Jus taking Gary's list for a moment, and asking one question: how old would each of these guys have been in 1948.

YOUNGER
Art Pennington – 25
Marvin Williams – 25
Bonnie Serrell – 26
Héctor Rodríguez – 28

MIDDLE AGE
Leon Day – 31
Lonnie Summers – 32
Bus Clarkson – 33
Theolic Smith – 34

OLDER
Ray Dandridge – 35
Pedro Formenthal – 35
Jesse Williams – 35
Booker McDaniels – 36

In general, as noted above, it's interesting that among the younger guys, only Hec made it (Pennington and Williams being better hitters and probably better overall players than Hec or Bunny), and it took him four years. Clarkson, alone, among 30+ players made it, which may be either a feather in his cap or the playing out of context or both.

It's worth asking this question too: which of the MLB jumpers came back, and how old when they did?
-Max Lanier: jumped at 30, returned at 33.
-Sal Maglie: jumped as 28 year old rookie, returned at 33.
-Lou Klein: jumped at 27, returned at 30.
-Fred Martin: jumped as 31 year old rookie, returned at 34.
-Mickey Owen: jumped at 29, returned at 33.
-George Haussman: jumped at 29, returned at 33.
-Roberto Estalella: jumped at 34, returned at 38. (cue Sunnyday!)
-Danny Gardella: jumped at 25, returned at 30.
-Luis Olmo: jumped at 24, returned at 29.
-Roberto Ortiz: jumped at 29, returned at 34.
-Nap Reyes: jumpted at 25, returned at 30.

This is interesting in that most of these guys returned post-30, all post-29. Very different from Gary's list.

Re. Estalella, I would say that any case you (I) might want to make would be based on his being held back prior to 1946. In various trials with the Senators in 1935-36-39, he hit .279 with a .391 OBA and 38 XBH in about 400 PA. OPS+ 153-181-126. Yet he only got his 500th PA at the age of 31 in 1942.

In 1942-43-44-45, against diluted competition especially the latter 2 years, granted, his OPS+ was 130-123-125-142 in > 1700 PA, now at age 31-34. Career OPS+ 127 playing almost entirely after his prime.

So anyway, his is the very unusual case of a guy who got a shot in the MLs (unlike black guys at the time), but clearly was held back because he was Hispanic. And before 1946-47. But what is unusual is that, unlike Clarkson or Marvin Williams, you might *think* he got a fair shot. But you'd be wrong. Just like Don Newcombe had what looked like a fair ML career, but didn't.

Oh golly, I just realized what this meant.

I've been wondering what the heck Bobby Estalella did in 1948 and with the rest of 1949. I have never found any information about his ballplaying whereabouts. I'd long wondered what happened to him.

And the answer is in my post above and in Gary's explanation of it. Estalella was blacklisted due to MxL participation and therefore couldn't find any work in OB. In addition, he couldn't even play in the CWL in his home country, which Gary noted was in OB at that time, until the lifting of the suspension in 1949.

To back this up a little bit, his final MLB game was September 16, 1949, but played in only 8 all year. I would hardly be surprised if they were all in September...which was after the ban was lifted. I can't locate any documentation of his other games, so I can't confirm it, nor of his transactions for 1949 to know precisely when he reupped with MLB.

Cross posted to his thread.

Gary, nice collaboration with Eric and Kevin

Marc s
For the record, Hector [Rodriguez] played 124 games --113 at 3B, the rest presumably as a PH--for the ChiSox in 1952. His lines were 1-40-.265 or .265/.346/.307 (ouch) for an OPS+ of 82. He appears from just a tiny bit of info to have been a pretty good defender--at least average and probably better. No power, but 47 BB and 22 K. 15 XBH (all 2B except 1 HR) in > 450 PA.

The 1969 Baseball Encyclopedia (Big Mac1) gives the pinch-hitter record 8ab 2h. That leaves three games as a pinch-runner or a p-h without any at bat.

For all you people clamoring to know every detail of Clarkson’s brief ML career, here’s a rundown of all his games:

Apr 30 vs. Pit Clarkson replaced Cusick at ss, hit 1 for 2, 1 run scored; no chances.
May 1 DNP
May 2 vs. Cin Clarkson, ss, batting 8th, 0 for 3 with walk; 4 po, 1a
May 3 vs. Cin Clarkson, ss, batting 8th, 1 for 3; 1po 1a 1e
May 4 (1) DNP 1st game
May 4 (2) vs. StL Clarkson, ss, batting 7th, 0 for 3; 3po 1a, removed in late innings for Cusick.
May 6 DNP
May 8 DNP
May 10 vs. NY: Clarkson, ss, replaced Cusick, who was struck in the chin by a grounder in the eighth; no chances or plate appearances.
May 13 DNP
May 14 DNP
May 16 DNP
May 17 vs StL Clarkson pinch-hit single for Jones in 9th.
May 22 (1) DNP
May 22 (2) DNP
May 23 DNP
May 24 DNP
May 27 (1) DNP
May 27 (2) vs Phi Clarkson, ss, batting 6th, 1 for 3 with 1 rbi; 1po 1a
May 29 DNP
May 30 (1) DNP
May 30 (2) DNP
May 31 Manager Tommy Holmes fired; Charlie Grimm brought up from Milwaukee to replace him.
June 1 (1) DNP
June 1 (2) DNP
June 1 Grimm’s first transactions as manager were to bring up Johnny Logan from Milwaukee and send Clarkson & Gene Conley down on “24-hour recall.”

(At some point before June 11 Clarkson was recalled from Milwaukee.)

