As pretty as Dorothy Dandridge was, Ray was just as pretty at third.

Ray Dandridge Data

From Riley

Born August 31, 1913

Teams: 1933 Det/Nash/Newark, 1934-39 Newark, 1939 Venezuela, 1940-41 Mexico, 1942 Newark/Mexico, 1943 Mexico, 1944 Newark, 1945-48 Mexico, 1949 NY Cubans, 1949-1951 Minneapolis in American Association, 1952 PCL, 1953 lower minor league

From Holway

1933 .218 for Det/Newark (no data from Nashville); ut
1934 .408 (1st in league); 3b, no all-stars listed
1935 .340, 9 2b (5th), 9 3b (1st); 3b, all-star
1936 .287; 3b
1937 .342; 3b
1938 .346, 2 3b (1st); 3b
1939 no data, in Venezuela
1940 .347 in Mexico
1941 .367 (3rd) in Mexico
1942 .186 for Newark, .310 in Mexico; 2b in Newark
1943 no data, in Mexico
1944 .370 (3rd); ss
1945-48 no data, in Mexico
1949-53 no data, in the minor leagues

career
.350, 406-1160
19-49, .347 vs. major-league competition
mean avg. .327 for 10 recorded seasons
black ink/gray ink 6/16

career data according to Macmillan 10th
234 g, 878 ab, 283 hits, 26 2b, 13 3b, 5 hr, .322 ba, .399 sa

Obviously, his full batting totals from Mexico, his 11 seasons in the Cuban Winter League, and his 4 seasons in AAA minors are going to be crucial to any evaluation of him. If he had strong plate discipline to go along with his skill in hitting for average, I think he would have made a good major-league hitter, a lot like Stan Hack, it would appear, probably with better OBP but less slugging. If that’s the case, was his outstanding defense enough to make him a HoMer? We’ll see.

Here are Dandridge's stats pulled from a nice site that has some stats for the Minneapolis Millers from1941-1950 and 1951-1960 among other years.

Ray Dandridge

*Riley has his birthday as 8/31/13, so I inserted his age as of July 1st

Looks to me like Dandrige was a MLer in 1949, but even that is a small sample. By 1950 he looks to be borderline already. I agree we are going to need a lot of help on this one. I mean, .350 is OK, but 44 XBH in 878 ABs, does this make him George Kell?

Hopefully we can get his 2Bs, 3Bs and BBs in Minneapolis.

Unless Dandridge's (1) plate discipline and xbh aside from home runs were very poor or (2) Minneapolis was an extreme hitters park, Dandridge was not a borderline MLer in 1950-1951. He was at least an average player. He was around major-league average as a hitter (give or take 10 points of OPS+, depending upon the factors we don't know). That's easily enough hitting for a very good defensive 2B/3B player to hold down a major-league position.

Here are the OPS+ scores for the starting 2B/3B pairs in the NL in 1950:

Bkn 140/74
Cin 88/96
Bsn 87/140
NYG 130/123
Phi 80/108
Stl 83/124
Pgh 100/88
Chi 77/99

If Dandridge has a 95 MLE OPS+ (a conservative view of 1950, IIRC he was elected MVP that year . . . ) and offered excellent defense at second and third, then he could have improved any team in the league _except_ the Giants, who had the rights to them. Eddie Stanky and Hank Thompson had Dandridge very effectively blocked, particularly given Dandridge's age.

The Riley power numbers don't look good for Dandridge's case, I agree, but there's a big discrepancy between Dandridge's power numbers (as we have them) for Minneapolis and his power numbers in the NeL as we have them from Riley.

In Minneapolis, he had 35 hr in (very approximately) 2000 ab.
In the NeL, he had 5 hr in 878 ab.
His later rate is triple his earlier one.

Maybe this is a natural career progression as Dandridge learned to hit for power. Maybe it's park effects. Maybe it's a problem with Macmillan's data. The Mexican data, when we get it, should serve as a bridge between the two sets.

Anyway, this post is mostly to urge against jumping to conclusions about Dandridge when our data is so fragmentary.

Chris, I agree we need a lot more info and analysis and I said so above. Just trying to frame the question a little bit. I think you'll find the dimensions of the Miller's Nicollet Park somewhere, probably the same place you found Dandridge's record.

Stew Thornley, whom many of you know from his involvement in SABR, has a Web site with a very comprehensive history of the Millers. Check him out. I do know that both Ted Williams and Willie Mays spent their final MiL games at Minneapolis and both hit in the high .400s there. Mays was something like .467 for half a season.

Sunnyday2, also Yastrzemski, in 1960, which is close enough to Dandridge's period there to be relevant. Yaz was at best a marginal MLer in 1960, though, as evidenced by his 91 OPS+ in 1961.

OK, Nicollet Park was a "bandbox," according to Thornley, with a 279' 10" RF foul pole. Obviously this isn't the whole story as to park effects...

Williams only hit .366 at Minneapolis in 1938, with 42 HR and 142 RBI winning the Triple Crown. Mays hit .477 with 8 HR and 30 RBI in just 35 games.

Here is what Thornley says about Dandridge:

RAY DANDRIDGE, Second baseman-third baseman, 1949-52
Ray Dandridge became the first black player to play for the Millers in June of 1949. Dandridge had been signed, along with Dave Barnhill, off the roster of the New York Cubans of the Negro American League.
A 16-year veteran of the Negro and Mexican Leagues, Dandridge was regarded as the greatest third baseman in the history of the Negro Leagues. With the Newark Eagles in the late 1930s, Dandridge was a member of the “Million-Dollar Infield,” so called because it was said that’s what it would have been worth had the players been white.
In the early 1940s, Dandridge left Newark to play in the Mexican League. He became a close friend and a trusted confidant of league president Jorge Pasquel, who often sent Ray back to the states to recruit other Negro League players.
At one point Dandridge and Pasquel had a falling out after a salary dispute. Dandridge was on his way back to the United States when his train was stopped by the Mexican Army, who informed Dandy that Pasquel had changed his mind and was prepared to raise his salary. Despite the occasional hard feelings between the two, Dandridge remained loyal to Pasquel, and even turned down an offer by Bill Veeck to play with the Cleveland Indians in 1947. In 1948, though, Dandridge returned north and became the playermanager for the New York Cubans until he and Barnhill left to join Minneapolis.
Dandridge was 35 at the time, but claimed he was only 35. (After the color line had been broken, many older Negro Leaguers shaved a few years off their age, hoping for an opportunity to plate in the majors.) For Dandridge, however, that chance did not come.
Besides his age, there are other explanations as to why he was never called up by the New York Giants. The Giants already had three blacks on their roster, and an informal quota system was said to still exist. In addition, Giants president Horace Stoneham told Dandridge he was too good a drawing card in Minneapolis to be moved to New York. There was even speculation that his chances were hurt because of his involvement in the Mexican League, which, by the late forties, was viewed as an “outlaw league” when it began raiding players from the American and National League rosters. Although never given the chance to play in the majors, Dandridge was elected to the Hall of Fame in 1987.
In his rookie season at Minneapolis Dandridge hit .311 and had a 28-game hitting streak during the year. He also had three home runs and eight RBIs in one game in August.
Dandridge’s 1950 performance in the field and at the plate (a.311 batting average with 106 runs scored and 80 batted in), earned him the league’s Most Valuable Player award.

Just as an FYI on Dandridge, Wright, Troupe, and others. I'm working on assembling the hitting context for the Mexican League from 1939-1951. It involves basically just going page by page and noting every player's AB, H, TB, BB, then dumping them into a spreadsheet later on.

I'm midway through the letter D after a few hours on-and-off, so I hope to have a fuller context soon, but it could be a little while.... By the time Dandridge and Wright come up, however, I should have it done...I hope. I'm also in the process of moving to a new home.

Update: this process is now done. Because Dandridge isn't coming onto the ballot for another couple real-time weeks, and because I've got to drive five hours tomorrow morning, and it's already 1:33 AM, I can't post my analysis of his Mexican League numbers right now. With any luck, I'll be able to by late next week once my computer is up and running again after the move is completed.

