But, as always, they also had some other odd parks that they called home. There used to be an interesting article on the net about how the club used Belmar, New Jersey, as one of their home fields in 1935.

Google:
"new york cubans" belmar

If I take the what-would-he-be-like-in-MLB approach, then I don't think he pitches, but I do think he probably turns out looking like an Al Kaline type of player with a good glove, good power, decent average, good OBP. However, I can't justifiably rank him as high as I'd rank Kaline because I'm speculating and I don't have quite enough data to support that assertion.
. . .
I'm sorry, I should have qualified that by saying that in MLB I don't think he would have suddenly adopted switch hitting

1. The nature of the counterfactual (he would have played differently in MLB) may itself provide a reason not to rank him based on that projection.
--itself, beside the lack of data to support the projection

2. Another tack in the same direction: Translated as a player-manager, would he play every day in MLB?

# 26 Gary A wrote:

Somewhere I missed 5 wins and 4 losses, but haven't been able to figure out where.

It appears that you missed his 45/46 season.

Here is an analysis of Dihigo’s wins above team for the Cuban League. I think these data are largely consistent with Chris Cobb # 50 showing his wins above team were much more impressive starting about 1935. I think they are also consistent with Gadfly’s assertion (#61) that as a pitcher he could have been a 20-game winner at least some years from 1935 to 1942 (Gadfly said starting with 1931, but Dihigo didn't pitch in Cuba during 1931-34).

Yr W L TPct WAT % tm dec
24/25 2 3 .500 -0.6 10%
26/27 2 0 .633# 0.8 7%
26/27(T) 1 0 .433 0.6 3%
27/28 4 2 .606#* 0.4 18%
28/29 2 1 .714#$ -0.2 7%
29/30 1 2 .469 -0.4 6%
30f(U) 2 0 .692# 0.7 15%
35/36 11 2 .708# 2.5 27%
36/37 14 10 .551# 1.2 35%
37/38 11 5 .509& 4.0 28%
38/39 14 2 .537 7.7 30%
39/40 6 4 .510 1.1 20%
40/41 8 3 .633# 1.3 22%
41/42 8 3 .523 3.0 25%
42/43 4 8 .500 -2.7 25%
43/44 8 1 .667# 2.5 19%
44/45 3 3 .521 -0.1 13%
45/46 5 4 .617# -0.6 15%
46/47 1 3 .379 -0.5 6%
# = team won pennant
* = excludes 4 games won by forfeit
$ = excludes 13 games won by forfeit
& = excludes 6 games won by forfeit
</pre>
Regarding league quality, I spent some time looking over the rosters in the Cuban League during Dihigo’s career. In the 1920s the quality was excellent—most of the top black Americans also played in Cuba. Due to political unrest and economic miseries, the Cuban League hit a rough patch in the early 1930s – the league missed a couple of seasons and didn’t bring in Americans another. But during Dihigo’s peak pitching years from 1935-41 the rosters look good—top Americans like Suttles, Wells, and Gibson were spending their winters in Cuba. During WWII the number of Americans drops off somewhat, presumably due to travel restrictions. On the whole, I see the league quality as pretty good during Dihigo’s career, especially during his best years.

Figueredo lists Cuban League leaders in various hitting and pitching categories; here are Dihigo’s appearances and also a few other notes. Check out 1935-36!

1924-25: Games pitched (20 - tied)
1927-28: Hits (54), MVP award
1928-29: All-time record for most 2B hit by a player in a game (4)
1935-36: Batting (.358), Runs (42), Hits (63 - tied with Willie Wells), Triples (8), RBI (38), Pitching (.846 11-2), Complete Games (13), Wins (11), Shutouts (4), MVP
1936-37: Games pitched (30); at end of season, won twice in consecutive days to force playoff, after losing game 1 of 3-game playoff, came back with two days' rest to win game 3; MVP
1938-39: Pitching (.875 14-2), Wins (14); All-time record for most consecutive seasons winning 10+ (4)
1941-42: Complete games (11 - tied), MVP (his fourth)
1943-44 Pitching (.889 8-1), ERA (2.23) -- Note - ERA statistics not available through most of his pitching career.