June 11 vs. Pit: Clarkson, ph, flied out for Surkont (p) in eighth
June 12 vs. Pit: Clarkson, 3b, batting 5th, hit 1 for 3, 2 runs, 1po 2a
June 13 vs. Chi: Clarkson, 3b, batting 5th, hit 0 for 3, 1po 3a 1e
June 14 vs. Chi: Clarkson, ph, grounded out for Sisti in 15th inning.
June 15 (1) DNP
June 15 (2) DNP
Jun 17 DNP
June 18 vs. Cin (1) Clarkson, ph, 0 for 1, struck out for Cole in 6th
June 18 (2) DNP
June 19 DNP
June 20 DNP
June 21 vs. StL: Clarkson, ph, lined out for Mathews in 7th.
June 22 vs. StL: Clarkson pinch-hit for pitcher, walked in 8th.

(I haven’t found a record of Clarkson’s final return to Milwaukee.)

To sum up:
1) Clarkson started four games at shortstop, two at third base; he pinch-hit six times.
2) For Tommy Holmes, Clarkson played only shortstop, pinch-hit once, and batted 7th and 8th when he started; for Charlie Grimm he played third base, pinch-hit five times, and batted 5th in his two starts.
3) As someone pointed out in another thread, Grimm’s first act as manager was to send Clarkson down and bring up Logan; however, within a week and a half, he recalled Clarkson again.
4) Holmes played Clarkson four times in six games (he hit 2 for 11), then only three times in the next 16 games; Grimm, starting June 11, used him in four straight games (he hit 1 for 8), then in three of the next nine.

This thread is similar to the Beckwith thread in the following respect: if this guy was so great, how come nobody wanted him badly enough to keep a strangle hold on his services? Both were fine hitters (although Beck was substantially better), but still...as a number of posters have said, something doesn't smell right, feel right, look right. On the one hand his numbers are great--he's even outhitting Willard Brown in the Texas League. On the other, Holway (and additional sources, not just the racially-biased Sporting News) really don't take him very seriously. How to sort this out?

Compared to Beckwith, this subject is even tougher, because you're juggling balls in the air for minor league, major league, Mexican league and Negro League portions of Clarkson's career--not just the NeL problem, as was the case for Beck.

I believe I have some answers, as I've promised on the ballot thread, but it's going to take me awhile to frame the argument satisfactorily, and then try to back it up with examples. That's a whole lot of work. But I'll give you a tiny preview.

There's a great deal of merit in attempting to right the cultural (and moral) wrong of ballplayers being denied their proper place in the baseball world because of the color of their skin. And really, with the exception of whatever anecdotal evidence seems credible, number crunching--including or especially "what might have beens" about MLB--seems to be the only way we have to do it.

But beware! Statistical analysis has, at its very basis a (sometimes unstated), assumption about the structural integrity of the engine that's generating those numbers.

And the heart of the problem lies in the fractured nature of the blackball leagues' structural history. When my co-author Dom Denaro and I looked for a place to begin our work, we selected the Eastern Colored League from 1923 to 1927, precisely because the two-league structure in place at the time was financially and administratively stable enough so that we felt statistical analysis was a legitimate way to proceed. In other words, we made it as easy on ourselves as we could.

The second two-league era, from 1937 to 1946, is already a big step down from that, because you've got the flight to Mexico, WW2, and the relative weakness of the NAL all pecking away at the legitimacy of one's analytical enterprise. You may wonder at the fact that stars that shown as brightly in the heavens as Sam Jethroe and Willard Brown didn't make any appearance at all, on any level, in the Pittsburgh Courier's 1952 all-time all-star list. It's because there were questions in the minds of some of the voters about the quality of play in the NAL.

And it goes steeply downhill from there. I give very little weight to any NeL stats after 1946, and not just because Jackie was physically in Montreal. It's because the entire purpose of Negro League baseball had changed: it had become a showcase for guys aspiring to the new goal, rather than an entity unto itself. For those not young enough to have such aspirations, or too well-known to knock 6 or 8 years off their age, it became a way to earn a living, and that's about it. As to stats after 1948, just throw them into the garbage; black baseball was D-league baseball by then.

So what Denaro did was divide league blackball history into 7 sections, EACH with their own characteristics, and hence strengths and weaknesses (read "statistical evaluations"), relative to MLB. This idea has been informally adopted by some of the Negro League historians I know, and I hope will eventually become standard practice. In the meanwhile, the idea of ONE multiplier for the NeL evaluations relative to MLB no longer makes any sense to me.

And among those 7 eras, every year in the 1940s after 1941 except (arguably) 1946 has something very wrong with it. And even '40 and '41 have some important players south of the border. Those years can't be treated statistically in anything like a normal fashion--they just can't.

So that begins to explain those stratospheric batting averages in the 1940s for guys like Sam Jethroe, Luis Marquez and others. But it also applies to Clarkson (and even more to Marvin Williams when I get to him). Clarkson played for the Crawfords at a time when Satch had taken the heart of the team to Mexico, and they were changing cities on a yearly basis. It was a highly-questionable franchise at this point.

Then, for the Philadelphia Stars we're talking about years in the 1940s when it is reasonable to raise doubts about the quality of the league as a whole.

So where does that leave us? Were Jethroe, Marquez and Clarkson very good hitters? Of course they were. But when we start talking about Clarkson ranking among the great infielders of his era, red flags go up for me. Why? Because I don't know, and I don't know how to find out. What I do know is that some of the conversion mechanisms that put him there are suspect. Certainly the ones from the Negro Leagues data are. In the meanwhile, we have to give at least *some* weight to whatever seems to qualify as expert opinion, to help us regain our balance from the damage done to our confidence in how to compute the numbers.