GENERAL QUESTION: Is there a current thread in which anyone would like me to post all the pitching and batting information from the Mexican League? Or would it be useful to have its own thread for reference? Right now, the pitching stuff is buried in the Re-evaluating NgL Pitchers thread, and it's not in chronological order either, so maybe this would be for the best.

First, Dandridge has some detailed stats in "The Negro Leagues Book", so I'll put them in the yahoo group if someone wants to do MLE's.

Second, maybe we need an 'other leagues' thread as I think Gary will shortly have some Cuban League batting/pitching, and we've had PCL and other leagues also scattered about, so maybe a central place for them would be good - or maybe just post a spreadsheet into the egroup?

RAY DANDRIDGE, MLEs FOR MEXICAN LEAGUE SEASONS

Before we pack away the computer in the moving van early tomorrow morning, i thought i'd post up this....

MLEs based upon the following assumptions…

-That the Mexican League conversion rate was .90/.82 each year of Dandridge’s career. In fact, I suspect this is not true and that the league varied considerably, being much closer to 1.0/1.0 in 1940-1941 and in 1946-1947, and probably being under .90/.82 in 1945 due to wartime personnel issues.

-That Dandridge’s playing time converts in a straight line from Mexico to MLB. In other words that if he played 80 games for a team that played 80 games, I assign him 154 games in the majors.

And has the following caveats…
1. I have not yet assembled the data for the Mexican League in 1945, so I used an average of the surrounding four seasons batting and slugging averages as the league’s hitting averages.

2. My team decisions data often differs from published team won-loss records, and in several instances, Dandridge’s games played exceeded the number of decisions his team’s pitchers accumulated (per my database). In these instances, the discrepancy was typically not very big, and I made these choices to deal with them:
-1943: Dandridge exceeded my team decisions data by two games, however, his games matched the total games reported by Juan Fernando Rivera Pernia on the Baseball Archive. I went with Pernia’s total.
-1947: Dandridge’s games played exceeded the total of his team’s decisions by 6 games: 122 for Dandridge, 116 for Eric. Pernia’s standings show 119. I simply opted for 122 games.
-1948: Dandrige’s games played exceeded the total of his team’s decisions by two games, 88 to 86.. Pernia matches my decisions total. I opted to go with 88 games.

3. Dandridge’s walk rates relative to his league receive no conversion discount. This is because I have no NgL walk data to work with.

4. No regression.

5. OBP always calculated in simple form: hits + walks / hits + walks

Dandridge’s MxL OPS+                                                    
                           Mxl  mxl  mxl  mxl   mxl  mxl  mxl
year  ab  avg  slg  OBP    avg  slg bb/h  obp   obp+ slg+ ops+
--------------------------------------------------------------
1940 127 .346 .480 .371   .290 .420 .400 .364   102 114 116
1941 430 .367 .521 .415   .288 .396 .428 .366   113 132 145
1942 142 .310 .458 .380   .289 .394 .466 .373   102 116 118
1943 370 .354 .505 .399   .273 .367 .402 .345   116 138 153
1945 344 .366 .483 .408   .280 .377 .431 .358   114 128 142
1946 418 .323 .431 .375   .281 .381 .422 .357   105 113 118
1947 514 .329 .411 .362   .278 .366 .435 .356   102 112 114
1948 369 .369 .471 .393   .273 .378 .429 .349   113 125 137
==============================================================
    2714 .347 .468 .389   .280 .380 .426 .357   109 123 132


Dandridge’s MLE stats
     mxl  mxl                  
     avg  slg               
     conv conv mle  mle  mle   mle  mle  mle  mle  mle mle  mle
year rate rate avg  slg  bb/h  obp   G    AB   H    tb  bb   pa 
---------------------------------------------------------------
1940  .9  .82  .292 .366 .095 .311   38  145   42   53   4  149
1941  .9  .82  .305 .405 .194 .344  152  579  177  234  34  614
1942  .9  .82  .251 .339 .306 .304   67  256   64   87  20  276
1943  .9  .82  .309 .406 .196 .349  154  585  181  238  36  621
1945  .9  .82  .321 .396 .162 .354  142  540  173  214  28  568
1946  .9  .82  .272 .341 .251 .319  154  585  159  200  40  625
1947  .9  .82  .292 .375 .146 .321  154  585  171  219  25  610
1948  .9  .82  .327 .407 .104 .349  154  585  191  238  20  605
================================================================
               .300 .384 .426 .336 1016 3861 1159 1483 206 4067

Dandridge’s MLE OPS+
      Mle  mle     nl   nl   mle  mle  mle
      obp  slg    obp  slg   obp+ slg+ ops+
--------------------------------------------
1940 .311 .366   .333 .391     94   94   87
1941 .344 .405   .333 .375    103  108  111
1942 .304 .339   .329 .356     93   95   88
1943 .349 .406   .330 .360    106  113  118
1945 .354 .396   .339 .377    105  105  110
1946 .319 .341   .335 .368     95   93   88
1947 .321 .375   .345 .407     93   92   85
1948 .349 .407   .341 .398    103  102  105
============================================
     .336 .384   .381 .337    100  101  100

Dandridge appears to suffer a lot by comparison to his reputation. He doesn’t have much power at all after conversion to an NL environment, nor does he walk enough to provide Stan-Hack type value. These seasons represent almost the entirety of his peak, encompassing ages 26 to 34 (except for 1944).

I could also have botched my calculations, in which case, please let me know where my errors are!!!!

I'd love some feedback from the group on two issues about MxL translations.

1) I've pegged them generally at the .90/.82 conversion rate to maintain consistency with the NgL rates. Is this appropriate as a general strategy?

2) I'm rather convinced that the 1940-1941 Mexican League was very close to major league quality. 30-40 top NgL players were in the league in those seasons, and the league contained six teams each year. So six NgL players per team, plus the best Mexico could offer, plus players from around the Carribean. Shooting from the hip, I wouldn't wonder if this league was a 1.0/1.0 league or at worst a .97/.94 league.

3) 1946-1947 was not quite as good as 1940-1941, but better than surrounding seasons.

Anyone got any thoughts here on this? If true, my hypotheses could have a somewhat dramatic effect on the MLEs of Trouppe, Wright, Brown, Dandridge, Bell, and others, and I don't want to do something that would make it appear that I'm puffing up their stats if everyone thinks it's not a good idea.

Also, KJOK, I second your idea for an "other leagues" thread. We really ought to have all that information brought together in a convenient place on the boards. And there's tons of issues to sort out like the ones I've mentioned above.

bump

Here are all the .300 hitters on the Minneapolis Millers during Dandridge's run. Sorry about the tabs. 1951:

1951 American Association, Manager: Tommy Heath, Record: 77-75, 5th Place
PlayerGamesB.A.HRRBIPosition
Willie Mays35.477830cf
Artie Wilson17.390213ss-2b-3b-of
Harv Gentry10.36403of
Hank Thompson14.340713lf
Ray Dandridge107.3248613b
Ray Katt117.3081157c
Neill Sheridan75.306950of
Pete Milne131.300968of

1952 American Association, Manager: Chick Genovese, Record: 1952, OK all the .290 hitterrs:

79-75, 3rd Place
PlayerGamesB.A.HRRBIPosition
Clint Hartung105.3342793of
Bill Howerton67.3072461of
Roy Broome142.3051799of
Ray Katt123.3041568c
Dom Dallessandro*115.2957301b-of
Jake Early63.295832c
Bob Lennon50.295832of
Daryl Spencer142.2942780ss
Ray Dandridge145.29110683b

1949
1949 American Association, Manager: Tommy Heath, Record: 74-78, 4th Place
PlayerGamesB.A.HRRBIPosition
Ray Dandridge99.3626642b-3b
Jack Maguire122.34812713b-ss-of
Roy Hughes115.3129572b-3b-of-ss
Don Mueller28.311212rf
Sal Yvars84.307857c

1950 American Association, Manager: Tommy Heath, Record: 90-64, 1st Place
PlayerGamesB.A.HRRBIPosition
Bob Lennon16.36406rf
John “Spider” Jorgenson64.3308473b-ss-of
Bert Haas141.318241061b-rf
Pete Milne54.312327lf
Ray Dandridge150.31111802b-3b

All of this is cc. directly from Stew Thornley's web site. Thanks, Stew.