On the less statistical side, here is a description of Dihigo from The Pride of Havana: A History of Cuban Baseball by Roberto González Echevarría:

“Martín Magdaleno Dihigo Llanos was born in the sugarmill Jesús María (town of Cidra), in the province of Matanzas… on May 25, 1906. When he was around three the family moved to the Pueblo Nuevo neighborhood, in Matanzas proper, a couple of blocks away from the Palmar del Junco. Little is known about his childhood and adolescence, but it is clear that he began playing baseball very early, for by the time he was sixteen he had been signed by the professionals. Dihigo was a tall man for the period, about six-three, and played at around two hundred pounds when he filled out. He was relatively slender when he was young, but very powerful. Those who played with Dihigo or who saw him play say that he had it all: coordination, speed, power, arm, good hands, batting eye, intelligence, and poise. But he had something else difficult to describe, an intangible trait associated more with bullfighting, a sport in which there are no numerical records to evaluate an individual: He had grace. His blend of style, dignity, and elegance was apparently unique. Fausto Miranda told me that once he took a friend to see Dihigo at La Tropical stadium for the first time and the friend asked him to point out Dihigo to him when they got there. But Miranda said no, that his friend would be able to spot him just by the way he moved. And his friend did. Dihigo’s demeanor and style, which I imagine to have been like Roberto Clemente’s, are qualities much appreciated by Latin American fans.

“Dihigo also had leadership qualities, was personable, well-spoken, and had a sense of humor. With the years Dihigo developed into a sharp dresser who wore his suits well. In short, like his fellow matancero Méndez, he had class. He eventually became a successful manager and the first Cuban ballplayer to become a sportscaster after retiring. Dihigo was also endowed with a social and political conscience: He supported and aided his fellow players, and took a position favorable to the Cuban Revolution in its early years. He was rewarded with a position in the ministry of sports and was buried a hero when he passed away in Cienfuegos, on May 20 1971. His son, a sportswriter, is still in Cuba.”

By the way, in the table the "T" in "26/27(T)" stands for an alternative league called "Triangular" and the "U" in "30f(U)" stands for a league called "Unico," which played in the fall of 1930.

Was the Cuban League of the 20's and 30's much like the Dominican league of today, where palyers go down to play ball in addition to a full regular season in the States? Or were these players playing a good portion of their total games in Cuba (like maybe 75-80 games in the NeL, then the same in Cuba)?

Just wondering because if it is the former should we be giving players credit for these seasons? Are we giving them credit when they played in the NeL in the summer? We won't be giving credit to David Ortiz and Co. when they become eligible right?

Though at the same time if we have data from these seasons I guess we should use it as a 'replacement' for missing NeL, huh?

Here’s some comparative data on Dihigo’s Cuban pitching that I’d been meaning to pull together:

Cuban pitching comparisons

Martin Dihigo
Totals 231 g, 117 cg, 102-52, .662

Dolf Luque
Totals 216 g, 113 cg, 106-71 W-L, .599

Jose Mendez
60-14 official, 59-18 according to Holway
.811 or .766 winning percentage

Since these are basic w-l records, their meaning depends heavily upon teams and quality of competition, of course, but the team difference and competition difference would have to be quite large for Dihigo not to appear comparable to Luque as a pitcher.

It’s worth remembering that Mendez’s dominance was tremendous.

I haven't caught up with the whole thread yet so if someone else published something along these lines, I apologize for the redundancy.

jimd, you've looked into the effects of adding and subtracting star players to help explain WARP's league strength assessments -- is there a theoretical model we could look at for this?

The "back of the envelope" method I used is not that difficult to implement. It used three factors:

x1) the WARP value of the star being replaced
x2) the WARP value of the replacement player
x3) the total WARP value of the league

(x1-x2)/x3 gives the percentage. Say Ty Cobb was being replaced and measured 15 WARP-1. His replacement has 0 WARP (or 2 if you prefer). The total value of the league is 77*8 or 616. Effect is roughly 2.4% (or 2.1%). (WARP totals for the league are close to the number of league wins though teams do not obey this restriction, unlike Win Shares.)

