Is Cooper a Rixey? A W. Cooper? An Andy Petitte?

He's pretty close to a Nip Winters, with a few years longer career, but probably not as good of a peak....

1928 Pitching Andy Cooper:

W-12
L-6
R/G - 4.16
G-21
St-16
CG-9
Sho-2
INN - 127 2/3
BFP-508
H-120
HR-2
K-52
W-16
HB-4
WP-1

The 1928 Monarchs really had a great staff, in addition to Cooper they had:
William Bell, 9-7, 3.56 R/G, 147 Inn
Chet Brewer, 6-8, 4.79 R/G, 120 Inn
Bullet Rogan, 10-2, 3.71 R/G, 114 Inn

Might as well chip in Cooper's 1923 record with Detroit, from Patrick Rock's research. He was as close to a relief specialist as the Negro Leagues got that season:

*-led league
W-15 (2nd in league to Rogan's 16)
L-7
SV-6*
W Pct- .682 (team 41-30, .577)
ERA-3.69
G-36*
GS-19
CG-12
SHO-1
IP-183
H-168
HR-12
W-44
K-68

Here's more on Andy Cooper

Andy Cooper data
From Holway

1920 1-0 for Det; team 40-21; not in reg. rotation (MLE)**
1921 5-8 for Det; team 37-42; not in reg. rotation (MLE); –1.3 wins above team (WAT)*
1922 16-6 for Det; team 57-41; #3 starter (MLE); 4.1 WAT; 3.76 TRA 2nd; wp 3rd
1923 16-8 for Det; team 48-33; #1 starter (MLE); 2.5 WAT; 60 K 5th, wins 3rd, wp 5th; all-star
1924 14-6 for Det; team 38-33; #1 starter (MLE); 4.6 WAT; 76 K 5th; all-star
1925 11-3 for Det; team 55-39; #3 starter (MLE); 3.3 WAT; wp 2nd (also listed as 13-2), TRA 3.36 5th; 44 K 5th
1926 12-8 for Det; team 51-44; #2 starter (MLE); 1.6 WAT
1927 8-3 for Det; team 70-53; not in reg. Rotation (MLE) 1.9 WAT
1928 13-7 for KC; team 49-32 ; #2 starter; 1.2 WAT; all-star
1929 13-3 for KC; team 66-14; #4 starter (MLE); -0.2 WAT; 2nd in wp; TRA 3.61 4th; all-star
1930 15-6 for Detroit Stars; team 56-37; #2 starter (MLE); 3.0 WAT); wins 4th; all-star
*A Note in Holway says that Cooper hurt his arm just before the playoff between Detroit and St. Louis; arm trouble may explain the lack of data for the next two seasons.
1931 0-1 for KC Monarchs
1932 no data (Riley lists him with Monarchs)
1933 pitched for Monarchs; no data as Monarchs barnstorm
1934 pitched for Monarchs; no data as Monarchs barnstorm
1935 pitched for Monarchs; no data as Monarchs barnstorm
1936 2-0 for KC Monarchs; also manager. Team 7-0
1937 2-0 for KC Monarchs; also manager. Team 13-8
pitched 17 innings in a game that ended in a tie in playoff vs. Chicago. He gave up 2 runs in the first and then threw 16 scoreless innings. Foster gave up 2 in the seventh and left the game when he was hit in the chest by a line drive off the bat of Joe Rogan.
1938 2-0 for KC Monarchs; also manager. Team 32-15

*Wins above team have become an important part of my evaluation of NeL pitchers, so I've included the numbers in the data for Cooper; I'll post comparative data on this and on black and gray ink for Cooper, Winters, and Rogan.

**These #1, #2 starter designations are my estimates of the pitcher's major-league equivalent workload, based on number of decisions in NeL play in relation to other NeL pitchers and the number of games recorded for the team. I use them as the starting point for creating ip estimates. Looking at 1923 and 1928, I think my estimates are pretty good. Cooper was not a workhorse on the model of Winters or Rogan -- in 1923 it looks like he was being used the way the Cubs used Three-Finger Brown at his best -- but I think he would sometimes have been throwing the most innings on his team in the majors, but generally have been in the 190-240 innings range.

Riley data of note
1936 27-8 vs. all levels of competition

(this season gives a snapshot of the number of games barnstorming Negro-League teams were playing during these years, even though they are often undocumented)

Career
121-54, .691 according to Holway’s totals
130-58, .691 by adding up the seasonal data

Chris,

Cooper is around 20-21 WAT by your numbers. Do have any sense of how that compares to other NgL pitching candidates?

Thanks!

I don't have that data yet, but I'll be working on compiling it soon.

Wins above team, career

Joe Williams 30.5 or more (a conservative estimate, don’t have the hard data yett)
Andy Cooper 20.7
Dick Redding 19.9
Nip Winters 15.0
Joe Rogan 9.9



Nip Winters Wins above team by season

1921 0.3 WAT; 3-3 for 34-28
1922 –1.2 WAT; 7-10 for 20-24
1923 2.0 WAT; 10-3 for 32-17
1924 7.8 WAT; 27-4 for 58-23
1925 –5.2 WAT; 21-13 for 65-26
1926 8.5 WAT; 23-4 for 58-34
1927 6.1 WAT; 18-16 for 47-70
1928 0.1 WAT; 10-11 for 26-27, 17-21
1929 –2.3 WAT; 2-4 for 38-18
1930 –
1931 –0.6 WAT; 1-2 for 44-16, 1-1 for 3-9; 0-2 for 0-4
1932 0.1 WAT; 1-2 for 16-35


Joe Rogan Wins above team by season

1920 2.1 WAT; 10-4 for 45-31
1921 1.9 WAT; 20-11 for 54-35
1922 –0.4 WAT; 20-11 for 77-37
1923 –5.7 WAT; 20-19 for 78-49
1924 2.4 WAT; 17-5 for 60-27
1925 4.5 WAT; 20-2 for 63-20
1926 –0.1 WAT; 14-4 for 65-19
1927 2.1 WAT; 15-6 for 59-33
1928 3.1 WAT; 11-3 for 49-32