Also Dave Barnhill with Millers 1950. Note that in 1949 he was 7-10, 5.25 but then adjusted (?).

Dave Barnhill271401133.60

Barnhill in 1951:

Dave Barnhill33105724.46

Above for 1950 he was 11-3 in 27 G and 140 IP. Here he is 7-2 in 33 G and 105 IP. Sorry again about the tabs.

Best MLB comp for Ray: in terms of position, reputation, career length, and strengths (avg, power, walks, speed, defense): I nominate Pie Traynor.

Between Traynor and Kell, is this the kiss of death?

I thought I'd love Ray Dandridge but, so far, the meat isn't there. I'm open to being convinced, but right now he ranks at minimum behind Paige, Moore, Oms, Mackey, Redding, Bell, Mendez, and W. Brown; roughly equal to (but probably slightly behind) Taylor and Scales. (I haven't evaluated Wright, Easter, or Troupe yet.) That cracks my top 75, but not my top 50.

Here are Ray Dandridge's full AA stats. They pretty much confirm what his Mexican League numbers tell us, while also being important for considering whether he gets minor league credit.

Yr G AB H D T HR R RBI BB K SB AVE OBA SLG
49 99 398 144 22 5 6 60 64 17 17 4 362 388 487
50 150 627 195 24 1 11 106 80 41 26 1 311 353 405
51 107 423 137 24 1 8 59 61 29 18 1 324 367 442
52 145 618 180 27 1 10 86 68 32 36 3 291 326 387

League totals for the AA:

Yr AB H D T HR BB AVE OBA SLG
49 41079 11387 1805 421 920 5532 277 363 409
50 40119 10599 1760 391 884 4759 264 342 394
51 40441 11035 1869 407 922 5015 273 353 408
52 41254 11192 1851 444 946 4852 271 348 406

The Minneapolis park, as has been noted before, was a fairly extreme hitter's park throughout its history. The Millers led the league in runs scored from 1949-51, and in home runs from 1949-52 (at times by a lot--in 1949 they hit 202 home runs, while the runner-up had 121). During these years they finished 4th, 1st, 5th, and 4th. I'd say this is indicative of an extreme hitters' park, especially for home runs.

Put it all together, and Dandridge could well have been an average (or worse) AAA hitter at this time.

Apologies for the formatting--I don't know what the problem is.

Testing...
11111
88888


11111
88888


11111
88888

Multiple Space Test
1 1 1
88 88 88

1 1 1
88 88 88


 1  1  1
88 88 88


_1 .1 _1
88 88 88


Gary, where do you get your excellent minor league data from? Do you have a particular encyclopedia or several?

Apologies for the testing, GaryA, I was just trying to see how you got such a nice fixed-width font without it getting squished.

View->Source is telling me that you used the "code" tag? Anyhow, that looks great for numbers but all multiple spaces get converted to a single space.

The "pre" tag is what it normally used here and it does preserve multiple spaces (though no tabs) but the font gets shrunk and the fives look like sixes.

Anyhow, a mini-pet-peeve of mine is why we can't seem to get the nice fixed width font generated by the "code" tag with the multiple space preservation of the "pre" tag. If anyone out there knows, let me know.

Funny, I thought I had used "code" before and spacing wasn't a problem.

Doc--I have Marshall Wright's many compilations of minor league stats available (though pricey) from McFarland. He's got books on the American Association, International League, Southern Association, and Texas League (also the pre-major-league amateur National Association of the 1850s and 60s, which I recommend highly).

I also have Carlos Bauer's latest book of PCL stats, but haven't been able to find it since I moved (it's around somewhere).

Gary,

Thanks! I looked on amazon at the Marshall Wright books, but they only appear to go through the early 1950s. Are there any sources for post-1952ish AA, IL, PCL stats?

Thanks again.

Ray Dandridge's Cuban League record:

Season    Team       Pos   AB    R    H   2B   3B   HR  RBI  SB   Avg
1937-38   Almendares  3B  211   26   63    5    1    0   27  11  .299
1938-39   Cuba        SS  182   18   58    5    2    1   30   3  .319
1939-40   Cienfuegos  SS  116   15   36    4    1    1   11   1  .310
1940-41   Cienfuegos  SS  158   16   29    2    1    0   15   4  .184
1945-46   Marianao    SS  173   24   55    5    2    0   13   2  .318
1946-47   Marianao          4    0    1    1    0    0    0   0  .250
1946-47F  Oriente     3B   93   15   29   --   --   --   --  --  .312
1947-48PF Cuba             66   11   14    2    0    0    3   1  .212
1949-50   Marianao    3B  318   40   84   15    3    2   28  10  .264
1950-51   Marianao    3B  278   38   83   12    1    2   24   3  .299
1951-52   Marianao    3B  224   30   64    8    2    3   16   2  .286
1952-53   Marianao    2B  305   36   85   12    1    1   21   3  .279

Notes:
1937-38 - Led league in stolen bases
1946-47F – An alternative league (Liga de la Federación) supported by American organized baseball.
1947-48PF – An alternative league (Players Federation). (Beginning in the 1947-48 season, American organized baseball gained control over the regular Cuban League; the Players Federation was organized in protest and included most of the long-time stars.)
1949-50 – Led league in at bats. Selected as third baseman on season’s all-star team.
1952-53 – Led league in at bats.

Ouch. For those of you keeping score, that's a career CWL BA of .282.

With about 2,130 AB (and few walks), I figure that is about the equivalent of 3.3 ML seasons. So his numbers per 154 games look like about:

3 HR - 57 RBI - .282
182 H
22 2B, 4 3B, and with the 3 HR, 153 singles
12 SB

Keep in mind these are not MLEs, these are just projecting his record in Cuba on a per 154 games basis (guesstimating on the 154 by ABs since I don't have actual games).

Not good. The fact that he played SS for 4 years helps.

Compared to his AAA stats per 154:
11 HR - 96 - .319

Much better, in a hitter's park, possibly what might be called an extreme hitter's park.

Of course we know he hit .350 in the Nel and .347 in Mexico. So even his .319 must be recognized as coming at an advanced age and the .282 in Cuba must reflect some fairly extreme pitcher's parks.

Average the four = .324, the mean = .333. A .300 MLE seems a tad low compared to some others....

A reminder that through 1948 he was playing in La Tropical, a huge park where the league leader in home runs typically had maybe 4 or 5 home runs (except for the years that Josh Gibson or Mule Suttles was playing). So Dandridge's lack of power probably didn't hurt him too much. And batting averages overall seem to have been down from what they were in the 1920s.

Comparing him to other players in the Cuban League, it looks like Dandridge was almost always one of the better players--often first or second on his team in runs or RBIs--but never really an MVP candidate.

Right now, without MLE's of course, I would have to say that he was better than Judy Johnson. But he isn't better than the Monroe, Lundy, Scales group that is populating the 30's for me.

However, I could be swayed with some impressive MLE's. This looks to be another clasic case of a player who is overrated by the HOF, great fielder and good contact/average hitter.

I thought the juxtaposition with Daryl Spencer in 1950 was interesting, I mean Spencer at least was one name on those Minneapolis teams that we're familiar with. Even taking Dandridge's great year from 1949 vs. Spencer's year in 1950.

Dandridge 1949 6-64-.362 2B-3B
Spencer 1950 27-80-.294 SS

Granting that Dandridge was age 36 by then, but Spencer was just 21 on hardly a mature player. In 1953 his K/BB was almost 2, e.g., and by 1958 and again in 1960 it was <1. Don't know about 1950. But Dandridge only played 99 games that year, which probably helped his productivity when he was in the lineup.

Spencer ended up playinjg 1098 ML games, mostly with the Giants with the following line: .244/.327/.380/88. In his best years he only hit 101 and 102 OPS+.

