you're reading...
Negro Leagues

Evaluating Negro Leagues Third Basemen

[NOTE: These MLEs were updated 12/7/17 to include an improved conversion of DRA to Rfield and a correction to the formula for Rrep which had been adding a run or two per annum.]

We make our way to the left side of Negro Leagues infields today to have a look at third baseman selected for honor by the Halls of Fame and/or Merit. Later in the next few months we’ll examine those candidates who haven’t achieved baseball immortality. We refer you to our Major League Equivalencies (MLEs) for Negro Leagues batters for all the gory details.

Ray Dandridge

[Updated 3/31/18 to reflect recent improvements in fielding runs and incorrect Quality of Play factors for several seasons.]

[Updated 4/3/18 with additional 1937 data.]

If you knew only two things about Hall of Famer Ray Dandridge, it’s probably that he had bowlegs and that he had one heckuva glove. These two attributes may well have a causal relationship. If doctors could have straightened Dandridge’s legs out, he’d have picked up half a foot in height, but you sometimes hear that bowleggedness helps athletes. Bowlegs lower the center of gravity, potentially creating more balance, stability, and agility. Combine that with the athleticism needed to play third base well at the highest level, and Ray Dandridge’s defensive reputation makes a lot of sense.

History also remembers Dandridge as a line-drive machine who hit .300 annually. That’s what his Hall of Fame plaque tells us. It’s a true story. Between his Negro Leagues and Mexican Leagues careers, he hit well over .300 and in the years for which we are running MLEs, he hit .339 (summer leagues only). You can rack up a lot of value hitting .339.

Still, the last time I’d worked with Dandridge’s numbers, at the Hall of Merit, I found his case less than compelling. He appeared to be a guy with an empty batting average, mediocre speed, and a great glove. It appears now that I may have been mistaken. Or am I?

Ray Dandridge
Negro Leagues Stats | Minor Leagues Stats | Bio
Career: 1933–1953
Destination: NL 1934–1951
Missing data: 1939, all winter league data
Honors: Baseball Hall of Fame, Mexican Baseball Hall of Fame

Year Age Lg Pos    PA Rbat Rbaser Rdp Rfield Rpos RAA   WAA Rrep  RAR   WAR
===========================================================================
1934  20 NL 3B    190    5    0    0     2     1    9   0.9    6   15   1.5
1935  21 NL 3B    300    0    1    0     4     1    6   0.6    9   15   1.5
1936  22 NL 3B    400    1    1    0     5     2    9   0.9   12   22   2.2
1937  23 NL 3B    500   18    1    0     6     2   27   2.8   16   43   4.5
1938  24 NL 3B    490   20    1    0     6     1   28   2.9   15   43   4.6
1939  25 NL 3B    460   13    1    0     6     1   20   2.1   14   35   3.7
1940  26 NL 3B    470   13    1    0     6     1   21   2.2   15   36   3.8
1941  27 NL 3B    610   13    2    0     8     1   23   2.5   19   42   4.6
1942  28 NL 3B    540    7    1    0     7     0   15   1.8   17   32   3.8
1943  29 NL 3B    590   19    1    0     7     0   28   3.2   18   46   5.3
1944  30 NL 3B    500   15    1    0     6     0   22   2.4   16   38   4.1
1945  31 NL 3B    540   16    1    0     7     0   24   2.5   17   41   4.3
1946  32 NL 3B    580   22    1    0     7     0   30   3.4   18   49   5.5
1947  33 NL 3B    570   15    1    0     7     0   23   2.4   18   41   4.2
1948  34 NL 3B    620   23    2    0     8     0   32   3.3   19   51   5.4
1949  35 NL 3B    350    5    0    0     4     0    9   1.0   11   20   2.1
1950  36 NL 3B    150    1    0    0     2     0    2   0.2    5    7   0.7
1951  37 NL 3B    100    0    0    0     1     0    1   0.1    3    4   0.4
---------------------------------------------------------------------------
                 7960  204   18    1    98    10  330  35.1  248  579  62.2

Hypothetical MLB Career Rankings (1871–1960)
PA: 89th
Rbat: 121st
WAA: t-39th
WAR: 40th

I’m of two minds about this MLE.