Clarkson has not been considered one of the Negro League greats, then or now. He was not among the original 94 candidates studied by the HOF election committee in February of 2006. If they had extended their list to 200 names, I'm not sure Clarkson would have been on it. I think Alec Radcliff and Parnell Woods (among thirdbasemen) and Rev Cannady and Frank Austin (among shortstops) would have been considered before Clarkson, if that gives you the flavor of the thing.

Now, in fairness, that judgment TOO is unreliable. Historians of blackball, like historians of MLB, appreciate coherent careers within coherent franchise and league contexts. So somewhere inbetween the statistical outcomes and the anecdotal/historical opinions we are going to have to find our balance. And the devilish thing is, that judgment is going to vary from player to player, depending on his circumstances. And it's going to be at its most challenging for a player like Clarkson, whose circumstances were pretty chaotic.

Two more pieces of advice: when you're reading sources like Jim Riley's bio book, be aware that he's cherry-picking particularly dramatic stats. In the case of Riley, at least, he's not doing that in order to be a Negro Leagues partisan, he's doing that in order to create a book that's interesting. How many times, exactly, do you want to read that some guy hit .263, .258, .271, .255, and .267 in successive years?

Second, when Negro League historians are silent on the subject of defense, that means there was nothing good to say. Silence does not equal average value, it indicates negative value. An example: Dobie Moore was not an elegant shortstop, but he could play the position, and everybody who's written about Moore says so. So that equals average value: he wasn't a defensive wizard, but he wasn't a butcher either. When you read of a player like Jud Wilson that he played thirdbase like a hockey goalie, knocking balls down with one or another body part, just picture that in your mind's eye and make a judgment accordingly. Wilson would not have played 3B at the MLB level, because MLB managers had an idea what a thirdbaseman ought to look like, and Wilson was not it.

Once again, just to be clear, I'm not saying any of this to knock Bus Clarkson specifically or blackball generally. Given that I've spent twenty-five years of my life studying the subject, if I really felt that way, I'd be certifiably nuts. I'm just trying to say that this is, in some respects, a very complicated task, and one that requires a multiplicity of evaluation tools in order just to get you into the vicinity of a reasonable answer. In the exceptionally complicated cases, like Clarkson and some of the other black players of the integration-overlap era, it also requires a humility before the excruciating difficulty of the task that I confess I find very frustrating. I wish I could be the bearer of better news.

OK, listen guys: I just lied. This has not been a "tiny" preview; I've let most of the proverbial cats out of the proverbial bag. This should probably be moved to (or copied to) the ballot thread, where it will have the most relevance.

OK, that last post exhausted me. I'm going to amuse myself here for a moment by responding to sunnyday2's excellent survey in post 200 of the White Sox' 3B misery in the early 1950s. As I am ancient, I was actually at Comiskey Park in those years, witnessing the whole mess.

Although Billy Pierce and Nellie Fox also came of age in 1951, it was really the trade with Cleveland for Minoso that energized the Sox out of their three-decades-long second-division slumber. And yes, 3B was a huge hole (although I'd very much like to have somebody explain to me what the hell happened to Bob Dillinger, who had been a big star with the Pirates. Besides, the symmetry of "Dillinger" in a Chicago uniform--wasn't there some destiny to that?)

Anyway, it was Minoso, specifically, who convinced Frank Lane to obtain Hector Rodriguez from Mexico. Minoso referred to him, in the Chicago papers, as "the great Mexican thirdbaseman." He wasn't that far wrong: Rodriquez had a big reputation, on defense, and as a hitter for average. For whatever combination of reasons, Rodriguez simply didn't produce, especially on offense, and he was gone after one season.

In 1953, there was exactly the fiasco that sunnyday2 describes. But it produced my single favorite White Sox moment of all time. In a game against the Yanks, Casey brought Ewell Blackwell, the nasty sidewinder, in to face Vern Stephens--the thirdbaseman du jour--with the bases loaded and the game on the line.

Paul Richards, whom I always admired for his willingness to think outside the box, brought in Tommy Byrne (yes, Tommy Byrne), to pinchhit for Stephens. My thought, as Byrne stepped to the plate was: how humiliating for a guy who had been a big star in '44 with St. Louis, and a big bat in Boston's lineup, to be lifted for a *pitcher*.

Well, Byrne hit a long homerun into the rightfield stands. A perfect moment in the history of a very imperfect franchise.

If the question at this point is, what exactly does this burniswright character want from us, I'll try to give an answer here.

My metaphor would be: you are all judges at an evidenciary hearing; you need to rule on what evidence ought to be heard in court, and what ought to be thrown out. So, in the Clarkson case, let's go through the list.

First of all, the MLB data is meaningless, in the literal sense of the word: there's no way to derive any meaning from it about what Clarkson could or could not do.

The '37 and '38 Crawfords are league teams, but '39 Toledo and '40 Indianapolis are not. Now, the problem with the latter two is the question of who they're playing. The majority opinion, among Negro League historians, is that only league games should count on a player's record; that would mean everything from those teams is thrown out. There is a minority opinion, however, that argues that games against teams that *are* league teams should count. If anybody feels like doing a hellish amount of work, trying to find Craws games and Clarkson's record against league teams in '39 and '40, I tip my hat to them.

But, beyond that question, I don't know anybody who feels that, during the league era, games against all levels of competition (AALOC) should count. In the pre-league era, you've got no choice regarding AALOC stats; the only thing you can do, besides making individual-game judgments about substantial or insubstantial competition, is to compare an individual's statistical performance against that of his teammates.