Apparently he was a terrible fielder initially, too, fielding .927 in 118 games with the Giants in 1953, whereupon they moved him to 2B in 1954. He was about equally bad and they moved him back to SS where he remained a regular 2 years, then another year at 2B, then another year at SS, then played out the string as a UIF mostly at 3B. Apparently he had a very strong arm, but mediocre range and his fielding percentages were not good.

Does this tell us anything about Dandridge? Maybe not, except that Daryl Spencer seemed to get more out of that bandbox park than Ray was able to do, by way of XBH anyway.

Re. Dandridge in Mexico, I don't think the parks would take away 2B and 3B, would they? Yet he had precious few of those, either. I'm not sure we aren't looking at a Nellie Fox comp.

RAY DANDRIDGE MLEs a verrrry rough draft

Hey, everyone, using the same processes that I used on several other players, I've done Ray Dandridge. And it's not at all pretty. Dandridge walked at about 1/2 the rate of the leagues he was in (that we have such info for), and he had almost no power to offset this problem.

Once again, I've used all the available information that I can, and in Dandridge's case this comes in handy because he played a bunch of partial seasons. I think rather than looking at Nellie Fox or someone like that as a comp, we're looking at Don Mueller, the very fellow who blocked Dandridge's path to the big leagues.

Now my math and my methods could be all wrong here---and my MLEs way off base---but even if that were true, it would still be hard to say much positive for a guy sucking up outs with little pop to show for it.

So with trepidation....

year lg age pos avg  obp  slg    g   ab    h   tb  bb   pa OPS+    WS
----------------------------------------------------------------------
1933 nl  19 3b .190 .209 .281   42  161   31   45   4  165   39    .6
1934 nl  20 3b .392 .423 .425  135  514  202  219  27  541  128  25.7
1935 nl  21 3b .295 .322 .371  136  515  152  191  20  535   86  14.0
1936 nl  22 3b .271 .298 .303   95  363   98  110  14  377   62   6.1
1937 nl  23 3b .303 .334 .366  122  462  140  169  21  484   90  13.1
1938 nl  24 3b .306 .338 .382   88  335  102  128  16  351   97   1.2
1939 nl  25 3b .293 .325 .396   85  321   94  127  15  337   93   9.7
1940 nl  26 3b .208 .230 .263   79  302   63  79    9  311   36   1.5
1941 nl  27 3b .301 .338 .399  122  463  139  185  26  489  107  15.1
1942 nl  28 3b .196 .235 .246  146  555  109  136  28  583   41   2.1
1943 nl  29 3b .309 .349 .406  154  585  181  238  36  621  118   2.5
1944 nl  30 3b .325 .361 .420  154  585  190  246  33  618  119  22.6
1945 nl  31 3b .310 .344 .384  130  494  153  190  26  520  102  15.7
1946 nl  32 3b .287 .330 .362  129  489  140  177  32  520   96  13.2
1947 nl  33 3b .287 .330 .362  129  489  140  177  32  520   84  13.2
1948 nl  34 3b .308 .335 .368  127  484  149  178  20  504   90  14.0
1949 nl  35 3b .268 .286 .332   64  240   64   80   6  246   65   4.4
======================================================================
TOTAL          .292 .325 .364 1937 7357 2147 2676 365 7721   90 201.7
              
----------------------------------------------------------------------
1950 nl  36 3b .284 .299 .345  145   549  156  189  11   561 69  11.8
1951 nl  37 3b .277 .294 .348  111   423  117  147  10   433 72   8.8
1952 nl  38 3b .244 .260 .284  150   568  139  161  12   580 51   6.0
1953 nl  39 3b .218 .248 .238  106   403   88   96  16   419 28   2.0
1954 nl  40 3b .246 .280 .260  123   468  115  122  22   490 43   4.8
1955 nl  41 3b .306 .344 .346  123   468  143  162  27   496 84  12.8
======================================================================
               .284 .315 .347 2694 10236 2905 3553 464 10700 81 247.9 

Notice the big dip in his production ad his number of games in the late 1930s? I wonder if Dandridge was injured in some significant way?

Also, note that the s-fWS projections above assume an average fielder, and Dandridge was reputed to be well above average at third. I also assumed he played 3B each year, which is a broad assumption. Various sources report him playing at every throwing infield position, so he may have had a bit more fielding value yet.

I think the Judy Johnson comment above is quite apt, and I'll be especially curious to see what Gary and Gadfly say.

By the by, I'm now incorporating Figueredo and Hyning as source material which should help matters a bit, especially for Cuban league play where I've been able to figure the league AVG and SLG.

Dandridge's slightly anomalous 1934 season may be partly due to the Newark Dodgers' ballpark. I don't have figures for it, but I do recall that scores there were very high, and the Dodgers had several hitters with impressive numbers.

I don't know if this is the same park the Newark Eagles would use later--though I think it also had a reputation as a hitters' park.

I do think we'd be on pretty safe ground considering Dandridge an above-average third baseman, with some time at shortstop. Cisneros lists him as a shortstop for his whole Mexican career, plus he's listed at short for four seasons in Cuba and one season (1944) in the Negro Leagues.

Dr. Chaleeko,

Two comments and two questions.

First, these projections look like they're in the ballpark for Dandridge.

Second, there seems to be some problem with the display of the seasonal win shares column: the seasonal totals don't add up to the career total, and some seasons are obviously too small by a considerable amount: see esp. 1943.

The questions:

1) what data are you using for Dandridge's NeL seasons in the late 1930s?

2) what conversion factors are you using for Dandridge's Minneapolis years, and what park factors?

My guess about Dandridge's low numbers in the late 1930s is that the projections are a bit low for those years. They could be affected by the offense-level factors you are using. I also strongly suspect that extra-base hit data is underreported for the Newark team in at least some of these years.

I'd say it's probable that Dandridge's real productivity during these seasons was pretty close to what he accomplished in Mexico: 90-115 OPS+ seasons.

Anyway, with this range of offensive production, he'd have to have been close to Ozzie Smith levels of fielding talent to have a serious HoM case. Brooks Robinson had a 104 career OPS+, with five seasons above 120. I think Dandridge's batting projections would have to rise up to that level for his case for the HoM to get traction. Pie Traynor's at 107, but with a lower peak and shorter career than Robinson, and his case hasn't gone much of anywhere.

Chris Cobb,

You're right, something funky did happen to the display and the WS totals got all funked up. I think I know the problem, PEBKAC (problem exists between keyboard and chair...I used a find/replace to pull the "0." off of the AVG and SLG numbers, and it ate a few numbers form the WS too. Doh!) This issue should only be affecting the WS, not the MLE numbers.

So here's what the yearly WS should have said:

YEAR   WS
1933  0.7
1934 26.0
1935 14.3
1936  6.3
1937 13.4
1938 10.4
1939  9.9
1940  1.7
1941 15.3
1942  2.5
1943 20.8
1944 23.0
1945 16.0
1946 13.5
1947 13.5
1948 14.2
1949  4.5
=========
TOT 206.1


1950 12.1
1951  9.1
1952  6.3
1953  2.2
1954  5.1
1955 13.1
=========
TOT 254.0

For the late 1930s seasons, I'm using a combo of the data from the scans that KJOK has generously provided on the Yahoo site along with Dandridge's Cuban numbers. Each converted at .90/.82 (though the Cuban numbers could perhaps be converted at .94/.92 as you have elsewhere suggested.) The Cuban numbers are normalized to the CWL's actual leaguewide AVG and SLG numbers (compiled from Figueredo), but the NNL numbers are not normalized to the NNL league average because I don't have data for those seasons. I've just normalized it to the NL's numbers, which may cause some distortion, but I hope not too much. Part of my reasoning in using the Cuban totals is that having full league totals ought to help ameliorate some of the underreporting of xBH in the NNL records. Though clearly it won't make that problem disappear by itself.