First off, Dandridge’s defensive reputation is neither confirmed nor disconfirmed by the currently available Negro Leagues statistical record. What we know: Dandridge picked up about 7 runs of defensive value per 162 games at third base in his Negro Leagues career, which so far spans 130-odd games of data and about 1100 innings. [On the other hand, he played a terrible shortstop.] The trouble for us is that most of his career took place in Mexico, where we have no fielding stats.

We do know, however, what his raw fielding stats looked like in the American Association and the Pacific Coast League from age 35 onward. Here are some fielding numbers from those seasons for Dandridge, the other third basemen on his teams, and his leagues’ third basemen at large:

YEAR  AGE     G@3B   FLD%    A/G    DP/G
===================================================
1949  35
Dandridge      83    .981    2.0    0.37     
Teammates      68    .921    1.8    0.26
League       1325    .947    1.8    0.20

1950  36
Dandridge     136    .978    1.9    0.20
Teammates      22    .953    1.9    0.18
League       1248    .942    1.8    0.21

1951  37
Dandridge     103    .961    2.0    0.23
Teammates      47    .946    1.5    0.17
League       1247    .945    1.9    0.20

1952  38
Dandridge     144    .975    2.0    0.25
Teammates      11    .880    1.6    0.00
League       1232    .942    1.8    0.21

1953  39
Dandridge      22    .921    1.5    0.18
Teammates     358*   .944    2.0    0.18
League       1477    .943    1.9    0.19
*includes stats for multiple teams

Basically, until he was 39, Dandridge was better than his teammates and his leagues. These were AAA leagues that included many soon and recent big leaguers such as Andy Carey, Don Hoak, Billy Klaus, and lots of others.

ERRORS: Dandridge was far less error prone than his peers. From age 35 to 38, “Scoops” made around 42 fewer errors than the league. Let’s assume that each error carries approximately the same run value as a single. Singles are worth about half a run to the offense, and outs are worth about -0.09, which means each error is worth about 0.59 runs because roughly speaking it turns an out into a single, and that means Dandridge might have saved 24.8 runs’ worth of errors.

ASSISTS: In Win Shares Bill James showed that third base putouts have very little meaning in determining the quality of a fielder’s play. Assists, however, represent how many balls a guy can actually get to and turn into outs. Dandridge racked up about 0.15 more assists per game than his peers between ages 35 and 38. That doesn’t sound like much, but it’s about 67 more assists than the league. Like with errors, these assists represent ground balls that didn’t go through for hits and each represents -0.59 runs for the offensive team, so chalk up another 39.5 runs to Ray above the league’s average hot corner man.

DOUBLE PLAYS: Finally, double plays turned will also represent skill. Each deuce is worth about -0.37 to the offense. Dandridge participated in 23 more double plays than the league average third baseman, which means 8.5 additional runs saved.

TOTAL: Totaling up all those contributions, his defense over four years was worth 72.8 runs above average, or 18 a year. We’re speaking in rough terms here. But even if you want to discount it 20% for occurring at AAA, that’s still 14.5 runs per season…as a late-career fielder.

CONTEXT: Dandridge probably didn’t face an unusual number of righty hitters, except in 1952. These are approximate innings thrown by lefties on his teams plus the percentage of MLB at-bats taken by righty hitters (MLB because minor league splits aren’t available,  at-bats because BBREF doesn’t list innings by lefties and righties in its splits):

  • 1949: 7.9% / 62%
  • 1950: 18.8% / 64%
  • 1951: 28.5% / 62%
  • 1952: 47.5% / 60%

In other words, in the first three of these seasons, Dandridge probably played behind considerably fewer lefties than his league mates did, which means that his very strong fielding rates occurred in all likelihood despite fewer opportunities.