Naturally, Win Shares can be substituted for WARP.

We have x1 in Win Shares for the star players. x2 and x3 may be difficult to estimate in those units. It might be easier to estimate true Negro League Win Shares for the lost stars rather than use the MLE Win Shares.

If so, then normalize the team schedule lengths to a common number of games played to estimate x3. x2 would be based upon the winning-percentage of the worst teams (e.g., .200-.250 in MLB yields 8-10 Win Shares on a 162 game schedule based on 12 starting positions).

The approximation will get even cruder if too many players are being replaced because league replacement level will be affected at some point.

That means roughly 5% of the early 30s AL offense is Ruth, Gehrig and Foxx, doesn't it? Worth bearing in mind that others OPS+ figures may be suppressed in that league for a few years there.

That means roughly 5% of the early 30s AL offense is Ruth, Gehrig and Foxx, doesn't it?

Actually, about 5% of the total league value, so it's probably more like 10% of the total offense, at least by Win Shares, since these guys were contributing relatively little on defense.

What do you think of the idea that Dihigo may have been the most valuable player in baseball during 1935-38?

With an OPS+ above 140, he would have been one of the top 15 hitters. During that period it looks like he was probably also one of the top 10 pitchers. I don't exactly know how one would add the two accomplishments together, but it seems to me that if a player is that close to the top in both hitting and pitching, there's got to be a case that he's the best player overall.

Chris's MLEs only cover 1935-36, but from his Mexican League and Cuban League records it looks like he stayed near that level for a couple more years.

Following up on my last post, here's how I think about it. Say, as a full-time pitcher Dihigo's worth maybe 25 win shares / season, and as a full-time outfielder maybe 28. What his teams were getting, though, was a full-time pitcher _plus_ a three-fourths time outfielder. So we might estimate his total win shares as 25 + (3/4)*28 = 46. That compares to the major league leaders:

1935: Vaughan 39
1936: Gehrig 38
1937: Medwick 40
1938: Ott 36

Gibson and Paige are in the mix too, but it seems to me that Dihigo has got to have a case as most valuable over that period.

Dihigo was a good player, but nothing like a full time ML pitcher in 1935-36, he went 7-3 in both years. You can add maybe 3/4 of his hitting value to half his pitching value and get 12.5 plus 21 = 33.5. Good peak, but not league-leading.

Sorry to be a pain, but can someone quickly type in his career batting line from Macmillan? Thanks.

Martin Dihigo

391 g
1363 ab
435 hits
52 2b
19 3b
62 hr
34 sb
.319 ba
.522 sa

Say, as a full-time pitcher Dihigo's worth maybe 25 win shares / season.

I think this estimate is a bit high, if one takes the MLE estimate I cooked up as valid: ERA+ 115-125, full-time starter but not a workhorse, for which I'd peg him at 200-210 innings in the NL of 1935 1936.

In 1935, that comps him most closely to three pitchers:

Watty Clark, 207 IP, 120 ERA+, 14 win shares
Van Lingle Mungo, 214 IP, 109 ERA+, 15 win shares
Charlie Root, 201 IP, 128 ERA+, 16 win shares

A pitcher with an average effectiveness and average innings pitched would be earning about 12 ws/season at this point in time. Dihigo was above average, but not twice as valuable as an average pitcher.

Chris Cobb wrote:

I think this estimate is a bit high, if one takes the MLE estimate I cooked up as valid: ERA+ 115-125, full-time starter but not a workhorse, for which I'd peg him at 200-210 innings in the NL of 1935 1936.