As for park adjustments, I should have mentioned this in my post about the MLEs, but I've made no park adjustments at all because I don't have any specific numbers to go on. Those AA seasons, as well as the Cuban years, are normalized to the AA's (or CWL's) leaguewide AVG, SLG, and BB rates and converted at the usual .90/.82. I think if those AA seasons were park adjusted, they'd look really bad....

Hope that answers your questions, let me know how you think I should deal with those late 1930s seasons!

Dr. C,

A conversion factor of .90 is definitely too low for the Cuban League of the late 1930s. By then, the league was again bringing in many top American stars like Josh Gibson, Ray Brown, and Willie Wells, and the quality of play was back near its mid-1920s peak. I'd suggest using a conversion factor of .97 (though that still won't be enough for Dandridge to be a strong candidate).

Thanks, Brent, I'll take a look back at it and see how it changes his career totals and report back. His cuban play is extensive, so it may well have an effect.

Also, does anyone have leaguewide AVG and SLG numbers for the NNL from 1932-1942?

Thanks!

RAY DANDRIDGE MLEs VERSION 2.0

A few small repairs.

1) I finally got my 1945 Mexican League leaguewide averages inputted and computed, so Dandridge's 1945 MxL play is now more accurately normalized to his league environment.

2) On Brent's suggestion, I have upped the Cuban discount rate for seasons prior to 1949. I'm using .94/.92, the general rate suggested by Chris Cobb over on the Oms thread.

Here's the numbers.

year lg age pos avg  obp  slg    g   ab    h   tb  bb   pa OPS+  WS
--------------------------------------------------------------------
1933 nl  19 3b .190 .209 .281   42  161   31   45   4  165  39   0.7
1934 nl  20 3b .392 .423 .425  135  514  202  219  27  541 128  26.0
1935 nl  21 3b .295 .322 .371  136  515  152  191  20  535  86  14.3
1936 nl  22 3b .271 .298 .303   95  363   98  110  14  377  62   6.3
1937 nl  23 3b .303 .334 .366  122  462  140  169  21  484  90  13.4
1938 nl  24 3b .315 .347 .414   88  335  105  139  17  351 108  12.2
1939 nl  25 3b .306 .338 .444   85  321   98  143  16  337 109  12.3
1940 nl  26 3b .214 .237 .285   79  302   65   86   9  311  43   2.5
1941 nl  27 3b .301 .338 .399  122  463  139  185  26  489 107  15.3
1942 nl  28 3b .196 .235 .246  146  555  109  136  28  583  41   2.5
1943 nl  29 3b .309 .349 .406  154  585  181  238  36  621 118  20.8
1944 nl  30 3b .325 .361 .420  154  585  190  246  33  618 119  23.0
1945 nl  31 3b .305 .338 .385  130  494  151  190  25  519 101  15.6
1946 nl  32 3b .292 .336 .382  129  489  143  187  32  521 103  15.0
1947 nl  33 3b .267 .296 .348  109  415  111  144  17  431  70   8.9
1948 nl  34 3b .308 .335 .368  127  484  149  178  20  504  90  14.2
1949 nl  35 3b .268 .286 .332   64  240   64   80   6  246  65   4.5
====================================================================
total          .292 .325 .369 1917 7282 2127 2685 351 7634  91 207.5

Great stuff, Eric.

Me thinks Ray will have a tough time getting votes for the HoM.

I am not sure that it will matter much but do thsoe WS give him A+ fielding or just the average fielding that SFWS gives?

I presume that those seasons are only for 154 game seasons?

these are sfws that assume average fielding. i don't have the WS fielding norms in front of me, but I'd guesstimate that the difference between A+ and C is probably on the order of two or three WS per season. So Dandridge could be a 225-235 WS guy.

season is 154.

some stuff I collected on Ray Danridge
from baseball library .com
Ray Dandridge was a masterful third baseman, a stylist who could make all the plays. He was smooth and relaxed, with soft hands, a strong arm, and the versatility to excel at any infield position. "People would pay their way in to the game just to see him field," claimed Monte Irvin. Roy Campanella said, "I never saw anyone better as a fielder." Hoyt Wilhelm, who played against Dandridge in Cuba and with him in Minneapolis (American Association), asserted, "No matter how the ball was hit, he always made the throw so that he just did get the man at first." Others observed that a train could go through Dandridge's bowlegs, but that a baseball never did.
Dandridge started his pro career with the 1933 Detroit Stars and moved to the Negro National League's Newark Eagles, for whom he starred throughout the remainder of the 1930s. A spray hitter with good bat control, he seldom struck out, and skillfully executed the hit-and-run. In 1935, he hit .368. Looking for more money in 1939, he opted to play in Latin America. He went to Mexico in 1940, and spent most of the decade there. When he came back for a year in Newark in 1944, he batted .370, leading the NNL in hits, runs, and total bases. In 1945 he set a Mexican League record for hitting safely in the most consecutive games and managed his team to a pennant. In nine Mexican League seasons, he compiled a .343 average. Following the 1948 season, he returned to the States as player-manager of the New York Cubans.

During his time in the NNL, Dandridge registered a lifetime .355 average, and played in three East-West all-star games, hitting .545. He played winter ball in Mexico, Venezuela, Puerto Rico, and Cuba; in 11 seasons of Cuban Winter League action, he batted .282.

Soon after Jackie Robinson signed with the Brooklyn Dodgers, Bill Veeck contacted Dandridge about playing with the Cleveland Indians, but Dandridge refused to leave Mexico without a bonus. Later, in 1949, at age thirty-five, he was signed by the New York Giants and assigned to their Triple-A farm club at Minneapolis. He batted .363 his first year there, and won the league's MVP award in 1950, when he led Minneapolis to the league championship. Despite his achievements, the Giants would not promote him to the parent club.

While at Minneapolis, Dandridge provided advice and assistance to a young Willie Mays, who never forgot the help or the man. Returning to Cooperstown for Dandridge's induction into the Hall of Fame (he was elected by the Committee on Baseball Veterans in 1987), Mays stated, "Ray Dandridge helped me tremendously when I came through Minneapolis. Sometimes you just can't overlook those things. Ray was a part of me when I was coming along." (JR)
from HALL OF FAME BIOS
Ray Dandridge
A contact hitter who consistently batted .300 or better,
Dandridge played seven years with Newark of the Negro National League
also spent eight summers in the Mexican League
11 winters in the Cuban Winter League.
played 4 years with the Giants Triple-A club batting 318 overall and winning the AA mvp in 1950
when he was 36 Ray Dandridge had 362 batting avg. and led all AA. 3rd basemen with a .981 fielding%
HALL OF FAME PLAQUE reads
RAYMOND EMMETT DANDRIDGE
NEGRO AND MEXICAN LEAGUES
1933-1948
FLASHY BUT SMOOTH THIRD BASEMAN. DEFENSIVELY,
A BRILLIANT FIELDER WITH POWERFUL ARM.
OFFENSIVELY, A SPRAY HITTER WITH OUTSTANDING
BAT CONTROL. PLAYED FOR DETROIT STARS, NEWARK
DODGERS, NEWARK EAGLES AND NEW YORK CUBANS
IN NEGRO LEAGUES AND FOR VERACRUZ AND MEXICO
CITY IN MEXICAN LEAGUES. AMERICAN ASSOCIATION
MVP IN 1950 WITH .311, 11 HOME RUNS AND
80 RBI'S PLAYING FOR MINNEAPOLIS MILLERS.

A question on Dr. Chaleeko's conversions. I'm struck by the low totals for Dandridge's games played during the 1930s. I assume that these are based on the difference between his games played as compiled by Holway and the team totals.

My question is, are these differences real, or might they be an artifact of the way the data were compiled? That is, my understanding is that Holway compiled his statistics from box scores. Did he miss some games, and if so, could it make a full time player appear to be half time?

If Dandridge actually was missing so many games, is there a story? Injuries?

I'm not trying to push Dandridge's candidacy, but these estimates of his production from ages 21 to 28 seem awfully low in view of his reputation.