Andy Cooper Wins above team by season

1921 –1.3 WAT; 5-8 for 37-42
1922 4.1 WAT; 16-6 for 57-41
1923 2.5 WAT; 16-8 for 48-33
1924 4.6 WAT; 14-6 for 38-33
1925 3.3 WAT; 11-3 for 55-39
1926 1.6 WAT ;12-8 for 51-44
1927 1.9 WAT; 8-3 for 70-53
1928 1.2 WAT; 13-7 for 49-32
1929 –0.2 WAT; 13-3 for 66-14
1930 3.0 WAT; 15-6 for 56-37

No season-by-season breakouts for Redding

Brief Analysis. This statistic, by itself, is interesting but not conclusive. Because competition was so uneven in the NeL, we have to ask how good teams were, and how good the other pitchers on teams were, in order to make judgments. Rogan has the lowest total, but he pitched on teams that were always very good in part because they had great pitching staffs. KJOK has highlighted the quality of their 1928 staff; that was not unusually good for the Monarchs in this decade. Winters occasionally had great pitching teammates and sometimes pitched for great teams; Cooper generally pitched for good teams, but except for his two KC seasons (28-29), his staffmates were not as distinguished. Redding’s teams were often not good. Williams’ teams were, for his career, about average without him.

Very interesting stuff. For Redding you might find the information I posted about his Cuban career useful (once again, he played on generally poor teams in Cuba).

I do have reservations about that 20-19 record Holway has for Rogan in 1923. Patrick Rock's research on that season is *very* thorough--I believe he was able to locate all but a small handful of box scores, mostly the second games of double headers. If you actually see his yearbook for Replay Baseball, you'll know what I mean--it's very detailed and credible. Anyway, he's got Rogan going 16-11, the Monarchs 61-37. If I'm figuring it correctly (a big if), that would make Rogan's WAT -1.1 instead of -5.7, putting his total at 14.5, very close to Winters.

I suppose it could be that Holway counts games Patrick doesn't. I'll email Patrick to ask him about it.

Gary,

It'd be great if you could get an explanation for the differences between the two sets of numbers.

The 16-11 does strike me as more likely.

I should add, however, that I infer that one reason Rogan is fewer WAT that year is that he was used very heavily against the Monarchs' better opponents. If he wasn't pitching so well that year, why else would he have so many starts & decisions? I don't suppose Patrick's documentation goes so far as to list who Rogan pitched against?

Actually, he includes a schedule with scores and starting pitchers. I can go through it and let you know who he started against.

I can also derive park factors from it (would take a little longer).

Knowing Rogan's opponents would be splendid!

Park factors would be excellent, too, more for general value than for the immediate question of evaluating pitchers.

OK--Rogan started 24 games against these teams, with team's winning pct. followed by Monarchs' record against them in Rogan's starts:

8 Chicago American Giants (.623), 4-4
4 Indianapolis ABCs (.607), 2-2
4 St. Louis Stars (.388), 4-0
3 Detroit Stars (.577), 3-0
3 Cuban Stars (West) (.435), 1-2
1 Birmingham Black Barons (.395), 1-0
1 Milwaukee Bears (.246), 0-1

Fifteen of 24 starts were against teams that finished over .500. Eight, or one-third, were against the American Giants, the second-place finishers and winners of the three previous NNL pennants.

The weighted winning percentage of Rogan's opposing teams is .527--it's .466 for the Monarchs' other pitchers.

In Rogan's 24 starts, the Monarchs were 15-9. They scored 5.71 runs per game behind him, and allowed 3.75.

Home/road breakdowns: W-L, R-OR, R/G-OR/G

road: 9-3, 85-44, 7.08-3.67
home, Association Park: 2-5, 23-27, 3.29-3.86
home, Meuhlebach Field: 4-1, 29-19, 5.80-3.80

Rogan's 24 starts came against 17 opposing starting pitchers. He opposed John Finner, of Milwaukee and St. Louis, four times, and Ed Rile, the American Giants' best pitcher, three times.

Twelve of the 17 pitchers finished the season over .500. If you add up their individual W-L records, it comes out to 115-81, .587; if you weight them according to their number of starts against Rogan, the winning pct. gets lowered slightly to .561, mostly because of Finner's starts (he was 4-8 overall).

Hope all that makes sense. In sum, Rogan did tend to start against the better teams and their better pitchers. He started 8 of the Monarchs' 22 games against the American Giants, but only 1 of their 15 games against the last-place Milwaukee Bears.

1921 Andy Cooper
NNL Detroit Stars

W-4
L-6
SV-0
TRA-5.44 (NeL 5.20)
G-16
GS-9
CG-4
SHO-1
IP-82.7
H-97
HR-5
R-50
BB-11
K-28
HP-6
SH-10
SB-15
DP-5
OOAVE-.293 (NeL .263)
OOBA-.328 (NeL .324)
OSLG-.420 (NeL .361)

Cooper's rookie season--just thought I'd throw it in, for what it's worth.

Some preliminary park factors for 1921. These are simple park factors, just total runs in a team's home games divided by total runs in their road games (which include neutral site contests):

Indianapolis (Northwestern Park): 104
Bacharach Gts (Inlet Park): 97
Chicago (Schorling Park): 55
Columbus (Neil Park): 120
Cincinnati (Redlands Field): 111
Cleveland (unknown): 110
Detroit (Mack Park): 111
Kansas City (Association Park): 115
Pittsburgh (unknown): 75
Hilldale (Hilldale Park): 105
St. Louis (Giants' Park): 104

Joe Green's Chicago Giants were a travelling team.

It does seem weirdly dominated by hitters' parks, but this is partly because of very unbalanced schedules. Neutral-site games generally seemed to have lower scores, which also works to slightly inflate park factors.

Yes, the Chicago number is accurate. Here are the breakdowns:
Home 34-7, 168 runs, 96 runs allowed--6.44 total runs/game
Road/Neutral 17-21-4, 240 runs, 255 runs allowed--11.79 total runs/game

Maybe Bingo DeMoss wasn't that bad after all...

Obviously, we need a lot more information to really make sense of most of this. I'll type in 1923 scores next, and I can also do 1920 and 1922 without too much trouble. That'll give us four straight seasons.

Chris, will you be doing an estimated record analysis on Cooper, it'd be a big help with him, for comparing him to Redding, Mendez, et al. Even a rough guess would be nice . . . I really don't have any idea of what to do with him right now.