When we take these latter-career performances into account, we must revise upward from his so-so DRA showing in just 136 Negro Leagues games. Brooks Robinson was about a +18 fielder in the same age range we’re talking about with Dandridge. I’m not about to claim that the latter was as good as the former. There’s not enough proof, and I’m using some general-application thinking and not getting down to his specific context. But how about we call Dandridge a 12-runs-per-season fielder across his career? Doesn’t seem like crazy talk, so that’s what I’ve done. It yields 144 fielding runs, which would be about eighth among all retired third basemen from 1920 through today. Want it to be 10 runs per 154, then it’s about 122 runs and ninth among that group. Dude could pick it.

Since originally posting this article, I’ve returned to the MLE to look once again at the fielding component. Since first posting Dandridge’s MLE, I’ve created an adjustment to fielding numbers that accounts for the difference in STDEV between the Negro Leagues Database’s DRA and the Baseball Gauge’s DRA. Applying that, I arrive at 8.11 runs per 154 games. The MLE above has been altered to reflect that, which reduces Dandridge’s WAR by 6 wins from the prior version of this MLE.

Now here’s the troubling part for me. Either Dandridge’s offensive game was much better than I’d have thought, or there’s something about his batting style that happened to play up in the leagues he played in. For starters, Dandrdige’s OPS+ in the Negro Leagues was 112, which ain’t amazing and boils down to an average or slightly below-average big-league hitter. However, his Mexican League seasons are good, and his first minor league season is solidly above average. He did slip below average in 1950 upon turning 36. What makes this difficult is that his Negro Leagues record is less comprehensive than his Mexican League record, and the latter all occurred in his peak and prime years. Meanwhile the minor league record starts during his baseball sunset. But taken together this all ends up looking like a guy who averaged a little more than 10 batting runs above average over his career and racked up nearly 200 batting runs. He did it by hitting .330 with 30+ doubles a year and by walking juuuust enough (around 6% of his plate appearances) to avoid becoming an OBP suck if he hit .280 or .290. In fact, Dandridge’s real-life batting average slipped below .300 only twice during the years we are giving him MLEs for, and just once below .285.

But see, I’ve got this thing about .300 hitters who don’t walk much. So I looked up the batting runs for retired players from 1920 to 2017 who hit .300, walked less than 700 times, and hit fewer than 100 homers in at least 5,000 trips to the plate. Let’s throw out Tris Speaker and George Sisler who did much of their damage in the 1910s. That leaves these guys:

  • Earle Combs: 242 batting runs in 6513 PA
  • Riggs Stephenson: 211 runs in 5134 PA
  • Tommy Holmes: 156 runs in 5563 PA
  • Edd Roush (age 27 on): 155 runs in 5336 PA
  • Sam Rice (age 30 on): 149 runs in 8734 PA
  • George Kell: 131 runs in 7529 PA
  • John Stone: 111 runs in 5008 PA
  • 13 other guys ranging from 13 runs (Lloyd Waner) to Pie Traynor (95)

So, you see, this kind of hitter existed for sure, and some even were outstanding hitters. Thing is, do I believe that Ray Dandridge would likely have been the 31st best hitter (by batting runs) in baseball for 70-something years if plopped into MLB? The bell that goes off for me is seeing George Kell’s name. Dandridge is precisely the same player: just one season’s difference in PA, both annual .300 hitters, walked at damn near the same frequency, hit for very similar power (most of it tied up in doubles), didn’t lose too much value to the DP despite a righty bat, better than average baserunning. This really makes me uneasy for the accuracy of this particular player.

Finally, Dandridge surprised me on the bases. I’d long imagined him to be a roughly average runner, but he stole at a considerably higher clip than his leagues (about 50% more often, adjusted for his team’s tendencies to steal). This is generally consistent with a player who runs up 20 or so Rbaser in a career.

I’m not yet ready to cast a ballot for Dandridge. I think there’s more detective work to be done on him. It’s possible he’s a nearly uniquely valuable hitter who is simply a better version of George Kell. Or our method may be overcompensating in some way. Or perhaps there’s another explanation yet. Maybe guys who rely on contact can fare better in leagues where the quality of play may open more holes for balls to fall in. I don’t know.

Judy Johnson

Judy Johnson is the one Negro Leaguer in the Hall of Fame that I’m certain doesn’t belong there. As you’ll soon see in the MLEs, he’s the same as dozens of others with substantial major league careers at the hot corner.