I hate to argue with the numbers based on not that much, but I just have a feeling we may be underrating Dandridge. I want to quote James in the NBHA:

"Really a shortstop in terms of his ability, but played third base in some of his best years because he was a teammate of Devil Wells...Fast, an amazing third basemen, and a .350-.370 hitte, he was signed by the Giants when he was 35 years old (nearly 36) and was the best player in the American Association the second half of 1949 (hitting .363) and all of 1950 (when he won the league's MVP award), but the Giants (who needed a third baseman) wouldn't bring him up. Durable, consistent; did everything exceptionally well except that he wasn't a power hitter. Is in the Hall of Fame, and deserves to be."

I know that not all of that is borne out by the statistical record, but if Dandridge was really that good of a fielder (and no one has argued that he wasn't), then I think he's not getting enough credit for it. I don't know why you wouldn't count at least some of his minor league years.

And quite honestly, the MLEs look somewhat suspect to me, specifically 1940 and 1942. Those are just insanely bad years, right in the middle of his peak. They seem to be much harsher than you'd expect from the Mexican League translations up in Post #12.

I know James isn't a Negro Leagues expert, but his opinion isn't worthless, either. (Even though he ranks Judy Johnson #2, his comments are a lot more tempered.)

I'm not settled on Dandridge yet, and I do think he's behind Leach and Elliott among 3Bmen. But there are enough things about the record that don't make sense to me that I think I'll have him higher than everybody else around here.

Devin raises a good point, and others have probably raised it as well.

One of the difficulties of MLEs versus actual ML playing records is that MLEs smooth out the peaks. But they also (presumably) smooth out the valleys. Whereas a Mickey Vernon can have a really good year and then a not-good year, back to back, we don't generally see this in the MLEs. As has been said many times, the peaks and valleys wash out, and the longer term player is what comes into focus. For peak/prime voters, of course, this is not quite an answer.

All of that aside, what then when MLEs do in fact show big ups and downs of the kind that occur in the real world. Does it stand to reason that they reflect the real record or some artifact of the conversion? Some of each, probably, but like peaks and valleys washing out, this too is not quite an answer.

Devin's comments about Dandridge reminded me of Fred Dunlap. Dunlap was in his NL years a terribly consistent player with (raw/unadjusted) WS of 17-15-14-16-x-14-17 from 1880-1886. The x, of course, is his UA year of 38 WS.

So, was Dunlap a different player in 1884? Of course not, he played in a different environment. I chose to discount his record due to that fact by 65%. Some say discounting the AA 25-35% that same year is way too much, but if you discount the UA by 65% you get about 14 WS--the "same" as before and after, a common sense answer that in turn seems to justify the discount as about right.

So with Dandridge--was he the same player in 1940 and 1942 as he was in 1939, 1941 and 1943? Or was he a different player for some reason? If he was the same player, then the problem is with the conversion out of a different environment, no?

I think Chris, Doc and others are doing great work on the conversions, but nobody said it was easy, and I think the conversions are all open to debate and further adjustment, as Brent just said on the ballot thread. And I'm sure Chris and Doc would not disagree.

So hopefully we'll all keep an open mind on Dandridge and all of the NeLers, especially those who mixed in some Cuban and Mexican and Puerto Rican and ManDak and other play. Hopefully as everything else gets a little easier (as we get to where we are able to consider ML records apples to apples) there will be more study of these leagues and of the conversion rates.

One of the difficulties of MLEs versus actual ML playing records is that MLEs smooth out the peaks. But they also (presumably) smooth out the valleys.

The MLEs that I do for position players are smoothed, because they use regression. Dr. Chaleeko's MLEs for position players are not regressed, so the stark peaks and valleys that appear with short-season data are reproduced fully in the projections.

My projections for Dandridge would look somewhat different at the seasonal level than the good Doctor's do because of the regression. However, my career projections would probably be pretty similar. I haven't done full projections for Dandridge, but I've done some spot-checking to see if my conversion would be similar, overall, to Dr. Chaleeko's. And the spot-checking suggests that they would be.

It does seem proper to project Dandridge as an A-level fielder, though I don't see that additional credit as enough to put him over the top.

Playing time projections always merit careful scrutiny,too, because that's where the person doing the projections has to exercise the most independent judgment.

Raymond Dandridge played two seasons in Puerto Rico, both with Santurce. In 1941-42 he batted .288 and in 1953-54 his average was .241.

Sorry I haven't been able to respond sooner, I've been away from a computer for a couple days.

Chris is correct, I do not use regression (I don't know how to!!!), so sample size can effect my MLEs. On the other hand, I'm also incorporating winter league data which helps to smoothe things out by increasing the sample.

To offer a little more info on a couple places:
-Dandridge is shown as if he were an average fielder. The most that Dandridge could reasonably gain in any individual season by upping him to A++++ would be something on the order of two WS per 118 games (or 1000 innings). At 1900 games played, that would mean another 32 win shares for his career. I don't think it makes or breaks him, so I didn't recalculate.
-Playing time: I've approached this in a mostly hands-off kind of way, simply projecting each season to 154 based on the number of games played and team games. I do a little shortening at the early early and late late stages of a player's career, but that's it. That's why my projections often come in higher for PA and G than are realistic. I'm hewing close to the record because I'm afraid to manipulate the data too much.
-1940 and 1942 are, in fact, very odd looking. What happened? Let's take each one seperately.
In 1940, he was above-average in the Mexican League, where the league averages were .290/.420. He loses 10% in translation which roughly takes him back to average. Then in the CWL, he tanks, going .184/.220 in a .248/.309 league. AFter the conversion discount is applied, he's more Candice Bergen than Bill Bergen.... But there's a kicker here: he only played 67 of his team's 160 games. That screams INJURY to me.

Then in 1942, he essentially has the same kind of Mexican season as in 1940, but in Newark in the NNL he hits .165/.203. Putting them together and applying the discount you get a reall bad hitter. But here again there's a playing-time question. He appeared in 60 of his teams' 129 games. Another injury perhaps?

The mystery is that his surrounding Mexican seasons are typical of his Mexican sojourn, so these injuries seem to have not effected him in the long term. He still hit for average and the same level of power, and he still stole bases and legged triples. So I'm not seeing a back, wrist, or leg injury here. If there's an injury, perhaps it's an elbow or right shoulder? Or perhaps he had some kind of infection that sapped his strength temporarily.

Anyway, that's what the record tells me and why the translations show up as they do.

Someone on the 1959 ballot thread asked for a recalculation of Dandridge's WS with an A+ fielding rating. Here they are. Quick note: I assumed he was a depression-era 3B and adjusted his WS/1000 innings to match the depression-era levels.

year po OPS+  WS   WSA+3B
-------------------------
1933 3b  39   0.7   1.3
1934 3b 128  26.0  27.7
1935 3b  86  14.3  16.1
1936 3b  62   6.3   7.5
1937 3b  90  13.4  15.0
1938 3b 108  12.2  13.3
1939 3b 109  12.3  13.4
1940 3b  43   2.5   3.5
1941 3b 107  15.3  16.9
1942 3b  41   2.5   4.4
1943 3b 118  20.8  22.8
1944 3b 119  23.0  25.0
1945 3b 101  15.6  17.3
1946 3b 103  15.0  16.6
1947 3b  70   8.9  10.3
1948 3b  90  14.2  15.9
1949 3b  65   4.5   5.4
=========================
total    91 207.5 232.6

A little contextual information on Dandridge's NNL ballparks.

Looking in Green Cathedrals I checked on the dimensions of the parks Dandridge played in to see if any of them were unusually difficult hitters parks.
<pre>
|(DIMENSIONS | (FENCES
| IN FEET) | IN FEET)
YEAR TM PARK | LF CF RF | LF CF RF
---------------------------------------------
1933 DET Hamtrack Stdm 315 407
1933 NAS Sulphur Dell 334 421 262 16 16 45
1933 NWK Meadowbrook O 300 380 300 12 12 12
1934 NWK Meadowbrook O
1935 NWK Meadowbrook O
1936 NWK Ruppert Stdm 305 410 305
1937 NWK Ruppert Stdm
1938 NWK Ruppert Stdm
1939 NWK Ruppert Stdm
1942 NWK Ruppert Stdm
1944 NWK Ruppert Stdm
1949 NYC ????????????

