio)
rRep: +326
RAR: +1315
WAR: 89.8 (16th, between Gehringer and Foxx)

Essentially, this would make Lloyd either the 3rd greatest shortstop ever by my methods of reckoning such things, or very close to it. A-Rod and Wagner ahead, Vaughan right next to him.

Here’s what his year-by-year looks like so far w/r/t WAR:
YEAR AGE   PA  WAR
==================
1907  23  581  5.4
1908  24  595  6.7
1909  25  604 10.7
1910  26  609  7.5
1911  27  569  6.3
1912  28  612  5.6
1913  29  607  7.0
1914  30  610  4.5
1915  31  609  6.4
1916  32  603  5.4
1917  33  611  4.2
1918  34  502  4.2
1919  35  564  5.5
1920  36  525  4.3
1921  37  483  3.1
1922  38  362  1.1
1923  39  235  0.8
1924  40  153  0.9
1925  41   99  0.1


By and large, this translation is consistent with what's historically known about Lloyd and with the HOM community's assessment of him back in time.
   43. Dr. Chaleeko Posted: July 09, 2017 at 04:26 PM (#5490145)
Oscar Charleston
1917-1936
Ages 20-39
Translated positions: CF (1917-1925), LF (1926-1929), 1B (1930-1936)
Inbound translation context: National League 1917-1936
Career totals (with ranking among all MLB players from 1893 through 1965):
G: 2225 (21st, between Hooper, Simmons and Clarke)
PA: 9580 (20th, between Keeler and Wallace)
Runs Crated: 1599 ()
rBat: +455 ()
rBaser: 49.5
rDP: N/A
rField: +101
rPos: -72.6
RAA: +533
WAA: 55.3 ()
rRep: +327
RAR: +1567
WAR: 87.2 ()
Charleston appears to have been one hell of a baserunner. Even as he aged, he stole more than the league until late in his career. In the field, I assumed he had an average arm in all seasons (based on conversation at HOM), and his was about a 15-run/year fielder until age 26. Then we don’t have data again until age 28, when he was awful. So what I did was step him down from 15 to 5 to -5, then move him to LF. He had played LF in deference to Torriente in Cuba a couple times, suggesting that this wasn’t crazy. He put on weight rapidly in his late 20s (despite the strong steals numbers), so I stopped him down each of the four years in LF, beginning at 5.5 then 2 then 0 then -5. He appears to have been a good 1B for the rest of his career.

Here’s what his year-by-year looks like so far w/r/t WAR:
YEAR AGE   PA  WAR
==================
1917  20   29  0.2
1918  21  424  5.4
1919  22  565  6.3
1920  23  630  8.9
1921  24  629  9.5
1922  25  637 11.6 
1923  26  619  7.1
1924  27  636 12.6
1925  28  576  3.0
1926  29  609  4.6
1927  30  636  4.6
1928  31  594  3.5
1929  32  535  1.6
1930  33  482 -0.2
1931  34  550  2.7
1932  35  472  1.3
1933  36  382  2.2
1934  37  325  1.0
1935  38  200  0.3
1936  39   50  0.1

We are missing data for 1929 for all players. I used his career average production rate for that year. Also, he was absolutely terrible in 1930, but it's under 20 games. But it affects the surrounding seasons too. If he turns out to be very good in 1929 should that data ever come along, it will give him a boost. This raises the question of whether I should place less emphasis on a single year (remember I'm using 60% current year), but I kind of think not. Otherwise, it's all one over smoothed homogenous blob. This way you get a real good sense of his peak. YMMV.

Overall, this gives Charleston the look of the 4th ranked CF around. He's not as good as Speaker, Cobb, or Mays. But his peak is fantastic, he's got plenty of shoulder seasons even though his WAR looks like Griffey Jr's after 30. And, as i said, there's the possibility of a little improvement on these numbers. Shades of Gray shows 1929 as very strong.
   44. Dr. Chaleeko Posted: July 09, 2017 at 04:37 PM (#5490150)
Ray Dandridge
1937-1952
Ages 23-38
Translated positions: 3B only
Inbound translation context: National League 1937-1952
Career totals (with ranking among all MLB players from 1920 through 1965):
G: 1815 (38th, between Vaughan and half of Hank Aaron)
PA: 7793 (28th, between Doerr and the first half of Hank Aaron)
Runs Crated: 981 (86th, between Pafko and Clift)
rBat: +106 (133rd, between Al Smith and Irish Meusel)
rBaser: 24
rDP: -1.4 (1948-1952 only)
rField: +105
rPos: +7.1
RAA: +241
WAA: 22.7 (59th, between Slaughter and Travis Jackson)
rRep: +266
RAR: +914
WAR: 48.0 (52nd-t Earl Averill)
This conforms to our previous understanding of Dandridge's career. A little above average as a hitter, with good years when he hit for average. The PA total is based on similar long-time 3Bs from an era widely centered on his career.

Here’s what his year-by-year looks like so far w/r/t WAR:
YEAR AGE   PA  WAR
==================
1937  23  576  4.0
1938  24  576  4.4
1939  25  578  3.1
1940  26  573  3.1
1941  27  575  3.8
1942  28  569  2.5
1943  29  567  5.2
1944  30  538  3.1
1945  31  529  3.5
1946  32  502  3.7
1947  33  505  2.7
1948  34  438  3.3
1949  35  512  3.0
1950  36  436  1.7
1951  37  269  1.0
1952  38   50  0.1


Given these numbers, Dandridge would rank well below the in/out line at 3B. He is roughly similar to Ken Caminiti, Bill Joyce, Heinie Zimmermann, and a couple others. He's an occasional All-Star kind of player. Given that he's receiving rField proportionate to his PAs, if his fielding were especially good some years and matched with his better hitting years and a strong running year, he could have been a fringe MVP candidate.
   45. Dr. Chaleeko Posted: July 09, 2017 at 04:54 PM (#5490157)
Josh Gibson
1930-1946
Ages 18-34
Translated positions: C (1930-1936), C/1B (1937-1941), 1B (1942-1946)
Inbound translation context: National League 1930-1946
Career totals, rank from 1920 to 1965 with occasional notes among catchers:
G: 1860 (48th, between Wertz and Pafko—trails Yogi, Hartnett, and Ferrell, just ahead of Lom)
PA: 8016 (39th, between Ennis and Doerr—trails Yogi)
Runs Created: 1338 ()
rBat: +425 (—leads Dickey by 120)
rBaser: -15
rDP: N/A
rField: +15.3 (This includes his being a poor catcher through his first few years, then average, then above average, as well as an above average 1B)
rPos: -6.3
RAA: +419
WAA: 40.3 (—1.8 ahead of Yogi)
rRep: +274
RAR: +1293
WAR: 67.5 (—leads Yogi by 3)
Chris Cobb has translated Gibson with less PAs and a career as basically a catcher or a catcher util the early 1940s. Riley implies that Gibson’s knees were getting pretty bad by the end of 1941, so I’ve got him on the Buster Posey regime. Taking a bunch of starts at first each year but ultimately moving there a la Joe Mauer. In the majors a bat that good just wouldn't stay behind the plate. He wasn’t a great defensive catcher and was a decent 1B so it’s not crazy town.

Here’s what his year-by-year looks like so far w/r/t WAR:
YEAR AGE   PA  WAR
==================
1930  18   32  0.3
1931  19   88  0.5
1932  20  315  1.3
1933  21  385  3.6
1934  22  420  2.3
1935  23  490  4.9
1936  24  525  3.8
1937  25  539  6.7
1938  26  546  5.3
1939  27  553  5.6
1940  28  560  6.4
1941  29  567  3.7
1942  30  613  3.5
1943  31  633  7.3
1944  32  620  3.8
1945  33  580  4.2
1946  34  550  4.5


Although you may not see the gigantic totals you might expect from the Negro Leagues' best hitter, this is actually a fantastic hitter. He blows every contemporary catcher away. You might also think that the PAs are too low through age 23. They might be. But they do represent a realistic scenario based on young-debut catchers. Unless you want him to move to 1B, it's hard to fight with history too much. Toward the end, I've tarted decreasing his PAs owing to the decline in his overall health and knees.

Here’s an interesting comparison. Josh from age 26 on vs Campy’s career and Yogi 26-34:
Gibson, Campanella, and Berra from ages 26 to 34
JG (44.2):  7.3  6.4  5.6  5.3  4.5  4.2  3.8  3.7  3.5  
RC (34.2):  7.1  6.7  5.3  4.4  4.1  3.7  1.7  0.7  0.6  0.1
YB (41.5):  6.3  5.6  5.4  4.9  4.8  4.5  3.9  3.2  3.0

Remember I have Josh getting a lot of 1B time in this span. Anyway, he's the better of the three by a good margin.

Currently some of Josh's park factors are counter intuitive. The method for PFs that I outlined above shows Griffith Stadium as slightly hitter friendly. This is probably because of how good the Grays' offense was thanks to Gibson, Leonard, and others. If the NLDB should provide PFs at any time, this will likely provide some improvement for Josh.
   46. Dr. Chaleeko Posted: July 09, 2017 at 05:11 PM (#5490168)
Cool Papa Bell
1925-1943
Ages 22-40
Translated positions: CF
Inbound translation context: National League 1925-1943
Career totals:
G: 2003
PA: 8671
Runs Crated: 1044
rBat: +3
rBaser: +50
rDP: N/A
rField: -32
rPos: -29
RAA: -9
WAA: -0.3
rRep: +296
RAR: +724
WAR: 24.0
Before you react, get to the bottom...

Here’s what his year-by-year looks like so far w/r/t WAR:
YEAR AGE   PA  WAR
==================
1925  22  200  0.3
1926  23  436  1.4
1927  24  584  1.9
1928  25  630  3.2
1929  26  603  0.8
1930  27  490  0.3
1931  28  630  1.8
1932  29  585  1.7
1933  30  529  1.3
1934  31  456  1.2
1935  32  479  1.7
1936  33  424  0.2
1937  34  450  1.1
1938  35  500  1.7
1939  36  525  1.6
1940  37  550  2.1
1941  38  350  0.8
1942  39  200  0.7
1943  40   50  0.2

Basically, he’s translating as the faster cousin of Lloyd Waner. His batting stats are not very good. While he’s got a fine line for a CF (.318/.384/.429), much of it was compiled in the 1920s and early 1930s, and he had some terrible years tossed in too. But more important is that we are missing a bunch of his seasons: 1926, 1927, 1929, 1930, 1931. Because of that, I can’t claim too too much confidence in the numbers. But his hitting translates badly, his fielding is in the negative on the NLDB, and I actually made it more generous. I also improved the baserunning on the basis that a lot of steals weren’t reported by the local papers as per a Scott Simkus comment to me. There’s no player like what I’ve projected here. Not because Bell is great and unique but because his profile wouldn’t make it past 6500 or so PAs. I projected him that far out with the idea that the missing data could improve him a lot. It might not, who knows?

What I'd say is this: This is what the stats we have with a little historical adjusting yield. You can argue with the method I laid out several posts before this one. But the other players I've worked up have conformed strongly to our impression of them. So I think this is about missing years and a player who looked really great because he hit .300+ in a time when lots of people did. And he was really fast. I'd say, we should keep the case open until we have more numbers.
   47. Dr. Chaleeko Posted: July 09, 2017 at 05:16 PM (#5490169)
Oops, here's rankings info for Gibson:

Runs Created: 1338 (30th, between Vernon and Ashbury, top catcher, leads Yogi by about 80)
rBat: +425 (17th between Simmons and Greenberg—top catcher, leads Dickey by nearly 200)
WAA: 40.3 (19th, between Frisch and Appling—top catcher, 7 ahead of Yogitha)
WAR: 67.5 (20th, between Snider and Simmons—top catcher, leads Yogi by 8)
   48. theorioleway Posted: July 19, 2017 at 07:08 PM (#5496538)
Thanks Dr. Chaleeko, this is great! As someone who figured Bell was overrated historically but still a low-level HOM talent, the #s you have for him are shocking, even with the caveat that we have even less info for him than the little we have for other Negro League players. I hope you're able to continue this for other players!

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
phredbird
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.6251 seconds
60 querie(s) executed