Here's the problem with non-league blackball teams. The assumption is that most or all of their opponents are semipro teams. Once again, AALOC stats are not going to be allowed in. So, if you wanted to make an argument for *some* stats, it would have to be on the same basis as the pre-league era--by judging opponents. "Semipro" can have a wide range of meanings. At the high end, there were a few traditionally strong teams (e.g. The Brooklyn Bushwicks) that could give a league team a legitimate tussle. But at the other end of the spectrum you've got "Bob's Auto Parts" at 128 South Main Street in Clockerville.

Which end of the spectrum a non-league team played at also had a lot to do with their reputation. Since both the Homestead Grays and Kansas City Monarchs spent significant numbers of years barnstorming, this becomes a legitimate historical inquiry for two of the greatest clubs in NeL history. I'm not so sure about the latter-day Craws.

Moving along, '40 and '42 Newark count, as does '46 Philly, subject to my misgivings expressed in the previous post. But I would throw out '49 and '50 Philadelphia entirely.

In fact, I would be inclined to throw out ALL data derived from baseball that's judged to be below about the single-A or double-A level. There's just too little wisdom to be gleaned there. We are, after all, judging the great stars in baseball history; what does it matter what they did at the B-league level? If we were writing a history of the Western Montana League, that would be different; we aren't.

OK, let's move on to the war years. Oh boy, this is not something I'm happy about trying to tackle. Apparently, all or most of you are fine with filling in WW2 service years with mathematically-derived estimates. I'm not fine with it. It seems like a very mixed metaphor to me. If Bob Feller and Ted Williams miss a bunch of years defending their country, lets give them patriotism credit, or bravery credit, or citizenship credit. But baseball credit? For games they never played in?

It seems to me that years missed in military service fall into a larger category called "Games Missed Through No Fault Of Their Own." And in that case, we're already pretty far down a slippery slope. You have to give Eddie Waitkus his full year back in 1949; it's certainly not his fault that some lunatic stalker shot him.

AND (I admit this is a hobby-horse issue of mine), you have to acknowledge that Chino Smith was simply the greatest hitter who ever lived, since I can absolutely assure you that it wasn't his fault he died at age 30.

But here's where it gets sticky: do I really want to argue that the guys who went off to war to save us from Hitler and Hirohito get a bunch of win shares deducted because they did that? Even as I'm writing that, it creeps me out.

My suggestion is that everybody should take a very deep breath and figure this one out for themselves. Personally, I just can't get past the fact that these are imaginary at bats in imaginary games. For me, as unfair as it may sound, the war years are out. The rationale for my position is that s**t happens.

Now, have I painted myself into a corner with "s**t happens"? After all, that argument could conceivably cover the entire history of racial segregation in this country, and with it, Negro League baseball.

Obviously, I don't think so. And the reason is that racial segregation in baseball was, when you get right down to it, a criminal conspiracy. By any reasonable interpretation of The Constitution, you can't keep Josh Gibson out of MLB because he has a deep tan. It was just plain wrong, and it's a wrong that has to be righted.

OK, if you throw out '39, '40, '43 to '45, '49 and '50 for Clarkson, that still leaves a lot of baseball--just not very much Negro League baseball. I think that helps to explain why Negro Leagues historians haven't taken him seriously. But that doesn't mean that BBTF shouldn't take him seriously. A lot depends on how you evaluate the importance of his many and varied minor league affiliations, as well as his winter league play.

As to the latter, let me simply admit that I don't understand how win shares are adjusted for 12 months of baseball in comparison to other guys who played six months of baseball. You don't have to explain it to me--I'm just raising the issue.

So, in summary then, if one of our local sabermetrics wizards wants to re-calculate Clarkson with all the restrictions I've just put on his data, plus the negatives on defense that I suggested previously, that might be very interesting. And if he still comes out to be a great player, so be it. I'll just be very surprised.

P.S.: once again, I've raised a bunch of issues here that belong partly on the Clarkson thread and partly on the ballot thread. You guys handle that however you feel is appropriate.

I think adding a more substantial discount for 39, 40, 49, 50 than the rest of his years makes a lot of sense. Looking at post 103 it looks like we did that already for 49-50 but not 39-40.

His 1939 (8 win shares) wasn't really helping his case much anyway. I think we could conservatively remove that year entirely. Discounting 1940 at .86 instead of .90 would knock off a couple win shares and bring it more in line with an increasing career path from ages 24-27. It's the 1953-1954 and the flat war-credit peak years that don't fit the curve. Even if you swap the years in the timeline, the rejuvenation in the Texas League is impressive.

All the discussions on Bus Clarkson don't clear up Clarkson as much as they leave me more and more convinced that Willard Brown was a glaring mistake.

When my co-author Dom Denaro and I looked for a place to begin our work, we selected the Eastern Colored League from 1923 to 1927,..

Is this a book in process, or some research already completed?

KJOK: it was published in 2003, and is monograph length, not book length.

Here is my three nanoseconds of shameless self-promotion: AJ Publications is run by Eric Naftaly. Write to Eric at ajericn@mail.com and he will be glad to sell you a copy. Like Patrick Rock's 1923 NNL, it was originally written in conjunction with creating a baseball sim cardset/computer disk (for APBA), but it can stand alone as well.

June 1 Grimm’s first transactions as manager were to bring up Johnny Logan from Milwaukee and send Clarkson & Gene Conley down on “24-hour recall.”

(At some point before June 11 Clarkson was recalled from Milwaukee.)

June 11 vs. Pit: Clarkson, ph, flied out for Surkont (p) in eighth
. . .
June 22 vs. StL: Clarkson pinch-hit for pitcher, walked in 8th.

(I haven’t found a record of Clarkson’s final return to Milwaukee.)

Gary,
Do you care much? If so, what newspapers have you checked? I will do some baseball at the Harvard U library Dec 26-28.

Gary,
Do you care much? If so, what newspapers have you checked? I will do some baseball at the Harvard U library Dec 26-28.

The only papers I've looked at for this have been the Chicago Defender and New York Times. I guess I'd like to pin down the exact dates of these transactions sometime, but it's no big deal. At this point more general commentary on Clarkson might be more interesting, in Boston as well as Milwaukee (was he really a popular star there?). There was a suggestion in a Defender article that Charlie Grimm, who had declared he would not leave Milwaukee to manage the Braves just a couple of weeks before taking the job, sent Clarkson back to Milwaukee as a sort of peace offering to Milwaukee fans. It would be interesting to see if there's any truth to that.

A couple of minor points:

The '37 and '38 Crawfords are league teams, but '39 Toledo and '40 Indianapolis are not.

Clarkson did indeed play for league teams in 1939 and 1940. The Toledo Crawfords started in the Negro National League in 1939, then moved to the Negro American League on June 20 (see article, “Toledo Replaces Indianapolis in American League,” in Chicago Defender, July 1, 1939). The Crawfords don’t appear in the first-half NAL standings published in The Negro League Book, but that’s a mistake (the totals don’t balance—there are 64 wins, 50 losses). They appear in published NAL standings in the Chicago Defender (for example, July 15, 1939), and are constantly mentioned as a league team. In fact, Crawfords’ manager Oscar Charleston was chosen to manage the West in the East-West Game that year (the two teams were league all-star teams; the West was known as the “Negro American League All Stars”).

In 1940, the Indianapolis Crawfords were definitely in the NAL; they are listed in the NAL standings in the Chicago Defender that season, and also in The Negro Leagues Book. An overly optimistic article in the May 11 Chicago Defender dubbed them “the team to beat in the Negro American League.” By the end of June Clarkson had been awarded to Newark to compensate the Manleys for losing Satchel Paige.

His 1939 (8 win shares) wasn't really helping his case much anyway. I think we could conservatively remove that year entirely. Discounting 1940 at .86 instead of .90 would knock off a couple win shares and bring it more in line with an increasing career path from ages 24-27.

I’m pretty sure the Negro League translation factors of .90/.82 for ave/slg were originally derived by Chris Cobb from the careers of players who performed in the war-, integration-, and Mexico-weakened Negro Leagues of the 1940s. The “standard” system of creating NeL MLEs thus already recognizes the relative weakness of this era (though not of the NAL specifically); the trouble is with earlier eras (especially the one-league period of 1933-36), where the .90/.82 factors probably underestimate the strength of the Negro Leagues. Gadfly wrote at length about this, somewhere or other.

I’m pretty sure the Negro League translation factors of .90/.82 for ave/slg were originally derived by Chris Cobb from the careers of players who performed in the war-, integration-, and Mexico-weakened Negro Leagues of the 1940s. The “standard” system of creating NeL MLEs thus already recognizes the relative weakness of this era (though not of the NAL specifically); the trouble is with earlier eras (especially the one-league period of 1933-36), where the .90/.82 factors probably underestimate the strength of the Negro Leagues. Gadfly wrote at length about this, somewhere or other.

Exactly. For war years seasons, I applied a competition discount of 10% to the NeL numbers. I did not use any NeL data for seasons after 1948. The full description of the data I used for calculating the conversion factors is somewhere on the site (and somewhere in my hundreds of HoM files . . . ). Incidentally, only the BA factors was calculated directly from data: the SA factor was derived from it. It would be good to go back and revisit these conversion factors using data from the Hall of Fame project, since none of that had been published at the time I had done my work.

(I hope the formatting on this turns out all right. In the "live preview" the columns are okay, but the spacing between paragraphs is really weird.)

On the very limited (but still significant) issue of Bus Clarkson vs. Willard Brown: I thought I’d put together a comparison of them when they played in the same league (Mexican League, Puerto Rican Winter League, and Texas League). They were the same age and often played on the same teams.

In the Mexican League, 1940-41, at ages 25-26, this is how they hit:

 G  AB   H  D  T  HR   R  RBI  BB  K HBP  SB  AVE  OBA  SLG
Brown  98 419  136 25  4  10  69  85  14 20  0   18  325  346  475
Clarkson 101 406  136 27  6  20  79  96  48 72  1 8  335  405  579

In 1940, they played on the same team (Nuevo Laredo), Brown for 70 games, Clarkson 19. Dr. Chaleeko, in a study of Mexican League park factors, gave Nuevo Laredo a BRF of .97. In 1941, Clarkson went to Tampico for 82 games (BRF .95), Brown to Aguila of Veracruz for 28 games (also .95).

In the 1953-54 Texas League, ages 38-39, this is how they hit:

 G   AB   H  D   T  HR   R  RBI  BB  K  HBP  SB  AVE  OBA  SLG
Brown 282 1105  345 72  7  58  183 228  70 108  - 5  312  353  548
Clarkson 294  988  323 53  3  60  200 222 204  99  -   18  327  442  569

In 1953 they both played for Dallas; in 1954 Brown played for Dallas and Houston, Clarkson for Beaumont and Dallas. So roughly three-fourths of these two years were spent in the same home park.

I don’t have year-by-year stats for both of them from the Puerto Rican Winter League, but I do have their career stats. Brown played in Puerto Rico for ten winters between 1941/42 and 1956/57 (between ages 26 and 41); Clarkson for ten winters between 1940/41 and 1954/55 (ages 25 to 39). They spent eight seasons in the league together, four of them for the same team.

