What might the careers of Oscar Charleston and Josh Gibson have looked like?



MLB and Baseball-Reference now recognize the top Negro Leagues as major league baseball. A result of this project is we can now find leaderboards, including wins above replacement, for these leagues. The schedules were not the same as the National and American leagues though. We find teams playing 40, 60, 70 game seasons instead of the standard of 154. The Negro League players were not taking it easy, their actual schedules include lots of games against amateur or semi-pro teams, plus there are still probably a lot of league games that we just don't have a box score for.

The work required to give us the kind of historical record we have is really amazing. BBref had an hour and a half talk with some of the key researchers, and it is well worth watching.

I understand the resistance of some to doing projections, MLEs, and stuff like that for these players. What BBref has is factual, if incomplete. A projection might show Josh Gibson hitting 55 homers in a season, or 45, or 75. Who knows what he would have hit if playing a 154 game schedule? But whatever result you come up with is not factual, just a guess, and dependent on the assumptions that you build into the model.

So why am I doing it anyway? I know that other people could build an equally valid or better model that comes up with different results. I know it's not factual. But if I'm trying to rank the greatest players of all time, this kind of model helps. Better than having Oscar rank 210th in WAR for position players, with 48.6. I tried to keep the model as simple as possible. Here's the method for WAR:

  1. Start with the WAR and games played from the Negro Leagues data
  2. Go season by season and estimate how many games he would have played in a 154 game season
  3. For example, a player played 40 games for a team that played 60. So I'd project that he would have played 103 games out of 154.
  4. Take the total number of projected games, and multiply by his actual WAR per game
  5. For Oscar, I project he would have played 2892 games over 25 seasons. His actual numbers show 48.6 WAR in 916 games.
  6. That works out to 153.4 WAR. On the alltime list, that would put him just behind Willie Mays and just ahead of Ty Cobb. There's a lot of uncertainty here about which player should rank where, but to have these 3 in the same ballpark, I'll call that a win.

Once I figure out his season by season game total, I tried projecting what his stats might look like. The way I did this is to borrow stats from adjacent seasons. For example, Oscar played 71 games in the 1925 season. This would work out to 150 games if his team played 154, so I need to borrow 79 games worth of stats. In this case, I borrow 40 games from 1924 and 39 from 1925. Doing this will smooth out some of the extreme numbers that show up when you have shorter seasons. Oscar hit .427 in 1925. Borrowing stats from 1924 (.405) and 1926 (.308) brings that down a bit, to .392. That is still an extreme batting average, but consider the time. Just a year earlier, Rogers Hornsby hit .424 in a full season, including .453 in 80 games from July to September. If the top Negro Leaguers had played full seasons in that time, some of them would certainly have hit .400 too.

Anyway, here's a wild guess as to what Oscar's batting line might have looked like. I am including stats for some leagues that did not get included in the major league designation, but Oscar's stats in these leagues are actually below what he did in the ones that are considered major league. It's not at all similar to Benny Kauff, a .287 hitter in the National league, beating up Federal league pitching for a .357 average.

Year Age G PA AB R H 2B 3B HR RBI SB BB HBP TB BA OBP SLG OPS+ WAR
1917 20 132 573 504 111 181 23 18 12 108 31 45 4 276 0.359 0.416 0.548
1918 21 154 669 587 130 211 27 20 13 123 36 54 4 317 0.359 0.417 0.54
1919 22 154 671 590 146 222 32 22 15 135 44 59 2 343 0.376 0.435 0.581
1920 23 146 638 551 145 207 31 19 15 117 41 63 4 321 0.376 0.443 0.583
1921 24 148 652 559 172 223 34 23 24 150 47 70 6 375 0.399 0.471 0.671
1922 25 145 649 563 153 214 36 23 26 153 37 65 3 374 0.38 0.447 0.664
1923 26 154 683 580 145 218 48 15 27 170 45 80 2 377 0.376 0.453 0.65
1924 27 136 603 503 152 201 47 10 32 165 42 81 1 364 0.4 0.484 0.724
1925 28 150 668 544 179 213 49 8 39 179 51 99 4 395 0.392 0.488 0.726
1926 29 151 673 539 164 205 44 8 34 169 52 108 6 367 0.38 0.489 0.681
1927 30 154 670 544 144 196 38 12 30 143 50 107 6 348 0.36 0.47 0.64
1928 31 154 665 543 141 200 39 15 24 131 30 105 3 341 0.368 0.473 0.628
1929 32 150 658 548 132 190 42 12 21 124 19 90 4 319 0.347 0.442 0.582
1930 33 151 660 569 123 190 45 12 19 124 13 68 5 316 0.334 0.41 0.555
1931 34 154 676 599 119 193 44 11 18 119 21 57 5 313 0.322 0.386 0.523
1932 35 143 604 535 110 175 38 13 14 101 21 54 4 281 0.327 0.393 0.525
1933 36 119 487 432 98 142 29 9 17 98 15 46 4 240 0.329 0.398 0.556
1934 37 108 443 387 72 121 23 5 13 83 12 45 4 193 0.313 0.39 0.499
1935 38 78 313 271 42 81 18 2 7 52 10 32 3 124 0.299 0.379 0.458
1936 39 59 209 181 29 54 13 0 6 33 3 20 2 85 0.298 0.374 0.47
1937 40 38 122 106 16 31 8 0 4 18 0 11 2 51 0.292 0.37 0.481
1938 41 3 8 7 1 2 0 0 0 1 0 1 0 2 0.286 0.375 0.286
1939 42 72 228 195 34 59 12 0 6 30 1 27 2 89 0.303 0.393 0.456
1940 43 37 118 102 16 27 4 0 2 12 1 14 1 37 0.265 0.359 0.363
1941 44 2 6 3 1 1 0 0 0 0 0 3 0 1 0.333 0.667 0.333
Total 2892 12346 10542 2575 3757 724 257 418 2538 622 1404 81 6249 0.356 0.436 0.593 175 153.4