Johnson’s election came via the panel of former Negro Leaguers who elected the first nine, a kind of all-time team. Judy’s backers cited his batting average, outstanding defense, character, and leadership. In reality, Johnson may simply have been the nicest and most watched third baseman among the panelists. Ray Dandridge spent most of his career outside the Negro Leagues, in Mexico or the minors. Jud Wilson was a mean, angry son’a’gun, didn’t have a great glove, and, anyway, played as much or more at first base. John Beckwith played a lot of third, but about as much at shortstop, and he was crazier than even Jud Wilson. The guys making the decisions were well into retirement, at least in their fifties, probably their sixties, and fellows from the earlier than the mid/late 1920s and 1930s didn’t get as long a look because the voters on hand hadn’t played against them or else saw them in their latter days.

So Judy Johnson stood out because they needed to elect a third baseman to complete the team. And what’s the difference? In the 1970s when precious little Negro Leagues data had been researched, let alone published out. How would anyone really have known whether he’d earned immortality on the field or merely in the mind?

Today the data indicate that Johnson was a fine ballplayer, but not overly different from someone like Willie Kamm or Billy Werber .

Judy Johnson
Negro Leagues Stats | Bio
Career: 1921–1936
Destination: NL 1922–1934
Missing data: 1927, 1929
Honors: Baseball Hall of Fame
Year Age Lg Pos   PA Rbat Rbaser Rfield Rpos RAA   WAA Rrep  RAR   WAR
=======================================================================
1922  22 NL  3B  150  - 1    0       1    1    1   0.1    5    5   0.5
1923  23 NL  3B  250    0    0       1    2    3   0.3    8   11   1.1
1924  24 NL  3B  320   12    0       1    2   16   1.6   10   26   2.6
1925  25 NL  3B  480   11    0       2    3   16   1.5   15   31   3.0
1926  26 NL  3B  520   11    0       2    4   17   1.8   16   33   3.5
1927  27 NL  3B  550    6    0       2    3   11   1.2   17   29   3.0
1928  28 NL  3B  550  -12    0       2    3  - 6  -0.6   17   11   1.1
1929  29 NL  3B  550  - 4    0       2    3    2   0.2   17   19   1.8
1930  30 NL  3B  550  - 7    0       2    3  - 1  -0.1   17   16   1.4
1931  31 NL  3B  550  -10    0       2    3  - 5  -0.5   17   12   1.3
1932  32 NL  3B  550    0    0       2    2    5   0.5   17   22   2.3
1933  33 NL  3B  250  - 3    0       1    1  - 1  -0.1    8    7   0.8
1934  34 NL  3B  130  - 1    0       1    1    0   0.0    4    4   0.4
-----------------------------------------------------------------------
                5400    2    0      24   31   57   5.7  168  226  22.7

Hypothetical MLB Career Rankings (1871–1960)
PA: 333rd   
Rbat: t-896th
WAA: t-373rd   
WAR: 335th

Basically, Johnson appears to be a slightly above average guy, the sort who supports the better players on a championship team or who is the best player on a poor team. His bat was nearly dead average, I gave him average speed based on a stolen base record that’s at 89% of his leagues and with a nudge for likely underreporting of steals as noted by others.

Jud Wilson

[Updated 3/31/18 to correct a park factor in 1922, 1923 seasons.]

[Updated 4/3/18 with minor park factor corrections.]

His nickname “Boojum” was an onomatopoeia for the sound made by the outfield wall when one of his signature line drives ricochetted off of it. The man could hit blindfolded and his stat lines suggest a player somewhere in between Wade Boggs and Edgar Martinez. He hit for very high averages (lifetime .352), he drew a lot of walks, racked up doubles, and had enough home run power that opposing teams pitched him carefully. Like them, especially Edgar, he also got a late start in the big leagues.

On the other hand, Wilson was known as one tough cookie with a fuse that would light in a hurry even if it were underwater. He fought opposing players, teammates, umpires, whoever. One story has him dangling little Jake Stephens by the ankle out of a hotel window in the midst of a dispute of some sort. But Wilson ultimately liked to take his anger out on the baseball, which he did splendidly.