Without knowing what other parks in the league were like, these dimensions at least suggest that Dandridge wasn't playing in any super-extreme pitcher's paradises which might adversely effect his numbers.

I don't have much data on his Mexican parks, nor any on his winter league parks.

Re posting that park data with the proper formatting....

|(DIMENSIONS | (FENCES 
| IN FEET) | IN FEET)
YEAR TM PARK | LF CF RF | LF CF RF
---------------------------------------------
1933 DET Hamtrack Stdm 315 407 
1933 NAS Sulphur Dell 334 421 262 16 16 45
1933 NWK Meadowbrook O 300 380 300 12 12 12
1934 NWK Meadowbrook O
1935 NWK Meadowbrook O
1936 NWK Ruppert Stdm 305 410 305
1937 NWK Ruppert Stdm
1938 NWK Ruppert Stdm
1939 NWK Ruppert Stdm
1942 NWK Ruppert Stdm
1944 NWK Ruppert Stdm
1949 NYC ????????????

Ack! Very sorry. One last time, correctly, and with feeling.

                      |(DIMENSIONS | (FENCES 
                      |  IN FEET)  | IN FEET)
YEAR  TM PARK         | LF  CF  RF | LF CF RF
---------------------------------------------
1933 DET Hamtrack Stdm 315 407 
1933 NAS Sulphur Dell  334 421 262   16 16 45
1933 NWK Meadowbrook O 300 380 300   12 12 12
1934 NWK Meadowbrook O
1935 NWK Meadowbrook O
1936 NWK Ruppert Stdm  305 410 305
1937 NWK Ruppert Stdm
1938 NWK Ruppert Stdm
1939 NWK Ruppert Stdm
1942 NWK Ruppert Stdm
1944 NWK Ruppert Stdm
1949 NYC ????????????

Here's Ray Dandridge in the new system. Dandridge presents some peculiarities which causes me to wonder how well any translation attempt will characterize him.

1) This translation shows Ray with a top-notch glove, average 3B is 1 WS/38 games, I have him at 1 WS/25 games.

2) I translated at 4.1 PA/G instead of his actual 4.41 PA/G.

3) I've excluded from this presentation his seasons before 1934 and after 1952, they are wretched and he was young or old enough not to be in the league with crummy numbers. Realistically, only his defense could have gotten him into the big leagues as a young player. In 1934 at 3B that may have been possible in a league where Ossie Bluege was a fresh memory...or even active.

4) His 1934 rookie season is a big one (for him), but I think it's illusory. I wonder if the stadium was changed in some way that year. It's just a major outlier year in that period of his career (or any).

5) No data for 1939 except for his CWL season, so I took that as my sample. I assumed it was around the same quality as the NgLs or the PCL as Chris, Gary A, and others, have reasoned.

6) His Minneapolis years are nice superficially, but in figuring some rough park factors, I've come up with these for Nicolette:
1.13
1.10
1.07
1.05

This is consistent with the 1940 Minneapolis pf I've figured for Estalella, 1.10.

These factors do damp him down somewhat.

7) Here's the discount structure I used:
NgL, CWL: .90
AA: .86
MxL: .80

8) There's a few seasons where he looks like he's played partial years. I don't know why, I don't have a clue, I'm just following the numbers on this one.

9) The increase from 232 to 279 is approximately a 20% increase from prior MLEing. For what it's worth, SFWS sees about three more WS than I do.

10) It's pretty clear that he would not have gotten 500+ PAs in his final season, but I thought before I did any shaping it might be wise to show it as-is to the group and let anyone make suggestions about what they think that year's PT should look like. 100 PA? 200? 300? 400?

11) Translated into the NL, 154 game sked.



For the purpose of comparison, I used the SBE to find comparable historical players:



It's not a real inspiring group, though to be fair I didn't look for relative averages.

Anyway, given the 1934 issue, the last-year issue, the comps, and all that, I feel OK saying that this round of MLE probably overstates him a little bit. I've got his defense as the same all along, which is my common practice, but at such a high rate, this could mean I'm missing his downswing and overstating his last few seasons. He's a smidge better than Judy Johnson on rate and rather better on bulk, but that's not saying much.

Nicollet (note the spelling) had a short porch to be sure. It is where Joe Hauser hit a bunch of HR. (Nicollet was another of those French priest/missionary/explorers. Nicollet is also a main north-south avenue through the heart of downtown Mpls. and on which the ballpark was located.)

Here's a list that checks Dandridge's win shares in relation to his PA and OPS+. It contains all players within 500 PA of Dandridge's estimated 9100 and who had an OPS+ of 110 or lower.

Player – PA – OPS+ (bws/fws/total)
Brett Butler – 9545 – 110 (239.4/57.5/295)
Pee Wee Reese – 9470 – 99 (203.9/111.3/314)
Willie Randolph – 9462 – 104 (213.3/97.4/312)
Alan Trammell – 9375 – 110 (225.1/92.6/318)
Bid McPhee – 9409 – 106 (206.8/98.7/305)
Tommy Corcoran – 9368 – 74 (95.1/118.8/214)
Jimmy Dykes – 9347 – 96 (166.8/80.0/245)
Red Schoendienst – 9222 – 93 (165.8/95.3/262)
Eddie Yost – 9175 – 109 (222.5/45.6/267)
Joe Carter – 9154 – 104 (198.4/41.0/240)
Tony Phillips – 9110 – 109 (211.2/56.7/268)
B. J. Surhoff – 9106 – 98 *active after 2001
Larry Bowa – 9103 – 71 (91.7/88.2/179)
Ray Dandridge – 9100 – 89 (190.3/88.8/279.1)
Joe Kuhel – 9095 – 104 (209.4/35.4/243)
Tommy Leach – 9051 – 109 (232.5/96.1/328)
Marquis Grissom – 8959 – 92 *active after 2001
Tim Wallach – 8908 – 102 (164.0/82.2/248)
Fred Tenney – 8807 – 109 (210.5/37.7/249)
Tony Fernandez – 8793 – 101 (179.1/99.4/280)
Ruben Sierra – 8782 – 105 *active after 2001
Charlie Grimm – 8745 – 95 (158.0/38.7/198)
Dick Bartell – 8743 – 96 (156.6/96.9/252)
Donie Bush – 8743 – 91 (154.8/78.0/232)
Kenny Lofton – 8715 – 107 *active after 2001
Todd Zeile – 8649 – 103 *active after 2001
Stuffy McInnis – 8623 – 105 (193.0/36.0/227)

From this list, it looks to me like Dandridge's batting win shares are probably overestimated by as much as 15%. Here's the batting group to which Dandridge seems most similar:

Jimmy Dykes – 9347 – 96 (166.8/80.0/245)
Red Schoendienst – 9222 – 93 (165.8/95.3/262)
Ray Dandridge – 9100 – 89 (190.3/88.8/279.1)
Tim Wallach – 8908 – 102 (164.0/82.2/248)
Charlie Grimm – 8745 – 95 (158.0/38.7/198)
Dick Bartell – 8743 – 96 (156.6/96.9/252)
Donie Bush – 8743 – 91 (154.8/78.0/232)

By OPS+, Dandridge is the worst hitter in this group, but he has the most win shares by a considerable amount. It seems very unlikely to me that differences in offensive conditions would make Dandridge's hitting so much more valuable contextually than any of these players. By these comps, I would estimate Dandridge's BWS as being in the 158 to 165 range. 190 seems too out of line with this group to be plausible. It doesn't matter in Dandridge's case, since we're not going to elect him, but I wonder if the same comp problem arises for Bus Clarkson.

The ARod comparison suggested that might be the case, but there were too many variables affecting that single comparison to make it meaningful. It would be a good idea to generate a similar list for Clarkson to see if there is a similar discrepancy. I won't have time tonight, but maybe someone else would be interested? If not, I'll do it over the weekend and see what I find.