The Puerto Rican stats don’t include walks. For Brown, these are the BB/AB rates we have:
1940-41 MxL: 14/419, .033
1937-49 NeL: 109/1769, .062
1953-54 TxL: 70/1105, .063

For Clarkson we have:
1940-41 MxL: 48/406, .118
1946-47 MxL: 103/521, .198
1950-52 AA: 125/724, .173
1953-54 TxL: 204/988, .206

For Brown, I used the highest walk figure available, .063. For Clarkson I simply averaged the four rates (getting .174, nearly the same as his walk rate in the American Association). This is extremely advantageous to Brown, as when they actually played side by side in the Mexican and Texas Leagues, Clarkson’s walk rate was well over three times Brown’s.

This is what the Puerto Rican stats look like, with hypothetical walks added in:

  AB   H  D   T  HR   R  RBI  BB   K HBP  SB  AVE  OBA  SLG
Brown  1940 679 135 27 101 378   -  120  -   -   -   350  388  604
Clarkson  2063 621 109 15  98 448   -  359  -   -   -   301  405  511

Adding everything (Mexico 1940-41, Puerto Rico 1940-57, and Texas 1953-54), this is what we get:

   AB   H D  T  HR   R  RBI  BB   K HBP  SB   AVE  OBA SLG
Brown 3464 1160 232 38 169 630  -   204  -  -   - 335  372 570 
Clarkson 3457 1080 189 24 178 727  -   611  -  -   - 312  416 535

(Sorry, I'll try it one more time a different way.)

On the very limited (but still significant) issue of Bus Clarkson vs. Willard Brown: I thought I’d put together a comparison of them when they played in the same league (Mexican League, Puerto Rican Winter League, and Texas League). They were the same age and often played on the same teams.

In the Mexican League, 1940-41, at ages 25-26, this is how they hit:


In 1940, they played on the same team (Nuevo Laredo), Brown for 70 games, Clarkson 19. Dr. Chaleeko, in a study of Mexican League park factors, gave Nuevo Laredo a BRF of .97. In 1941, Clarkson went to Tampico for 82 games (BRF .95), Brown to Aguila of Veracruz for 28 games (also .95).

In the 1953-54 Texas League, ages 38-39, this is how they hit:


In 1953 they both played for Dallas; in 1954 Brown played for Dallas and Houston, Clarkson for Beaumont and Dallas. So roughly three-fourths of these two years were spent in the same home park.

I don’t have year-by-year stats for both of them from the Puerto Rican Winter League, but I do have their career stats. Brown played in Puerto Rico for ten winters between 1941/42 and 1956/57 (between ages 26 and 41); Clarkson for ten winters between 1940/41 and 1954/55 (ages 25 to 39). They spent eight seasons in the league together, four of them for the same team.

The Puerto Rican stats don’t include walks. For Brown, these are the BB/AB rates we have:
1940-41 MxL: 14/419, .033
1937-49 NeL: 109/1769, .062
1953-54 TxL: 70/1105, .063

For Clarkson we have:
1940-41 MxL: 48/406, .118
1946-47 MxL: 103/521, .198
1950-52 AA: 125/724, .173
1953-54 TxL: 204/988, .206

For Brown, I used the highest walk figure available, .063. For Clarkson I simply averaged the four rates (getting .174, nearly the same as his walk rate in the American Association). This is extremely advantageous to Brown, as when they actually played side by side in the Mexican and Texas Leagues, Clarkson’s walk rate was well over three times Brown’s.

This is what the Puerto Rican stats look like, with hypothetical walks added in:


Adding everything (Mexico 1940-41, Puerto Rico 1940-57, and Texas 1953-54), this is what we get:

Well, that didn't work. Bottom line is: in leagues they played together, mostly for the same team, in seasons that are spread out over their whole careers (from roughly ages 25 to 40), Brown hit 335/372/570; Clarkson 312/416/535. Brown has a 35-point advantage in SLG, Clarkson a 44-point advantage in OBP.

Cumberland Posey, picking his “All-American” Negro League team in the Pittsburgh Courier, November 7, 1942 (courtesy of Paul Wendt; see the Irvin thread for the link), wrote this on shortstops:

“We saw more good shortstops in 1942 than in any year previous, since 1929.

“Willie Wells of Newark; Tommy Butts of Baltimore; Williams of Kansas City; Sam Bankhead of the Grays; Brown of Memphis; Martinez of the Cubans: ‘Bus’ Clarkson of the Philadelphia Stars were the leading shortstops.

“It would be hard to find three better fielding and throwing shortstops than Butts, Martinez and Williams. Clarkson was of the Wagner and Lloyd type, his body, big hands, hard as nails, with a good arm and a long distance hitter.

“He was not as good a fielder and general all-around shortstop as Bankhead or Wells.

“WELLS STILL TOPS

“Sam Bankhead had a good year in his first full season at shortstop.

“Willie Wells of 1942 was a better all-around player than he has been in any years since we first saw him in 1927. That is saying a lot, as he has been placed in the immortal class yearly. He belongs in the same class with Lloyd, Lundy, Chacon, Stevens and Moore.”

bump

Using Dr. Chaleeko's MLEs (final edition), I calculated WAR for Clarkson following Sean Smith's method:

I calculated wOBA using the (1.8*OBP + SLG)/3 approximation; since Dr. Chaleeko only gave OPS+ and WS in his final MLEs, I used the 2nd-to-last MLE's, adjusted downward to match the final MLE's.
I assumed Clarkson was a -5 runs/600 PA fielder throughout his career.
I used a 9 run position adjustment for SS and a 0 run position adjustment for 3B, which is consistent with what other players (Arky Vaughan, Bob Elliott) received. Like Dr. Chaleeko, I assumed that Clarkson was a full-time SS through 1945 and a full-time 3B from 1946-1956.
I estimated the runs-to-wins converter using rolling three-year averages of the runs-to-wins converters for Arky Vaughan (1938-1941), Bob Johnson (1942-1945), and Enos Slaughter (1946-1956). In some cases I used 2 year averages because the player retired immediately following that season. I chose Vaughan, Slaughter, and Johnson because they appear to have had similar careers to Clarkson; I'm not really sure, however, if this is the best way to calculate runs-to-wins converters.

Here is Clarkson's WAR, by season:
1939 1.8
1940 5.2
1941 4.4
1942 5.3
1943 5.0
1944 5.5
1945 5.7
1946 3.4
1947 4.1
1948 4.8
1949 3.6
1950 3.2
1951 2.8
1952 2.9
1953 6.0
1954 6.1
1955 2.3
1956 1.7
Total 73.7

Clarkson's total of 73.7 is the second-highest on this year's ballot, after Bagwell. Dropping Clarkson's defense to -10 runs per 600 (which would make him an awful fielder over the course of his career) reduces his WAR by 7.3; decreasing his WW2 credit drops him another 1.2 WAR. Of these two adjustments, I believe that reducing Clarkson's WW2 credit is legitimate; his defense, however, is already rated poorly enough, and there is no reason to believe Clarkson was a historically awful fielder. -5 runs for the duration of Clarkson's career may in fact be too low.

The big issue with the MLE's is the treatment of Clarkson's 1953 and 1954 Texas League seasons. Clarkson should have been called up to the majors as a result of his huge numbers in 1953 and 1954; however, age and racism were probably sufficient to keep major league teams from signing Clarkson. If we cut those two seasons in half, Clarkson is still at 67.7 WAR, which makes him a very impressive candidate--in fact, Clarkson would have the highest WAR of anyone in the backlog at 67.7.

There is a side mention in the new Larry Tye book on Satchel Paige of Bus Clarkson being the best hitter in the Puerto Rico Winter League. Satchel intentionally walks Clarkson with the bases loaded early in the game to force in the run rather than pitching to Clarkson. That was the only run Satchel gave up that particular game.

Those conversions are eye popping. That's comfortably in the median of the shortstops, the 67.7 WAR matches Ernie Banks. Among 3B that's the same value as Paul Molitor. The conversions, even with the discounts, make him a median HoM guy. It's a different shape than most of these guys - there are hardly any below average glove infielders that aren't complete sluggers.

As Gary A noted above, Bus Clarkson wins the head-to-head hitting competition with Willard Brown. I think of Willard Brown as a lesser version of Andre Dawson (worse defense) and a mistake but he was elected to the HoM based mainly on his bat. If you can outhit a Hall of Merit outfielder and play SS, you're a pretty special player. Even if he's a 3B, he's still at the same general area on the defensive spectrum as CF Willard Brown.

Actually those WAR figures are slightly too high. I calculated Clarkson's wOBA using the (1.8*OBP + SLG)/3 approximation, while I calculated the league averages using the standard wOBA formula from Tango's website. For whatever reason, the approximated league average wOBAs are all about 10 points higher than the actual averages; I suspect that it's because the approximation is optimized for the current run environment and as a result it doesn't work as well for the 1940s and 1950s NL. However, I think that by using the approximation both for Clarkson and for the league average, the errors will cancel out.

So here is Clarkson's WAR, by season, but this time I approximated the league-average wOBA:
1939 1.5
1940 4.7
1941 3.9
1942 4.8
1943 4.4
1944 4.8
1945 5.0
1946 2.9
1947 3.4
1948 4.2
1949 3.0
1950 2.7
1951 2.4
1952 2.5
1953 5.4
1954 5.4
1955 1.9
1956 1.2
Total 64.4

If anything, I think I have Clarkson's WAR too low; he was probably a better fielder than -5 runs/600 PA every year. Later today I'll look up the fielding stats for some other SS, and possibly adjust Clarkson's defensive value upward. Also, I think that rather than assuming Clarkson was uniformly -5, it's more realistic to assume he started his career as an average or even slightly above-average fielder.

Also, I think that the lack of a great peak hurts Clarkson; as a result, his HoM case is not as strong as his 64.4 WAR indicates. His career path looks somewhat similar to that of HoMer Jake Beckley, who had 61.4 WAR and a 5.5 WAR peak. It took Beckley a long time to get elected, correct?

I would start him as average and decline toward negative until he moves to 3B, then start him back at above average and repeat the process.

I think Clarkson is a cut above Vern Stephens and Miguel Tejada. Is Jeff Kent a good comp?

I assumed the following fielding numbers for Clarkson (all are per 600 PAs):
1939 0
1940 0
1941 -1
1942 -2
1943 -3
1944 -4
1945 -5
1946 0 (switch to 3B)
1947 0
1948 -1
1949 -1
1950 -2
1951 -2
1952 -3
1953 -3
1954 -4
1955 -4
1956 -5

Which gives the following season-by-season WAR:
1939 1.7
1940 5.1
1941 4.2
1942 5.1
1943 4.6
1944 4.9
1945 5.0
1946 3.4
1947 4.0
1948 4.6
1949 3.4
1950 3.0
1951 2.6
1952 2.7
1953 5.6
1954 5.5
1955 2.0
1956 1.2
Total 68.6

Although again we should keep in mind that Clarkson's career total overrates him due to his low peak. In terms of career shape, Beckley is a very good comparison. Kent, with 59.4 career WAR but a much higher peak, is a good comparison in that both Clarkson and Kent were poor-fielding middle infielders with a lot of power.