In 1935, that comps him most closely to three pitchers:

Watty Clark, 207 IP, 120 ERA+, 14 win shares
Van Lingle Mungo, 214 IP, 109 ERA+, 15 win shares
Charlie Root, 201 IP, 128 ERA+, 16 win shares

I’ll observe that if we accept Chris’s estimate of 15 win shares per year as a pitcher rather than 25, that still gives Dihigo about 36 total win shares per year. If he continued to perform at that level for 1937-38 (as the Mexican and Cuban statistics strongly suggest), I still think that makes him the best player in baseball for 1935-38.

However, there are two reasons why Chris’s estimate may be too low—one a matter of opinion, the other an omission in the calculation.

First, although it is Chris’s opinion that Dihigo would have pitched only 200-210 innings, I think the actual data suggest a heavier workload. 1) Chris (#50) observed that Dihigo pitched 19 to 23 percent of his team’s decisions. That would translate to about 230-275 innings in a 154-game season. 2) For the Cuban League, Figueredo provides the starting and ending dates for each season. For the 4 seasons beginning in the fall of 1935, Dihigo has 69 decisions in 60 weeks of play; for a 25-week major league season that translates to about 29 decisions (or 19 percent of a team’s decisions). So I think 235-245 IP per season is more consistent with Dihigo’s actual usage than 200-210. Substituting this target for the one used by Chris, I located the following comps for 1935:

Curt Davis, 231 IP, 124 ERA+, 21 win shares
Willis Hudlin, 232 IP, 122 ERA+, 21 win shares
Johnny Marcum, 243 IP, 111 ERA+, 19 win shares

The second issue is that as a pitcher with a 140 OPS+, Dihigo also would have earned batting win shares while pitching. (When I said he might have earned 25 win shares per season as a full-time pitcher, I was including the batting win shares earned while pitching.) How many batting win shares would a pitcher with an OPS+ of 140 earn? I found a few pitcher seasons in the general neighborhood:

Ruth 1916, 122 OPS+, 150 PA, 4.64 bWS
Ruth 1917, 162 OPS+, 142 PA, 6.01 bWS
Young 1903, 125 OPS+, 146 PA, 4.02 bWS
Johnson 1925, 163 OPS+, 107 PA, 4.99 bWS
Spahn 1958, 131 OPS+, 122 PA, 4.68 bWS

We should adjust these down a bit for Dihigo, since these pitchers generally pitched more innings than we are assuming for Dihigo (and they also did some pinch hitting). But I think it's reasonable to add 3 to 4 batting WS per season to whatever estimate is selected for Dihigo’s pitching WS.

So we can quibble about how many IP to project, but even with conservative assumptions, I see Dihigo with 39 to 40 win shares per season.

So I repeat, shouldn’t we be thinking of Dihigo as perhaps the most valuable player in baseball during 1935-38?

And if his peak really was that good, and he played for 23 years (ok, probably only 18 of them were major league quality), why shouldn’t he be ranked at least second on every voter’s ballot?

This has probably been gone over, but what quality of defender was Dihigo? What positions did he play early in his career? How good was he considered there? Where did he play later in his career (when not pitching), and how good was he there?

This is our toughest challenge since Frank Grant I imagine...lot of spottiness to his record, lot of unique things about him. The hitting is good enough to be on the ballot, but his position and quality of his defense are going to determine where he is on my ballot.

Michael,

See # 6 for Holway data, # 8 for detailed info on 1928, and # 25 for the Cuban League. He played every position, but seems to have spent more time in the infield in the 1920s, and in the outfield during the 1930s. What I've read about his fielding skills has been very positive, but I doubt that he would have been an A+ fielder at any position. I suspect he was probably more like Pete Rose - someone who could go out there and fill in wherever needed. But others here know much more than I do, and they should speak up.

So I think 235-245 IP per season is more consistent with Dihigo’s actual usage than 200-210.

I agree that this is more consistent with his actual usage. It's my view, though, that as the quality of competition that a pitcher faces rises, and their effectiveness diminishes in consequence, they are able to throw fewer innings, both because they have to face more batters per inning (more hits and walks given up) and because they have to pitch at closer to maximum effort more of the time. It's also the case that the offensive environment in the major leagues was higher than in the Negro Leagues, which would also place more strain on pitchers.

For that reason, I think 200-210 innings is a reasonable estimate of ML innings for a pitcher who would be throwing 240-250 innings in the NeL.

There's disagreement on this point, of course, as we saw when we discussed MLEs for Willie Foster a while back, but this is the assumption I was working on when I made my estimate of Dihigo's MLE innings in 1935 and 1936.

Current Negro League Backlog

1. Dihigo
2. Suttles (career) or D. Moore (peak)
3. Moore or Suttles
4. Lundy
5. Beckwith (bat) or Bell (complete)
6. Bell or Beckwith
7. Monroe
8. Redding
9. Mendez
10. Mackey

11. Chino Smith (whattapeak!)
12. Taylor or Poles (probably should be ahead of Chino but whattapeak!)
13. Poles or Taylor
14. A pitcher but which one--Cooper? Donaldson? Winters? I'll just take them alphabetically--Cooper
15. A 2B but which one--Allen or Scales? I like Allen
16. Donaldson
17. Scales
18. Winters
19. Oms
20. Judy Johnson

21. Carr
22. Petway
23. DeMoss
24. Heavy Johnson
25. Fats Jenkins (higher if you count his hoops)

I would urge everybody to do this, and you will discover (I think) that you (I) don't know a damn thing about any of these pitchers after Redding and Mendez. I'm sure there should be 2-3 more pitchers in the top 25. The rest, I'm OK. But the pitchers? Elect the best, ignore the rest has certainly been my motto.

I'd love to hear from the Dean/Gomez crowd what seperates Dean/Gomez from Nip Winters...other than our collective fogginess on Winters's record. He basically looks like the NgL answer to Diz.

On the 1950 Ballot thread # 93, John Murphy wrote:

Just to be clear, I have Dihigo behind Cronin, but Beckwith ahead of the AL President.

I do believe, in my gut, that I am shortchanging him, though, and mentioned as much on my ballot. His career was so weird (like John Ward's) that I need a little more time to get a handle on him.

...When the stats are somewhat sketchy, people will pause before making a commitment to vote for a particular player. Anecdotes alone won't cut it by themselves.

On this thread, # 48, Gary A wrote:

One thing about Dihigo: his career is actually more extensively documented than any other NeL career to this point. Counting up everything from the NeL and the Cuban, Mexican, and Dominican leagues (as well as the 49 at bats Holway gives him against major leaguers), we have 6,119 at bats (1,865 hits for a .305 average, and 142 home runs) as well as a 261-143 record as a pitcher (.646). And we don't even have any information from Venezuela, where he played when he was 26-28 years old.

In Dihigo's case, in addition to the prejudices that David C. Jones alluded to on the 1950 ballot thread # 70, 87, 91, and 92, we also have another type of prejudice. Dihigo's contributions are just so different from anything most of us are familiar with that it becomes difficult to comprehend then and they get discounted.

I mean, just look at his 1938 season in the Mexican League--Dihigo won the batting title with .387/.482/.599 and led the league in pitching, 18-2/0.92 ERA/184 SO and 32 BB in 167 IP. Well that was probably a poor quality league, so we just ignore it.

Except he did almost the same thing in Cuba in 1935-36--led the league in batting (.358), runs, hits, triples, RBI, pitching (.846 11-2), complete games, wins, and shutouts, and (by the way) was elected MVP. But even though he was playing against several major leaguers and top NeL players in Cuba, we don't know what the appropriate quality factor was, so we just ignore that too.

With Jud Wilson, I think when we were able to look at his MLEs and say he looked sort of like Boggs, but more aggressive on the basepaths, I think that helped a lot of voters form a picture of him and feel comfortable voting for him. If Dihigo had just stayed in right field, maybe we could all say he was Clemente with better strike zone judgment and he'd sail into the HoM. But what do we make of someone with all of the diverse talents Dihigo actually exhibited during his career?

In # 110, 111, and 116 I've put forward a case that Dihigo was arguably the most valuable player in baseball in the mid to late 1930s, and so far haven't seen any arguments to convince me otherwise. (My only hesitation is that I haven't yet looked closely at Paige or Gibson for that period.) In the context in which Dihigo was playing, his diversity of skills was just incredibly valuable. I'm a bit disappointed that more voters don't seem to recognize this.

Although I wouldn't give Dihigo as many win shares as Brent has proposed, I should say that I agree with him that Dihigo was quite possibly the most valuable player in baseball in the late 1930s. It's hard to say exactly what the MeL numbers mean, but they're obviously good, and the Cuban numbers for 1937-39 don't give us any reason to suspect that Dihigo was in decline. He appears to begin to decline around 1940, but the statistics suggest a solid peak from 1935-39.

The NeL statistics, of whose meaning we have a better sense, can give us a look at the height of his peak.

Based on the pitching estimates I've made, the batting and playing time MLEs, and giving Dihigo credit as a straight B outfielder, here's what Dihigo looks like in WS for 1935 and 1936, figuring 32 pitching starts and 90 outfield starts (I figure a game of rest after each pitching appearance at major-league standard).

Year--BWS--FWS--PWS = Total
1935--17.7--2.3--16 = 36
1936--24.3--2.3--17 = 43

I think it's almost certain that Dihigo was the top player in baseball in 1936, and he was probably the most valuable player in 1935-36 combined (top ML WS over the two years are Vaughn 74, Gehrig 72, Ott, 71). It seems very likely that his next two and possibly 3 years would have been worth at least 30 win shares each.

Dihigo Positions:

In 1935, for the all-star game selection, he is listed as a RF.

For 1936, in the few boxscores I have, he's playing 2B and batting 5th.

Wow - that is one of the best threads I've read since we started this project.

I'm still sure where I'll rank him, but the discussion here has been phenomenal. Great job guys.

Placement of Dihigo's stateside pitching on the career leaderboards:

WINS t-103rd with 29 wins

LOSSES t-107th with 26 losses

DECISIONS t-102 with 55 decisions

WINNING PCT .527
(50+ decisions minimum) 74th
(25+ decisions minimum) 111th
(10+ decisions minimum) 169th

ADJ PCT OF TEAM DECISIONS 11.8%
(50+ decisions minimum) 113th
(25+ decisions minimum) 173th
(10+ decisions minimum) 255th

WAT 92nd with 2.2

WAT PER DECISION .04
(50+ decisions minimum) 49th
(25+ decisions minimum) 73th
(10+ decisions minimum) 113th

PLACEMENTS ON YEARLY WINS LEADERBOARDS

1935 t-7th in EWL with 7.

1936 t-4th most wins in NNL with 7.

Here's a little more on Dihigo's pitching in Mexico.

1940
Pitched in hard luck. -.4 WAT despite ERA that was 20% less than team total and 25% better than league. Peripherals all suggest a better record than he got stuck with.

1941
No data on him.

1942
Huge year, 22-7, 6 WAT with a 2.53 ERA in a 4.38 league, on a 4.37 team that went 48-40.

1943
A very good year. 16-8 with 2 WAT and an ERA of 3.10. But team ERA was 3.22. League was 3.90.

1944
12-10, .5 WAT, 3.14 ERA for a 4.09 ERA team. I haven't compiled the league figures yet.

1945
Haven't compiled figures for the league or him.

1946
11-4, 3.3 WAT, 2.83 ERA for a 4.18 ERA team in a 3.94 ERA league.

1947
The last season I have data on him. 4-2, 1.4 WAT, 4.37 ERA for a 4.72 ERA team in a 3.96 ERA league.

All told, it's a very interesting and positive career in Mexico, probably better than I had personally credited him with initially.

Hang on, I looked back, and it turns out that I'd simply missed Dihigo in my spreadsheet in 1941.

1941
9-10, .5 WAT, 4.10 ERA for a 5.08 ERA team in a 4.77 ERA league.