I will be doing an estimated record analysis of Andy Cooper, but I doubt I will be able to get to it before December 25. The extra week of discussion for 1942 should help me get caught up with MLEs for the NeL players and get ready for the steady stream of major NeL candidates that's on the horizon.

Next on my list are MLEs and win shares for Cooper, win shares for Rogan, and revised win shares for Beckwith and Moore, with MLEs and win shares for Charleston, Bill Foster, Lundy, and Judy Johnson in progress for 1943. As Gary A. gets us more information about parks and league offense levels, I'll try to keep my estimates up to date with the best available data.

I know it goes without saying, but this work couldn't be done right without Chris and Gary and KJOK all offering information and expertise on a regular basis.

Thank you guys so much for everything you do to help the rest of us make better decisions about men who it would otherwise be insanely difficult to find good quality information on.

Thank you guys so much for everything you do to help the rest of us make better decisions about men who it would otherwise be insanely difficult to find good quality information on.

I'll second that. Thanks guys!

I second that. Fellow HOM voters may think I'm simplistic anyway, but I would be COMPLETELY throwing darts against a wall on Negro leaguers without your help -- since I have a naturally suspicious nature, this would result in Josh Gibson and Oscar Charleston getting in my PHOM and nobody else!

I know it goes without saying, but this work couldn't be done right without Chris and Gary and KJOK all offering information and expertise on a regular basis.

I think you should say Chris, Gary and Gadfly, with some limited knowledge from KJOK, mostly stats passed on from Gary!

THANKS

I'll second that THANKS to Gary and Gadfly, who are making so much useful information available to us.

I just post what's available in public sources and muck about with the numbers a bit.

I'm just glad the Negro Leagues are getting serious attention here.

OK, I finally got park factors from Patrick Rock's 1923 NNL data. These are for NNL teams, plus three associate teams (Cleveland, Birmingham, and Memphis):

Indianapolis (Washington Park): 85
Chicago (Schorling Park): 81
Detroit (Mack Park): 79
Kansas City (Association Park): 113
Kansas City (Muehlebach Field): 104
Kansas City (overall): 109
Milwaukee (Borchert Field): 88
St. Louis (Stars Park): 155
Toledo (Swayne Field): 89
Cleveland (Tate Field): 108
Birmingham (Rickwood Field): 130

The last three are based on small samples.

The Memphis Red Sox' Lewis Park doesn't have a park factor, because the Red Sox played no road games against top black teams. However, in Memphis games teams averaged a total of 6.37 runs, compared to 10.95 for all NNL games.

ANDY COOPER'S PLACEMENT ON CAREER LEADERBOARDS (per Holway's data with some help from Gary A.)

WINS 7th with 130

LOSSES 23rd with 59

DECISIONS 10th with 189

WINNING PCT .688
(50+ decisions) 10th
(25+ decisions) 18th
(10+ decisions) 33rd

ADJ PCT OF TEAM DECISIONS 17.8%
(50+ decisions) 67th
(25+ decisions) 89th
(10+ decisions) 126th

WAT 2nd with 21.8

WAT PER DECISION .115
(50+ decisions) 11th
(25+ decisions) 24th
(10+ decisions) 45th

ANDY COOPER'S PLACEMENT ON YEARLY WINS LEADERBOARDS

1922 t-4th in NNL with 16 wins.

1923 Led NNL and NgLs with 16 wins.

1924 t-2nd in NNL with 14 wins, t-5th in NgLs.

1925 t-4th most wins in NNL with 11, t-8th in NgLs.

1926 t-4th most wins in NNL with 12, t-6th most wins in NgLs.

1928 t-4th most wins in NNL with 13, t-4th in NgLs.

1929 5th most wins in NNL with 13, 6th most wins in NgLs.

1930 t-3rd most wins in NNL and NgL with 15

Andy Cooper's OPP+ and hOPP+
(for more info on OPP+ and hOPP+ see the re-evaluating NgL pitchers thread)

In 11 of the 15 seasons he spent in the NgLs, Andy Cooper had an OPP+ above 100. For his career:

Cooper's teams had 1061 decisions.
Average teams in his leagues had 927.77.

Career OPP+ of 114.

Cooper's teams had 1061 decisions.
An historically average team would have had 765.

Career hOPP+ of 139. That's actually ten points higher than Nip Winters's.

I'm hoping someday to do OPP-neutral W-L records and WAT, but I'm not seeing it happening right away.

ANDY COOPER MLEs (ROUGH DRAFT)

I’ve never done an MLE for a pitcher, and having now tried a couple, I can tell you that there’s a lot of alchemy involved due to the crazy spread of variables. So here’s a rough try at Andy Cooper’s MLEs. Hopefully our experts will weigh in and offer some feedback.

First I’ll tell you what I did to in the interest of full disclosure. I based my methodology on Chris’s work as described in the Winters and Foster threads, but I veered off his course a bit because I didn’t feel very comfortable assessing the pitcher’s team’s offense and defense. Therefore, these numbers are not adjusted for RSI, nor for DEF as Chris’ are.

Also, in looking for a best-fit method, I tried a bunch of different methods, not really knowing which was most appropriate. This one is the one that seemed to work “the best.”

In each instance, use the methodology on a season-by-season basis.

MLE METHODOLOGY

Step 1: Assuming league-average offensive support in any given season, multiply pitcher’s number of decisions by 4.5 to estimate the total RS in his games.

Step 2: Assuming he threw mostly complete games in the Negro Leagues, estimate the total RA in his games by using Pythagorean formula (I used an exponent of 2 to keep things simple) with 4.5 as the RS and matching his actual winning pct.

Step 3: Assuming a 2-1 pitching–defense breakdown in run responsibility, estimate his contribution to his total RA by multiplying the estimated RA by .67.

Step 4: Convert the figure from step three to a major-league normalized figure by using the .85 conversion rate previously used by Chris Cobb.

Step 5: Assuming league-average defensive support, add 1.5 runs to the result of step four. This (I hope) represents the pitcher’s MLE DERA.

Step 6: Use Pythagorean formula to determine his MLE winning percentage, using 4.5 for the RS total and the DERA for the RA total, with exponent 2.

Step 7: Estimate the pitcher’s decisions by multiplying his MLE innings times 8.7.

Step 8: Apply the result of number 6 to the decisions in number 8 to figure W-L record.



NOTES ON INNINGS ESTIMATION
This was kind of tricky. I looked at three things, Cooper’s rotation slot viz and MLB starter in the same slot, how he fared among NNL pitchers in decisions, and what the aging pattern of his MLB contemporaries (born within three years of him) looked like. FYI: Cooper pitched in the Negro Leagues from age 24 to age 43.

In the latter half of his career, Cooper appears to have had that peculiar every-other-year quality to him where he’d put up a big year in year 1, then his workload would diminish in year two, only to surge again in year three.

The Problem of 1931–1938
The problem is a paucity of recorded decisions. Cooper’s rise, peak, a good chunk of his extended prime are well-documented by Holway, but thereafter things fall apart quickly. Cooper’s total reported decisions for 1931–1938 total seven. The Monarchs of this period don’t total too many more, however. In 1931, Cooper took just one of the Monarch’s twenty-seven league decisions, perhaps indicating an injury or sudden decline in effectiveness (or simply a change in role). But countermanding that is that Cooper’s two reported decisions in 1936 representing 29% of the Monarch’s seven decisions. He quickly falls away as they increase their decisions in the next two seasons (two decisions each in twenty-one and fifty games in 1937 and 1938 respectively), indicating that, perhaps, these are his real swan-song seasons.

I don’t know the answer here, but I suspect that Cooper was pitching in 1931–1936 as his team barnstormed, but that his workload was dropping from its heights in the 1920s. The pattern established by the eleven contemporary MLB pitchers I studied showed that in late career, guys’ workloads would suddenly drop in their late thirties and plateau at a new, lower workload level for two to four years before finally dropping off altogether around age 40.

So for 1931—when the Monarchs played roughly half the games of most of the other NNL teams—I decided that I would base his usage partly on the documented usage and partly on a hunch that the Monarchs were probably still pitching him in non-league games as the NNL collapsed around them. I could be wrong, please correct me if so.
So he basically gets half a year in rotation. For his effectiveness, I averaged his previous two MLE DERAs and raised them by 5% to account for age-related deterioration in skills.

1932–1935, the Monarchs barnstormed, so I pretty much just took a whack at it, using the age-based contemporaries as a guideline for workload, as well as the every-other-year pattern of Lefty’s career. For effectiveness, I made him more effective during the larger innings years (as was the case in the rest of his career). This pattern starts in 1932 and I adjusted for age by increasing his DERA by 5% over the his 1931 DERA. In 1933, I raised his DERA by 10% over 1932. Then in 1934, when he again has a larger workload, I raised it 5% over 1932. Finally in 1935, his DERA rises by 10% over his 1934 DERA to signal it’s getting close to the end.

1936–1938, since Cooper started 2 of the Monarchs’ 7 league-level games in 1936, I made this season, his age 40 season, his last big year workload wise, with a DERA 5% higher than his 1934 DERA. In 1937 and 1938 he falls away to oblivion with his DERA in 1937 rising 10% over 1935 and 1938 rising 10% over 1937.

I know it might seem funny that I didn’t diminish his effectiveness year over year, but as I thought about it, older pitchers often don’t age linearly, they move along in fits and stops, losing a little here, adjusting and rallying, then losing some more, until they just can’t adjust enough. Hopefully that model makes sense to everyone.

Without further ado, here’s the Andy Cooper MLEs.

     MLE   MLE     EST     MLE   MLE
YEAR DERA  PCT     INN     WINS  LOSSES
---------------------------------------
1920 1.500 0.900   35      3.6    0.4
1921 6.037 0.357   175     7.2   12.9
1922 3.696 0.597   250     17.2  11.6
1923 4.036 0.554   300     19.1  15.4
1924 3.848 0.578   285     18.9  13.8
1925 3.373 0.640   250     18.4  10.3
1926 4.429 0.508   275     16.1  15.6
1927 3.696 0.597   175     12.0   8.1
1928 4.132 0.543   275     17.1  14.5
1929 3.223 0.661   180     13.7   7.0
1930 3.768 0.588   245     16.6  11.6
1931 3.675 0.600   125      8.6   5.7
1932 3.860 0.576   165     10.9   8.0
1933 4.250 0.529   140      8.5   7.6
1934 4.053 0.552   170     10.8   8.8
1935 4.458 0.505   125      7.3   7.1
1936 4.260 0.527   200     12.1  10.9
1937 4.900 0.458    75      3.9   4.7
1938 5.390 0.411    25      1.2   1.7
=======================================
TOTALS            3470    223.2 175.7
                       WIN PCT  0.560

Please everyone let me know what you think and whether the innings are reasonable and whether this MLE method seems like it’s working OK.

If Cooper only pitches in 7 League Monarchs games from 1931 and was 40 in 1936, I think this assumes too high a usage level in the MLE, and thus too high a number of decisions/wins. Had he been a major league pitcher, the 1920s figures may well be about right, but continuing to 42 was not something most pitchers did once they started losing effectivesss. This makes him Clemens/Spahn, which few were.

Well, he went 15-6 in 1930 and got 2 of 7 decisions in 1936, so I'm not sure there's any evidence that his usage seriously declined until 1937. We think he was pitching regularly for the barnstorming Monarchs 1931-35 at least. See post #5 for his NeL numbers.

And it seems to me we've seen plenty of sore arms, etc., in the NeLs, plenty of short careers. So when a NeLer is able to have a nice long career, maybe that's real. If he coulda pitched in the MLs in the '20s, well, then maybe he coulda continued to do so in the '30s too.

Besides, you're forgetting the Jack Quinns and Jack Powells. Not everybody pitched at the level of Spahn and Clemens into their 40s but was it so rare in the MLs or so common in the NeLs that we can say with any confidence that Cooper's longevity is only an artifact of weak NeL competition? I don't know.

Even then his case doesn't depend a whole lot on the 1930s, at least not to me as a peak/prime voter. Even throw 1921 in there and he has a 10 year record of 124-58 or MLE 156-119, but then pitched regularly for another 5 years at the very least.

Years NeL WL MLE WL

1920-21 6-8 11-13
1922-30 118-50 149-106
1931-38 2-1 59-48
1937-38 4-0 5-6

The real question, I think, is not the number of decisions but the WL pct. that a good conversion rate and a normalized team environment would give. Out of all the extra decisions you get by normalizing to 154, Doc only has Cooper winning about 33 percent. That doesn't make it the right conversion but it seems to make it not grossly generous.

It's the 1931-38 conversion of 2-1 to 59-48 I have a problem with. When Chris gets back next week, I'd be very interested in his view; not at all to say that he's "more" of an expert, but that to have one set of MLEs has the virtue of consistency, and the issue of whether very long NEL careers can be translated into the ML is one he's thought about a lot.

On the figures above Cooper's within my consideration set but not quite on the ballot; if the 59-48 should in fact be (say) 30-24 then his record becomes 194-154 in 2962 innings, which drops him off the bottom unless his MLE ERA+ is very exciting. Dizzy Dean is not on my ballot.

Karl, I'm neither certain nor uncertain about whether Cooper would or wouldn't have pitched into his 40s in the majors. The evidence is completely mixed, and I concur that I look forward to Chris Cobb's analysis and experience-honed intuition since this is my first time trying this.

As fodder for conversation, through 2001, the SBE lists the following 40 year olds as pitching 200 innings in their age-40 season:

INNINGS PITCHED                 IP     
1    Cy Young                  343.1   
2    Phil Niekro               342     
3    Grover C Alexander        268     
4    Warren Spahn              263     
5    Charlie Hough             252     
6    Tom Seaver                238.2   
7    Early Wynn                237     
8    Eddie Plank               235.2   
9    Tommy John                234.2   
10   Red Faber                 234     
11   Gaylord Perry             232.2   
12   Jack Quinn                229     
13   Don Sutton                226     
14   Joe Niekro                225.1   
15   Dazzy Vance               218.2   
16   Nolan Ryan                211.2   
17   Rick Reuschel             208.1   
18   Johnny Niggeling          206     
19   Bob Smith                 203     
20   Tom Candiotti             201 

Since then, Randy Johnson, David Wells, Jamie Moyer, and Roger Clemens have also turned the trick. In addition, Kaiser Wilhelm threw 243 innings in the Feds in 1914.

Dr. C, that list is about what I would have expected; of approximately 3 million innings pitched in the majors since their inception, about 4800 (0.16%) have been pitched by your 200-inning 40 year olds. It's not impossible, but it's not what you'd call common.

Dr. C, that list is about what I would have expected; of approximately 3 million innings pitched in the majors since their inception, about 4800 (0.16%) have been pitched by your 200-inning 40 year olds. It's not impossible, but it's not what you'd call common.

karl, the conversion from 2-1 to 59-48 would certainly be questionable if indeed Cooper had 3 decisions in 8 years. What he had, however, was 3 decisions in games against other NeL teams--the point being that Cooper and his team did not play in the NeL those years. I wish we knew what Cooper actually did during those years--30-40 decisions or maybe hundreds?--but he certainly did not pitch 3 games.

PS. Was he 42 years old in 1938? I haven't seen a DOB.

If he was, then he was 24 in 1920. Maybe he coulda/woulda pitched elite ball prior to age 24?

But anyway, wipe out 200 IP and a 12-11 record at age 40 and what do you have? Not a significantly diminished candidate.

Although I don't have Riley at hand, IIRC, Cooper was listed with a birthdate of March, 1896. He died in 1941.

I don't remember what Riley said about Cooper's pre-league days, but I'll check it out when I get back home.

Just spotted something in the Chet Brewer thread worth noting:

Posted by Gary A on May 18, 2005 at 07:21 PM (#1346911)I should say that his 1934 record represents four of the few games the Monarchs played with top NeL teams--they mostly barnstormed in Canada and the Pacific Northwest with the House of David.

Which is to say that if this example is reflective of the period 1931-1937 for the Monarchs, our projects will be speculative by necessity.

>He died in 1941.

Well, then I think those MLE IP in 1942 are pretty questionable! ;-)

FWIW, Riley says nothing about Cooper prior to 1920, but he does mention that in the winter of 1934-1935 Cooper was part of the team touring the orient (which included Biz Mackey as we earlier discovered).

Between Riley and Holway I've been able to come up with a few more little things about the Monarchs and Cooper in this period.

-In 1932 KC apparently folded and reorganized. They played in the Southern League garnering 14 decisions. Cooper had none of them.

-Chet Brewer was a Monarch during this time, and in his notes, Riley says that Chet won 30+ games against all levels of competition in 1933 and 1934.

-Cooper, however, is credited by Riley with a 27-8 record in 1936 against all levels of opposition.

Hopefully this sheds a little more light on how often the team was playing.

Based on Dr. Chaleeko's MLEs, here are a couple of additional calculations:

Andy Cooper's career MLE DERA was approximately 4.03 (in 3470 innings).

A couple of major league pitchers with similar career DERA and innings pitched statistics are Mel Harder (3426 innings, 4.10 DERA) and Wilbur Cooper (3480 innings, 4.05 DERA).

Using the short-form formula for calculating win shares, I get the following:

Year WS
1920 6
1921 0
1922 19
1923 19
1924 20
1925 22
1926 13
1927 13
1928 16
1929 17
1930 18
1931 10
1932 12
1933 8
1934 11
1935 6
1936 11
1937 2
1938 0
Total 223

Looking at the comparable pitchers, though, it looks like the short-form formula may be producing estimates that are a bit on the low side. Harder was credited with 234 WS and Wilbur Cooper with 266.

Wilbur Cooper is at the very bottom of my 50 person consideration set and Mel Harder just missed cracking that group. If Andy Cooper is just a tad below them, he's somewhere between 55 and 75. "Now starting for the Hall of Very Good All-Stars, Andy Cooooper."

When I reviewed A Cooper's MLE winning percentage, innings, and record and compared them to big similar big-league records, Harder, Drysdale, Blue, Cooper, Hunter, Tiant were the names that came up.

The caveat here, however, is threefold.
1) Chris Cobb hasn't weighed in yet on the MLE process, and I'd like to see if he has any tweaks before making any final judgements.
2) Pursuant to number one, I didn't make any adjustments for the batting or fielding skills of a pitcher's teammates, which Chris does, and which could potentially influence the direction of the MLEs.
3) I don't know what kind of matchups Cooper was handed. Was he started against the top starters of other teams? Was he pitted only against top teams? Or only against lesser teams? This question is significant, because it appears this is a much bigger issue in the NgLs than in MLB.

It's possible that points 2 and 3 cancel one another out ("offsetting penalties" as they say in the NFL). Then again, this guy also had a lot of career WAT despite pitching for excellent teams virtually his entire career.

I still feel like something is missing. Well of course something is missing. Data is missing, pre-1920 and for some of the best players who were barnstorming in the 1930s. So it is frustratingly difficult to feel comfortable, to feel like I really know how good these guys--the NeL pitchers--were.

With Cooper, we know he is #7 all-time in NeL wins, and he is #2 in WAT. And he is top 10 in the % of team's decisions that he took (i.e. a workhorse).

We have already elected Foster, Foster, Williams, Rogan and Dihigo, and Paige and R. Brown are looking good. Was there really no other pitcher worthy of the HoM--basically only W. Foster and Williams between 1915 and Paige and Brown in the '40s?

Some MLEs

Cooper 223-176, .560, 223 WS
H. Smith 174-123, .588
Byrd 208-180, .536, 197 WS
Redding 267 WS
Winters 163
Bill Foster 210-155, .575, 254 WS

Cooper actually looks a lot like Bill Foster whom we elected very easily.

Cooper 7th in W, 10th in dec., 10th in %, 2nd in WAT
Foster 4th 4th 14th 5th
Byrd 6th 3rd 31st 4th
Winters 8th 8th 29th 6th
Rogan 4th 9th 6th 9th

Granted, two fairly comparable players can really separate on a ballot (Van Haltren-Ryan). And some really good players I once voted for are no longer in my top 50. But if I've got 10-12 ML pitchers in my active consideration set, I am not comfortable that Mendez and Redding are the sum total of NeLers who should be, too. Some of these guys (above) were at least very very good. We just don't quite know which ones.

I raised this before but in a different manner and got taken to school by Gadfly, but here goes again...

While there were plenty of black all-star to HOM level position players in the 50's and 60's, there were very few pitchers. Don Newcombe, Bob Gibson, and Juan Marichal are the only ones that come to mind right now. I am not sure if we coudl count Stachel Paige in that group or not. He was certainly HOM level but did he pitch that way for the Indians? Either way, that total is a lot lower than oen would expect when you think of all the black position players that dominated the time period.

How much of this is a bias amongst MLB decision makers of the time? Is it even possible that the Negro Leagues for some reason or another weren't as good at developing pitchers as they were at developing position players?

Not saying this even comes into my thinking when I rank NeL plitchers (Redding is on my ballot and Mendez is in a position to return if we ever get back into the backlog) but it might be worth bringing up again. I also dont' know how we could tell such a thing through MLE's. And it is possible that the 50's/60's generation was simply a lost generation for black pitchers.

As much as we look over these guys I just am not impressed enough with any of them to hoist them up to my ballot. Right now it looks like every one of them (Brewer, Cooper, Byrd, Winters. etc.) falls more into that W. Cooper, Mel Harder, Tommy Bridges, Addie Joss group of guys who don't crack my top 60.

J,

A question and a comment. First the comment.

This is not a well-informed opinion, but it seems to me that many of the best NeL pitchers of post-WWII never got a chance in the bigs--Jim LaMarque among others. Over on the NeL Pitcher thread there are lots more names. Needless to say, a lot of biography is needed to support this statement and I don't have it....

It just seems to me that if a guy wasn't a star or ready to step directly into the MLs, the he didn't get a chance. We know that even the Dodgers had a quota of black players above which they wouldn't go. So the guys who got a shot early on were Jackie, Campy, Willard Brown (well not a shot so much as a few AB but he was a big star), Doby, etc., not anybody that needed seasoning. And the obvious, successful NeL pitchers at the time were already a bit long in the tooth--R. Brown, H. Smith, Brewer, Byrd. The next generation maybe wasn't quite ready, I don't know, but nurturing them was not in the contract the MLs made with the NeL players.

The question. How does this relate to Brewer, Cooper, Byrd, Winters? Is it the idea that if there weren't very many great black pitchers in the '50s and '60s that therefore, there may not have been prior to 1945?

If this is the inference, I don't think it follows, especially since we don't really know why there weren't more black pitchers in the '50s and '60s. It is entirely plausible that black pitchers were imagined in the same way black QBs were. Whereas black teams of course had black pitchers and black QBs.

I think the real hypotheses here (for pre-1945) might be (and I list them in order of likelihood at least as I would guess them right now):

1. The NeLs blew out a lot of arms in the way that the 19th century MLs did.

2. The translations for Cooper, Redding, Byrd, Winters, Brewer etc. etc. etc. are too low--whether systematically or randomly, which is an argument in and of itself.

Only then would I list your hypothesis as the third and least likely option:

3. Black American culture failed to produce VG to great pitchers in the same proportions as white America did.

For post-1947, to me it is still hypothetical but yet well-nigh unto a certainty that the ML establishment didn't want black pitchers, generally speaking.

jschmeagol,

I mentioned in reponse to your previos message along this line that this was a long-term problem well into the early 1990s...it sort of started to subside after the Al Campanis comments woke the baseball world up a little bit to its own institutionalized racism.

But to tackle the more pragmatic aspects of your question, that is the part about the missing generation of black pitchers...here's one take on it.

Once integration started, the majors feasted on the Negro Leagues. Although less picky about hitters, with pitchers they scooped up those pitchers with live arms who were young enough to be projectable. So guys like Jim Lamarque were picked over because they were closer to 30 than not, whereas the Newcombes were pre-peak or peak pitchers.

Within three years of integration the Negro Leagues withered to the point where they were no longer of major-league caliber. So all these older pitchers either:
a) continue to cash ever-smaller checks in the crumbling Negro Leagues
b) go pitch in the lower minor leagues but never get much of a chance to climb the ladder (McDaniels, Wright, et al)
c) go play in the Mandak/Candian league (an integrated minor league where Willie Wells among others went, though obviously he's not a pitcher)
d) go play in Mexico, Cuba, wherever in Latin America they find a good gig (Day, Manning, McDuffie, et al.)
e) quit baseball (Bill Byrd and Hilton Smith)
f) play semipro ball to scrape together a living near their home towns (Jim Lamarque)
g) some combination of all of these (which many did).

So there's an enourmous pool of proven talent out there that the big leagues (foolishly, IMO) never tapped and whose fire was extinguished from the baseball universe.

Meantime, how were conditions for those pitchers who made it into "organized" baseball? Well, Newc made it big time, of course, and Joe Black and Dan Bankhead and Connie Marrero and Satch and others were all big leaguers, but numerous others never got very far. Two examples:

-Johnny Wright went 37-16 from 1937-1947 with 3.2 WAT. He was signed by the Dodgers right around the same time as Campy and Newc. He was dismissed from the Dodgers' farm system due to attitude problems, and no one wanted a piece of him after that. Now I don't know if Wright was a problem child or an enfant terrible, but somehow I don't think it would have taken much sassing for a black man to get the boot from a white team in the late 1940s.

-Dave Barnhill, who will be an interesting HOM candidate, went 45-26 in the NgLs with 13.6 WAT from 1941-1948 and signed with Dandridge and Irvin to play at St. Paul for the Giants. Only Irvin made it to the show. Barnhill had three up-and-down years with St. Paul before being cut. But you have to wonder how much "seasoning" a guy would need having already pitched at the highest possible level he could for almost 10 years.

Anyway, there's plenty of stories like these where NgL pitchers in or near their prime were signed to play in the minors and never made it.

So combine the lack of 30+ guys getting a chance with the seeming quota or whatever system and you've got a big generation gap through the late 1940s and the 1950s.

In 1955, the only pitchers I personally (and perhaps erroneously) recognize as being dark-skinned are Connie Johnson, Pascual (or was he a lighter toned Latino?), Newc, Black, Sam Jones. So five, maybe ten guys among 16 teams. About one-half a black pitcher per team.

By 1965, the majors included Gibson, Tiant, Pascual, Marichal, Mudcat Grant, Downing, Rudy May, Segui, Odom, Jenkins, and Cuellar, plus lesser lights and probably a few I'm forgetting. Say 10-15 among 20 teams. About 3/4 of a black pitcher per team.

In 1975, the majors included JR Richard, Don Wilson, Vida Blue, Dock Ellis (or was it Doc Medich?), Odell Jones, Lynn McGlothlin, Elias Sosa, Ray Burris, Donnie Moore, Borbon, Reggie Cleveland (IIRC), Jessie Jefferson, Jim Bibby,
and Mike Norris as well as many of those mentioned in 1965. Probably 20-30 black or dark-skinned pitchers among 24 teams. About 1 black pitcher per team.

In 1985, the majors included Luis Leal, Ken Dixon, Oil Can Boyd, Ed Correa, Rijo, Dave Stewart, Jose Guzman, Andujar, Gooden, Youmans, Lee Smith, Charles Hudson, Fred Toliver, DeLeon, Guante, Alejandro Pena, Ken Howell, Dennis Powell, Pascual Perez, and some previously mentioned hurlers. (I'm only listing guys who I remember as being darker-toned players, there were many, many lighter-skinned Latino pitchers and I have no idea how MLB would have treated them in the immediate aftermath of integration.) Probably at least 25-35 guys among 26 teams. A smidge more than one black pitcher per team.

In 1995, the majors included Borbon Jr, Terrell Wade, Bobby Jones, Robert Person, Slocumb, Gene Harris, Yorkis Perez, Willie Banks, Pedro, Carlos Perez, Mel Rojas, Urbina, Bobby Jones, Carrasco, McElroy, Mike Jackson, Johnny Ruffin, Pedro A. Martinez, Jamie Navarro, Anthony Young, Donovan Osborne, Ken Hill, Rene Arocha, Esteban Loaiza, Ramon Martinez, Astacio, Daal, Felix Rodriguez, Marvin Freeman, Armando Reynoso, Salomon Torres, Melido Perez, Mariano Rivera, Benitez, Rhodes, Alan Mills, Juan Guzman, Jose Mesa, Flash Gordon, Roberto Hernandez, Baldwin, Mahomes, Latroy Hawkins, Mike Harkey, Darren Oliver, and many more that I don't know about. Obviously quite a bit more, among 28 teams. Probably at least two black pitchers per team.

And today I'd bet that number is a little higher yet.

So it seems like the majors have been moving toward greater racial parity among pitching staffs (espeically if you consider the influx of Japanese and Hispanic pitchers), but it edged up very slowly for African Americans and dark-skinned Latinos. I mean I don't really think that it's unlikely that in 1955, for instance, there weren't 5-10 more black pitchers than those in MLB who weren't equally qualified or more qualifed to be pitching than the rotting carcases of Grizzled Veterans A through P. Same for 1965 and 1975 and 1985.

Anyway, to close this message up, that's how come I see a generation gap for black pitchers, and that gap doesn't appear to have begun closing until very recently.

Camilo Pascual was a fairly light skinned Cuban, ditto Pedro "Pete" Ramos, not like one of my other heroes, Tony Oliva, whom I assume everybody knows was a dark skinned Cuban.

Can't wait to learn more about Ray Dandridge. There was just a story about the Minneapolis Millers-St. Paul Saints rivalry, specifically from WWII through 1960 when they were affiliated with the Giants and Dodgers, respectively. The hook being the Twins playing the Dodgers this weekend.

They listed HoFers who played for or managed each club--the Millers had 17 over the years, the Saints just 6. Most of them were here for a cup of coffee--including Teddy Ballgame and Yaz (the Millers were affiliated with the Red Sox before and after the Giants) and Willie Mays (hit .486 in 77 games!). Duke Snider and Newk played in St. Paul. Bill McKechnie is the only HoFer to play or manage (in his case, manage) both.

But of all of them, the two who played here the longest, both for Minneapolis, were Dandridge (4 years) and Rube Waddell (3 years).

If Dandridge is really a deserving HoFer, doesn't that say that it was a mistake (a racist mistake) to have been kept in AAA for 4 years 1949-52? (I have to admit, however, that I don't know how old he was.)

Anyway I think the Doc is probably right. Older NeLers never got a chance, superstars did, and everybody else got sort of a chance, sort of not. If they couldn't stand the racial slurs they were out. And there were no NeLs to go back to. So we here at the HoM, who would like to give them a fair shot at immortality, have nothing to go on.

At what point exactly should the NeL's no longer be considered a major league. Obviously most of the good players were still there in 1948, but by 1953 how strong were they? 1958?

Good points guys. I get this idea sometimes and it seems on the surface that it could explain our lack of interest in NeL pitchers, i.e. they sim;ly werent' as good as their position player brethren, but over the idea doesnt' stand up.

Random fluctuation over small sample size guys; we're only talking 5-6 people. There may have been more NEL Homers than should demographically have been the case, i.e. 20-25 rather than 12, but only say 5 of them were pitchers. If it was 5 out of 12 you'd say pitchers were overrepresented. Doesn't have to be anything more than the luck of the die-roll.

To the question of when the NeL ceased to be a "major" league: The question assumes it was ever a ML and that is probably in the eye of the beholder. Some say the UA of 1884 was a major league, others say no. I say that player records (uber-stats like WS) from it should probably be discounted 65%. Call it one-third major or two-thirds minor, same either way.

To me the old "major" AA of the 1880s was from 35% to 0% worse than the NL. At 0% it is reasonable to call it a major. What do you call it at 35% below NL quality?

How about the PCL? Some say it was .9 to .95 the quality of the MLs. At that rate it was every bit as major as the old AA or even the NL of the 1910s.

So finally, was the NeL a major league? Does it matter? Major is just a word. If the discount should be 5%-8%-15%, whatever, whether that's major or not. But (assuming we agreed on the conversion rate, which we don't) we would know how to treat the players regardless of whether we know whether to call it a major.

The question, then, is what is the proper conversion rate at different times? If your threshold for major status is .9, e.g., then when did it dip below .9?

One thing I noticed with Dr. Chaleeko's MLEs in # 28 and my win shares estimates based on them in # 41 is that the win shares are less volatile year to year than the DERA. The reason is that the estimated innings tends to go up or down in sync with the estimated DERA. So a season that's good in terms of a low DERA tends not to show a big jump in WS because its offset by a lower estimate of innings.

This strikes me as a somewhat unusual pattern - usually we think of pitchers getting more work in the seasons that they are successful. I assume this pattern must be coming from Cooper's actual NeL statistics, but it seemed a bit unusual. If he had followed a more typical pattern of pitching more innings in the years when his DERA was lower, his career DERA also would have been lower.

I guess I will rephrase. At what point do the NeL's dip to a quality where its star players are no longer serious HOM candidates because the best black and dark skinned latino players are either in MLB or in the white MiL's? I would say with some certaintly that the lower minors of the 20's and 30's are of a haven't been viewed as serious HOM candidates. At what point are the NeL's at that level?

So far we have treated these NeL as if all players who were stars in those leagues shoudl be considered for the HOM. At what point does this stop? Does it ever stop or do the NeL's go bust before this happens? I would presume that it happens fairly soon, wihtin the next decade or so at the most.

I'm not the person to give an absolute answer to this, but I think the generally accepted answer is after 1948 or 1949. Holway stops his book at 1948. That doesn't mean you have to ignore what a borderline candidate was doing at that point if they weren't in the majors or the "official" minors, but that's pretty much when the juice had gone out of the NeL.

I think the generally accepted answer is the 1949-1950 time period is when the Negro Leagues quality took a severe nosedive, and then after 1950 the league was probably similar to what we think of as AA quality and probably even less than that by 1953...

Brent,

I based my innings allotments on a combination of Cooper's decisions and his rotation slot for 1920-1930. He exhibited the pattern you mentioned, so I just kept it up during the less-well documented 1931-1937 period (as noted in the methodology). So some of this is conscious on my part, but it grounded in a real pattern in his career.

In 1975, the majors included JR Richard, Don Wilson, Vida Blue, Dock Ellis (or was it Doc Medich?), Odell Jones, Lynn McGlothlin, Elias Sosa, Ray Burris, Donnie Moore, Borbon, Reggie Cleveland (IIRC), Jessie Jefferson, Jim Bibby,

Not that it takes away from your very good point, but Reggie Cleveland is a white guy from Saskatchewan. Also Dock Ellis is right and Doc Medich is white.

Glad that eight years of research finally paid off ;-)

It's fun re-reading these old threads.

Brooks Lawrence was another black MLB pitcher from the 1950s who was pretty good.

Several NgL reputations have changed a lot from the 1980s, when Holway's book was basically the only serious source, to today. Oscar Charleston, for example, was just another good CF to 1980s reps; Cool Papa was considered the big star CF. Holway was a big fan of Ray Dandridge, if I remember right. In any case, the only third basemen I can remember hearing about from that far back were Dandridge and, if I remember the name right, Judy Johnson. Ray was surely, at that time, considered, by those white baseball fans who cared, to be the best NgL 3B ever. His rep has dropped off some from that, due to many more stats showing up. Also, at the time, 3B was still not considered an "equal" position to the others, in baseball of any color, which is one reason why there are so few in the Hall of Fame. 3B was considered the spot where you played the guy who had an arm but wasn't really good enough to play shortstop, or, at least, to beat out your incumbent shortstop.

On the other hand, the only reason I looked at this thread is that I have literally never heard of Andy Cooper before, that I can remember, so what do I know? - Brock Hanke

Because of my support of Hilton Smith, I re-examined Cooper. It seems to me that he suffered a catastrophic arm injury in 1930 that limited his ability to pitch at the highest level of competition thereafter.

Yet that alone shouldn't disqualify Cooper. Dave Stieb is an interesting comp: He injured his arm in early 1991, at the start of Season 12 as a plus pitcher (starting 1980), and we inducted him. Cooper's injury occurred at the very end of Season 9 as a plus pitcher (starting 1922).

Stieb's two extra quality seasons go a long way in elevating him, but Cooper is closer to HoM worthiness than we've been giving him credit for.

Hey, everyone,

Please find my latest MLEs for Andy Cooper here. The fully articulated MLE method is linked in the the article.