There are reasons for the missing peak. Sometimes the MLE's smooth out the years, not sure if that's the case here. Another is his peak ages are covered by war credit.

DL/227:

I think that you're right--the MLE's do tend to smooth out the years, which means we should give him a "bonus" on his peak (although it's still quite low).

Alex,
I believe we consider this a known artifact of MLE by Chris Cobb; he often includes a reminder. The method describes shorter seasons, and incomplete records of any-length seasons, on the scale of complete records for full-length major league seasons. The degree of regression to the mean stipulated by the method yields too little full-season variability. Simple replication to match the length of the major league season ("pro-rating") would yield much too much full-season variability.

Thanks, Paul. I understand what you mean about the regression of peak years.
With that in mind, do you think that Clarkson's peak season of 5.9 WAR should be considered equivalent (or nearly equivalent) to Palmeiro's peak season of 7.4 WAR? Clarkson has 6 seasons of over 5 WAR and 16 over 2 WAR; Palmeiro has 5 and 14. Therefore, it seems to me that Clarkson has a pretty strong case to rate ahead of Palmeiro.

I was not a Willard Brown fan here, and suspect he could be a mild HOM mistake (very good player, but worth voting in?).

And I've never put Bus Clarkson on my 15-man list, iirc.

But I'll keep taking in the new info, and we have no time limits on changing our minds.
So let the new concepts continue........

we have no time limits on changing our minds

Well, we have to do it before we die.

Hey, gang, please click through for my latest MLEs on Bus/zz/ter Clarkson.

Actually, you should know the baserunning will be updated in a couple weeks or so and won't necessarily be quite so high as I have him on the date of publication of this MLE.

Next up in our Negro Leagues candidates review, Bus Clarkson.

BUS CLARKSON: SS/3B
Year Age  Lg  Pos      G   PA  Bat  Bsr  DP  Fld  Pos  RAA   WAA  Rep  RAR  WAR  WAR162
=======================================================================================
1938  23  NL  SS     114  490    5   0    0    0    7   12   1.2   15   27   2.9   3.0
1939  24  NL  SS     129  550    9   0    0    0    8   16   1.7   17   33   3.5   3.7
1940  25  NL  SS     128  550   17   0    0    0    8   25   2.6   17   42   4.4   4.6
1941  26  NL  SS     122  520   10   0    0    0    7   17   1.8   16   33   3.6   3.8
1942  27  NL  SS     134  570   11   0    0    0    8   19   2.1   18   36   4.2   4.4
1943  28  WW2                      
1944  29  WW2                      
1945  30  WW2                      
1946  31  NL  SS/3B  134  570   22   0    0    0    4   26   2.9   18   44   5.0   5.3
1947  32  NL  SS/3B  135  580   15   0    0    0    4   19   2.0   18   37   3.9   4.1
1948  33  NL  SS/3B  127  540   14   0    0    0    4   18   1.8   17   34   3.6   3.8
1949  34  NL  SS/3B  126  540   15   0    0    0    4   18   1.8   17   35   3.6   3.8
1950  35  NL  SS/3B  133  570   16   0    0    0    4   19   1.9   18   37   3.7   3.9
1951  36  NL  SS/3B  100  430    9   0    0    0    3   12   1.2   13   25   2.6   2.8
1952  37  NL  SS/3B   97  410   17   0    0    0    3   19   2.1   13   32   3.5   3.7
1953  38  NL  3B      99  420   22   0    0    0    0   22   2.1   13   35   3.4   3.6
1954  39  NL  3B      84  360   22   0    0    0    0   22   2.2   11   33   3.4   3.6
1955  40  NL  3B      35  150    5   0    0    0    0    5   0.5    5    9   1.0   1.0
1956  41  NL  3B      24  100    3   0    0    0    0    2   0.3    3    6   0.6   0.6
---------------------------------------------------------------------------------------
                    1721 7350  211   2   -3   -3   63  270  28.2  229  499  53.0  55.7

Depending on how on apportions war credit, Clarkson may appear to be a slightly stronger candidate than Ben Taylor. If one gave no war credit, then Clarkson would be about as good a candidate as Ray Dandridge. If one gives him merely an average player’s credit in about 130 games a year, he’d be worth 28 WAA and about 60 WAR in 162 notation. If one gives him his career average rate of play in 130 games a year for his three missing seasons, he’s worth roughly 35 WAA and 68 WAR in 162 notation. That latter figure is basically the same as Joe Cronin or Bobby Wallace’s raw BBREF, though both look better than Clarkson once you adjust for schedule. Still that’s seemingly impressive company. Here’s the but…but Clarkson may not be as good a candidate as those guys.
A) Cronin’s WAR7 is 44 and 46 in 162 notation. Wallace’s WAR7 is 42 and approximately 44 in 162 notation. That’s considerably higher than Clarkson’s 30 WAR7 in 162 notation.
B) His fielding is based on 17 games in the Negro Leagues and 474 minor league games. That’s a small sample, and it’s also skewed toward late in his career. The shortstop fielding data is in the negative, and without early-career fielding data, it’s unclear that we really know how good a fielder he is.
C) I’m pretty skeptical that the last four years of his career would be as strong as they appear, specifically around his hitting. It’s the best hitting of his career by far. These are seasons in which he played in the AA, the Texas League, and the PCL, so they should be pretty accurate, but it’s nonetheless hard to swallow that he suddenly became a significantly better hitter starting at age 37.

In addition, Clarkson played during times that we have represented effectively already. He’s primarily a shortstop, which isn’t an area of particular need. Though he would be helpful at third base. Finally, Clarkson’s lack of durability is an Achilles heel for him.

[Cross-posted from 2020 discussion thread.]