And while we're at it, let's take a look at the greatest catcher to ever play the game, Josh Gibson.

Year Age G PA AB R H 2B 3B HR RBI SB BB HBP TB BA OBP SLG OPS+ WAR
1930 18 90 362 328 66 109 19 7 18 86 0 25 2 196 0.332 0.383 0.598 135
1930 18 1 4 4 0 2 0 0 0 0 0 0 0 2 0.5 0.5 0.5 171
1931 19 154 627 566 113 188 32 13 28 135 2 49 3 330 0.332 0.388 0.583 179
1932 20 131 538 486 102 163 30 12 23 111 3 44 2 286 0.335 0.393 0.588 174
1933 21 143 565 507 118 180 33 13 31 132 7 51 1 332 0.355 0.415 0.655 187
1934 22 1 2 2 0 1 0 0 0 1 0 0 0 1 0.5 0.5 0.5 176
1934 22 119 495 438 110 151 32 9 28 123 12 53 1 285 0.345 0.417 0.651 197
1935 23 84 361 312 86 113 20 7 22 102 14 45 0 213 0.362 0.443 0.683 181
1936 24 93 404 343 107 134 20 10 34 134 12 55 1 276 0.391 0.476 0.805 221
1937 25 94 416 351 120 139 23 12 38 142 6 59 2 300 0.396 0.485 0.855 240
1938 26 131 592 496 172 196 32 16 52 206 10 85 2 416 0.395 0.485 0.839 243
1939 27 74 322 266 83 98 13 6 23 99 6 50 1 192 0.368 0.47 0.722 219
1940 28 37 177 149 48 64 12 4 16 58 4 28 0 132 0.43 0.52 0.886 259
1940 28 1 4 2 1 0 0 0 0 0 0 2 0 0 0 0.5 0 42
1941 29 142 649 538 156 206 42 8 49 190 11 110 1 411 0.383 0.488 0.764 213
1942 30 107 461 379 121 148 21 9 30 149 7 77 0 277 0.391 0.493 0.731 257
1943 31 150 632 532 162 210 33 16 36 192 7 98 1 383 0.395 0.49 0.72 223
1944 32 102 422 364 97 138 18 13 22 114 5 57 0 248 0.379 0.463 0.681 208
1945 33 114 459 401 90 138 18 12 24 109 2 55 0 252 0.344 0.423 0.628 192
1946 34 101 405 353 79 121 18 9 23 97 1 49 0 226 0.343 0.423 0.64 204
Total 1869 7897 6817 1831 2499 416 176 497 2180 109 992 17 4758 0.367 0.448 0.698 206 120.6

It's hard to make a case that Gibson would hold the homerun records, even if he might have been the greatest hitter to ever live. What limits him in this counting stat is that he was a catcher, and so would not have played anywhere close to 154 games per year, and that he died shortly after his 35th birthday. I thought Gibson might have been a part time catcher, playing a lot of first and outfield to keep his bat in the lineup. But the defensive data we have shows he played a little over 90% of the time behind the plate. His projected OPS+ of 206 ties him with the Babe for best ever. A WAR of over 120 puts him well ahead of any other catcher in history, Johnny Bench is second at 75.

Note that I'm not doing any quality of competition adjustments here. That's a work in progress.

This page was last modified 02/25/2022

All content Copyright © 2008-2022 by Sean Smith.