Eric, do you see anything that would account for a discrepancy, either a justification for the difference or an error that would have created it?

Pee Wee Reese, I note, has a much higher rate of bws/PA than Tim Wallach, despite being similar offensively. He has 10 points of OPS+ on Dandridge, but his rate is still only slightly higher than Dandridge's is. We can't expect exact consistency in these matters from win shares, but Dandridge's totals are outliers in comparison to the outlier here, it looks like.

Pee Wee Reese, I note, has a much higher rate of bws/PA than Tim Wallach, despite being similar offensively.

Comparing the two, Reese's OPS was weighted toward OBP, whereas Wallach's was heavier on the SLG side of things. Also, Reese probably was helped by his base stealing, which was a negative for Wallach.

I agree that Eric's current WS formula seems to be overstating WS, given the other statistics shown in the players' translated batting lines.

Wait, I think I might have an idea on this. The A-Rod example kind of made me think of it.

I'm currently calculating this as if the team had a typical number of runs scored for the league as a unit. This means that the player in question begins squeezing the other players away from average and toward replacement. The phenomenon would be worse with better players.

Maybe what I should be doing is instead assuming that the other seven slots in the lineup are average hitters and that the pitcher is average for pitchers, then tack the player onto that group (also picking up any "missing" time within his own lineup slot) and figure WS.

I refigured Clarkson this way (with a couple of other small modifications as well that were made along the way since the last draft, and I get this:



Wow, that's a big difference! If you choose to further discount his TxL seasons, you'd probably be down close to the area of 330 WS, which is just 15 more than the original MLE projection.

Here's Dandridge by the same method:



He's not so good as Clarkson, so I suspect there's less de-puffery because he's a lot closer to average already. Still, a reduction of 5%. Looking back at Chris's list of most-comparables above. He's a little closer now, less of an outlier in that group, though admittedly still at the top end of it.

Ray Dandridge – 9100 – 89 (180/85/265)
Red Schoendienst – 9222 – 93 (166/95/262)
Tim Wallach – 8908 – 102 (164/82/248)
Jimmy Dykes – 9347 – 96 (167/80/245)
Dick Bartell – 8743 – 96 (157/97/252)
Donie Bush – 8743 – 91 (155/78/232)
Charlie Grimm – 8745 – 95 (158/39/198)

As noted above, I think Dandridge is problematic to translate in general, so I'm not going to make any pronouncements here, though I might add a couple names to the list:

Ossie Bluege 7452 - 85 (114/70/183)
JIm Gilliam 8321 - 93 (178/68/247)
Maz 8979 - 84 (105/112/219)
Frank White 8467 - 85 (103/106/211)
Tony Taylor 8501 - 88 (132/64/198)

Obviously Dandridge is coming out better than these long-career guys with somewhat similar profiles. Could be that the MLE over estimates how long he could remain in the league as a glove-first guy.

Baby steps....

Wow, that's a big difference! If you choose to further discount his TxL seasons, you'd probably be down close to the area of 330 WS, which is just 15 more than the original MLE projection.

That's still damn impressive, Eric, and will keep him in the #1 spot on my ballot.

Please click through for my latest Ray Dandridge MLEs.

Some of his data is missing, and the MLEs will be updated as new stuff comes to light.

Currently looking at 70.6 WAR, but I’m not 100% confident. The MLE protocol I’m using returns that figure, however, you will see in the article that while the fielding seems solid, the batting value we have estimated may be a problematic. Always glad for other’s thoughts.

RAY DANDRIDGE: 3B
Year Age  Lg  Pos  G  PA    Bat  Bsr DP Fld  Pos  RAA   WAA Rep RAR  WAR  WAR162
=================================================================================
1934  20  NL  3B   35  150    0   0   0   1   1     2   0.2   5   7  0.7  0.7
1935  21  NL  3B   82  350    2   0   0   2   1     6   0.6  11  17  1.7  1.8
1936  22  NL  3B   92  400    1   0   0   3   2     6   0.6  12  18  1.8  1.9
1937  23  NL  3B  141  600    7   1   0   4   2    13   1.4  19  32  3.4  3.5
1938  24  NL  3B  142  610   15   1   0   4   2    22   2.3  19  41  4.3  4.5
1939  25  NL  3B  140  600   10   1   0   4   1    16   1.7  19  35  3.7  3.8
1940  26  NL  3B  133  570    9   1   0   4   1    15   1.6  18  33  3.5  3.7
1941  27  NL  3B  152  650   11   1   0   4   1    17   1.8  20  37  4.0  4.2
1942  28  NL  3B  144  610  - 5   1   0   4   0     0   0.0  19  19  2.2  2.3
1943  29  NL  3B  150  640   24   1   0   4   0    29   3.3  20  49  5.6  5.9
1944  30  NL  3B  138  590   16   1   0   4   0    20   2.2  18  38  4.2  4.4
1945  31  NL  3B  137  590   10   1   0   4   0    14   1.5  18  33  3.4  3.6
1946  32  NL  3B  146  620   23   1   0   4   0    27   3.1  19  47  5.3  5.6
1947  33  NL  3B  135  580    4   1   0   4   0     9   0.9  18  27  2.8  2.9
1948  34  NL  3B  131  560   22   1   0   4   0    26   2.7  17  44  4.6  4.9
1949  35  NL  3B   98  420    4   0   0   3   0     7   0.7  13  20  2.1  2.2
1950  36  NL  3B  105  450    1   0   0   3   0     4   0.4  14  18  1.8  1.9
1951  37  NL  3B  110  470  - 4   0   0   3   0   - 1  -0.1  15  14  1.5  1.6
1952  38  NL  3B  109  460  - 7   0   0   3   0   - 3  -0.4  14  11  1.2  1.3
1953  39  NL  3B   63  270  - 8   0   0   2   0   - 6  -0.6   8   2  0.2  0.3
---------------------------------------------------------------------------------
                 2383 10190 135  10   0  67  11   223  23.8 318 541 58.1 61.2

As you can see, Dandridge isn't a bad hitter at all, but he's a somewhat limited by his lack of walks and power. What's most surprising about him? His fielding: It's only very good, not fabulous. Much of Dandridge's reputation lay in his outstanding glove. To get to these 67 fielding runs, we have a combination of 159 fielding games in the Negro Leagues and three seasons of minor league play (1949-1951). Check out my blog for an explanation at greater length regarding how I used the minor league fielding numbers. In a nutshell, his Negro Leagues fielding comes up as about 2.5 Rfield per 154 games, which is good but not hosannahs. In the minors his numbers as a late-thirties infielder are quite good, more like 4 Rfield and change per 154 games. Obviously, he was a good fielder for a long time, but I haven't yet come across any evidence of his being an outstandingly good fielder.

When you tally it up, he comes to 33.4 WAR7/61.2 career adjusted to 162. In my own reckoning, he's not very similar to anyone. I think that's because the MLE tends to flatten peaks. The closest my own numbers come is probably Ron Cey: 38/54. From a raw BBREF WAR perspective, and now in terms of career WAA and WAR, Darrell Evans (no 1981 adjustment, no DRA) is at 24.3/58.8 v. Dandridge's 25.1/61.2 per 158. But with those adjustments I have Evans at 40 WAR7/69 career against Dandy's 33.4/61.2. (Sorry to flip back and forth between WAR7/career and WAA/WAR.) We might also think of him as Larry Gardner with a longer career.

So here's my bullet points on Dandridge. This is, of course, just my opinion based on my worldview of the HOM and my reckoning of every available player:
1) The fielding we do have, about 620 games, says he's not a world-class fielder.
2) That's problematic for him because we do have about 6300 PA of offensive data that says he's merely a good hitter.
3) If my reckoning is correct, we are well above and no worse than a little above the number of guys we need from the prewar period.
4) There are, in my opinion, at least four superior or equivalent 3Bs to him still on the board: Bell, Leach, Bando, and Cey.
5) Taylor would be a better selection than Dandridge overall...and as you'll see above, I don't support Taylor at this time.

[Cross-posting from 2020 election thread.]