Wilson’s defense is an interesting sort of mess. On one hand he played more first base than third. On the other hand, he played better at third than at first. His career in top-flight baseball started at age 26, and he played first base. He became a regular at third at age 32. Moving the wrong way on the defensive spectrum is, um, highly irregular. So for this purpose we’ve started him at third and kept him there until the point in his career where he returned to first base, age 38.

Jud Wilson
Negro Leagues Stats | Bio
Career: 1922–1944
Destination: NL 1922–1937
Honors: Baseball Hall of Fame, Hall of Merit

Year Age  Lg Pos   PA  Rbat Rbaser Rfield Rpos RAA   WAA  Rrep RAR   WAR
========================================================================
1922  26  NL  3B  560   19    0       3     4   26   2.5   17   44   4.2
1923  27  NL  3B  590   23    0       4     4   30   3.0   18   49   4.8
1924  28  NL  3B  570   35    0       4     4   43   4.3   18   61   6.2
1925  29  NL  3B  590   24    0       4     4   32   3.0   18   50   4.8
1926  30  NL  3B  540   39    0       3     4   46   4.7   17   63   6.4
1927  31  NL  3B  580   53    0       4     4   60   5.9   18   78   7.9
1928  32  NL  3B  530   38    0       3     3   44   4.3   17   61   6.0
1929  33  NL  3B  570   35    0       3     3   42   3.8   18   60   5.4
1930  34  NL  3B  600   42    0       4     3   49   4.2   19   67   5.9
1931  35  NL  3B  550   47    0       3     3   54   5.4   17   71   7.3
1932  36  NL  3B  550   22    0       3     2   28   2.8   17   45   4.6
1933  37  NL  3B  550   19    0       4     2   25   2.8   17   42   4.8
1934  38  NL  1B  570   24    0       0   - 5   18   1.8   18   36   3.7
1935  39  NL  1B  560   27    0       0   - 5   22   2.2   17   40   4.0
1936  40  NL  1B  400    8    0       0   - 4    5   0.5   12   17   1.7
1937  41  NL  1B  100    2    0       0   - 1    1   0.1    3    4   0.4
------------------------------------------------------------------------
                 8400  458    0      43    24  525  51.3  262  786  78.0
Hypothetical MLB Career Rankings (1871–1960)
PA: t-67th
Rbat: 25th
WAA: 18th
WAR: t-18th

Wilson’s overall defense in the Negro Leagues at third base comes in at about 3 runs saved per 154 games. I knocked it down to 2 per 154, then beginning at age 34 decremented by a run per year until he switches positions at 38. His defensive reputation suggests he played the ball of his chest as necessary, so he’s no Brooks Robinson. But he made enough plays to show up as a mild positive. At first base he looked like a net-neutral glove.

Overall, a lefty Edgar Martinez who could stay on the field is a pretty good comparison if you want a mental visual. Edgar racked up about 70 more batting runs in 274 additional plate appearances to what we’re MLE’ing Wilson at. But Martinez loses that and more in the positional exchange between third base and DH, costing him 9 WAR versus Wilson’s career total of 77. But this is hair splitting. Same kind of player, even if Wilson’s bat didn’t quite match Edgar’s. You get the picture.

* * *

So, the hot corner. Pretty interesting bunch of guys. Next Wednesday it’s on to our first batch of shortstops. We’ll check in with one of Blackball’s baddest dudes, John Beckwith, the criminally under recognized Grant “Home Run” Johnson, and the great John Henry “Pop” Lloyd.

Advertisements

Discussion

2 thoughts on “Evaluating Negro Leagues Third Basemen

  1. Did you work up Oliver Marcelle or John Beckwith?

    Posted by Michael Webber | November 16, 2017, 11:48 pm
    • Hi, Mike!

      Yes, I’ve run them both. You’ll see Beckwith in the SS article next week. Marcelle will be coming a few months from now. Don’t have him in front of me due to biz travel, but IIRC he comes out better than Judy Johnson, but not electable.

      Posted by eric | November 17, 2017, 12:18 am

Tell us what you think!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

Institutional History

Advertisements
%d bloggers like this: