Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Monday, July 25, 2005

Willard Brown

Willard Brown

Eligible in 1958.

John (You Can Call Me Grandma) Murphy Posted: July 25, 2005 at 01:51 PM | 114 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 1 of 2 pages  1 2 > 
   1. John (You Can Call Me Grandma) Murphy Posted: July 25, 2005 at 01:55 PM (#1496134)
He da Hombre!
   2. Dr. Chaleeko Posted: July 25, 2005 at 02:42 PM (#1496226)
Brown's case is really quite interesting and potentially tricky. I think I've elaborated in other places, but I don't remember. Anyway, I don't have my sourceworks with me, but here's what I remember about him off the top of my head.

1) Came up at a young age with the KC Monarchs as a SS.

2) Basically led the NAL or was top 5 in everything all the time, except walks.

3) Went to Mexico for parts of two years, then to war for a couple of years.

4) Every time he came back to the NgLs he still had game.

5) Got an unfair trial with the STL Browns in 1947 and hit about .179 with a homer in about a month's worth of games.

6) Returned to the NgLs and dominated again into the early 1950s.

7) Went to the Texas League and completely dominated it in the mid-1950s as his career wound down.

As a player, he seems to have had a lot of tools and skills. He
-hit for high average
-hit for lots of power
-stole bases at a good clip
-had enough range to play SS in the first half of his career, then CF later.

The weakest part of his game, by far, was a hacktastic approach at the plate. In Mexico he drew very very very few walks in leagues where walks were more frequent than in the US.

So, if you were to ask what group of players he might belong to, the answer would be a clutch that includes the type represented by Alfonso Soriano, Vlad Guerrero (remove his IBB and you'll see what I mean), Sosa, Robin Yount, and Kirby Puckett. In other words he looks like a guy who makes the very most of a low-walks skill set. I intuit Brown to be a much better defender than Soriano, better than Guerrero and Sosa, but not as good as Puckett. I mentioned Yount due to the career arc and position change, however, Robin walked more than Brown would have.

I think we'll really need Chris's translations to get a good idea of what Brown looked like, but at this early stage, he'll be debuting on my ballot in the middle of it at least.
   3. karlmagnus Posted: July 25, 2005 at 07:53 PM (#1497011)
Would you put Soriano on the middle of your ballot? Will be very interested in Chris's conversions, but my instinct is short uneven career and no walks doesn't do it. However, vamos a ver.
   4. sunnyday2 Posted: July 25, 2005 at 08:14 PM (#1497046)
Walks aren't everything. Here is a partial list of players with more than 150 percent as many walks as Jake Beckley.

Cupid Childs
Roger Connor
Miller Huggins
Dunny Hoy
Ralph Kiner
Ron Cey
Enos Slaughter
Jesse Burkett
Jim Gilliam
Max Carey
Roy Thomas
Eddie Joost
Norm Casgh
Roin Fairly
Robin Ventura
Bill Dahlen
Lu Blue
Jimmy Sheckard
Harry Hooper
Toby Harrah
Max Bishop
Donie Bush
Chili Davis
Billy Hamilton
Pee Wee Reese

And really there are another 80 of them.
   5. sunnyday2 Posted: July 25, 2005 at 08:25 PM (#1497067)
Willard Brown was born in 1915. James has him as the #2 RFer in the NeLs after Dihigo and three slots ahead of Oms (Ted Strong and Wild Bill Wright squeeze in between).

He says "Tremendous right-handed power hitter, maybe comparable to Jose Canseco, Juan Gonzalez, Andre Dawson, or Frank Robinson."

Obviously I'd like to which one he comps the best!

"Hit hundreds and hundreds of home runs in the NeLs, Puerto Rico (where he still holds many records) and Mesico. Quincy Trouppe says, 'Brown is one who challenged the feats of Josh Gibson.'"

As I said, rates ahead of Ted Strong, whom James comps to Dave Winfield, and also ahead of Oms, whom James comps to Paul Waner. It doesn't seem that you'd have Canseco ahead of Winfield and Waner, or Gonzalez, or Dawson. So maybe Frank Robinson is your man. OTOH F. Robby and even Canseco walked a bit, while Juan Gone and Dawson didn't walk much, so maybe the latter are better comps. But the raw power that is attributed to Willard Brown sounds more like Canseco.

I'll be interested to see if Brown hit many 2Bs. Dawson and Gonzalez both had good doubles power and/or hustled to get down to second when they could.

James says Brown was perceived sometimes as not hustling. Buck Neil, he says "said he hustled. He was just so easy-going that he sometimes rubbed people the wrong way." Canseco was, shall we say, lackadaisical about his game and it shows up in relatively fewer doubles than Juan Gone and the Hawk but roughly comparable HR totals.
   6. sunnyday2 Posted: July 25, 2005 at 08:33 PM (#1497084)
PS. Don't be fooled, accept no substitute.

IOW there was a Willard Brown who played 1887-1894, a C and 1B, who apparently was OUR Willard Brown's bizzaro double--i.e. a pretty good fielder but an 82 OPS+ hitter with little or no power (6 HR in 1700 PAs). His real name was William, however, and he sometimes went by Big Bill. Why exactly he ever went by Willard I don't know, but that's how he's listed in the encyclopedia.

OUR Willard debuted July 19, we just missed his anniversary, and he played 21 games. Not sure if that was 21 out of 60-70 through the end of the season or if they let him go early,

When Doc says he got an "unfair" trial, my understanding is that among other things, they said, don't bring any bats, we'll supply the bats. But he used a big heavy bat and they didn't have any to his liking, and made no effort to get any. Jeff Heath had the heaviest bat on the team and Mr. Heath made it known that over his dead body would Mr. Brown use any of his bats.
   7. Dr. Chaleeko Posted: July 25, 2005 at 08:35 PM (#1497087)
Karl,

I think Soriano is an inferior player to Brown. Like I said, I'll post his info tonight, but he is absolutely ALL OVER the NgL leaderboards. I wouldn't consider his career uneven at all; I'd consider him consistently excellent in that he essentially dominated most every league he played in year in and year out.

The walks may keep him from being an upper echelon player, but his total package is, at first blush, pretty darned impressive.

Sunnyday, I'd forgotten about Dawson, he feels like a very good comp as well, but I think Brown probably hit for higher average and would have played CF longer. IIRC Brown was on the doubles leaderboards very frequently; I'll verify this evening.
   8. Dr. Chaleeko Posted: July 25, 2005 at 08:41 PM (#1497094)
To add to Marc's tale of Brown's trial, Brown borrowed a teammate's bat (possibly Heath), and homered with it for his one and only MLB homer. The teammate picked up the bat and broke it over his knee after the homer.

Unlike the Dodgers who allowed Jackie, Campy, Newc, Jethroe, and others a short adjustment period in the minors, Veeck brought Brown and Hank Thompson up right away. Apparently there wasn't much communication between the front office and the dugout either because the then-current SLB players were definitely not on the same page about it. The whole thing reeks of the same "visionary thinking" (read: exploitive sideshow mentality) that brought Eddie Gaedel to the plate.
   9. karlmagnus Posted: July 25, 2005 at 09:06 PM (#1497125)
Presumably Brown's fellow players didn't steal his shoes, so 0 walks in 67AB is not impressive. I think this is however DIFFERENT "visionary thinking" to Gaedel, since correct me if I'm wrong (Baseball Reference doesn't have ownership, silly omission) but Veeck didn't own the Browns yet in '47.

In reality, walks aren't everything, but I will take some convincing about yet another NEL player from the most over-represented era in the HOF. My mind is however open, to be filled with data by Chris.
   10. sunnyday2 Posted: July 25, 2005 at 09:09 PM (#1497133)
It was Heath and it was in fact one of Heath's bats that was busted but in such a way as to be legal to use. I had understood that he (Heath) busted it on the dugout wall.

Muddy Ruel was the manager of this mess.

Thompson, age 21, unlike Brown, survived this baptism of fire. He hit .256 in 27 games, mostly at 2B. But of course the Browns got rid of him, and he came up with the Giants in '49 and stayed for 8 years, mostly at 3B.

Meanwhile the Browns ran Jerry Priddy and Bob Dillinger out there at 2B and 3B in '49, then Owen Friend and Bill Sommers in '50, then Bobby Young and Freddie Marsh in '51, Young and Jim Dyck in '52 and '53, then Young and Vern Stephens in '54. (Stephens had played with Brown and Thompson in St. Louis in '47).

In Baltimore, by the time Thompson retired, the Orioles had also tried out Wayne Causey, Billy Gardner and George Kell at the 4 and 5 holes.

In 1953, the Browns' last year of existence, 3B was actually shared among Dyck, Stephens and Bob Elliott--the latter two being traded for one another. Stephens hit .321 in 46 games with St. Louis, while Elliott hit .260 in 67 games with the White Sox after the trade.

Thompson hit .302 with 24 HR in New York that year. In '54 he hit another 26 HR, then hit .354 in the World Series. He took over at 3B mid-way through 1951 BTW, basically for Don Mueller as Bobby Thomson moved from 3B to the OF and Mueller to the bench. (Hank did hit just .143 in that Series).

Through it all, I doubt that anybody in St. Louis had the presence of mind to wish they'd hung on to the guy.
   11. Dr. Chaleeko Posted: July 26, 2005 at 02:31 AM (#1497944)
I've finally gotten a chance to look at the source books. Here's what they say:

Born June 26, 1911, so a little older than I thought when he came up.

1934: Signed to minor league Monroe Monarchs

1935: Signed to the KC Monarchs, played SS, no data.

1936: 11 for 30 in seven league contests; t-5th in the league in AVG.

1937: .361 average; Led NAL in HR (9), HR/550 (24), doubles (13), triples (8); 6 SB tied for fourth, though KC may have been only team reported.
-8/20 in playoff series
-1/2 in WS (with CHI)
-5/13 vs. White players
-8/55 in Cuba.

1938: Switched to CF for KC. Led league in HR (6), HR/550 (29), SB (10); Second in league with 3 triples and fifth in AVG (.362).

1939: CF. Led league in doubles (9) and SB (3); Second in HR (2); Third in HR/550 (9); Fifth in AVG (.343).

1940: Mexico, Nuevo Laredo, OF. 294 ABs, .354, 18 doubles, 4 triples, 8 HR, 61 RBI, 10 BB, 15 K, 13 SB, .524 SLG.

1941: a) Mexico, Aguila, OF. 125 ABs, .256, 7 doubles, 0 tirples, 2 HR, 24 RBI, 4 BB, 5 K, 5 SB, .360 SLG

b) KC, CF. 1st in HR (2), doubles (4), HR/550 (11); 2nd in triples (4); fourth in AVG (.347)

c) 2nd in P.R. Winter League with .409 avg.

1942: KC, LF, led league in HR (9), HR/550 (24), SB (3); second in triples (3); third in doubles (4), .310 average.
-7/17 in playoffs
-2/3 vs. whites.

1943: KC, CF. Led league in HR (6); second in HR/550 (28); third in doubles (5), .309 average.

1944-1945: Military service.

1946: KC, LF, Led league in HR (13), HR/550 (102!!!), second in AVG (.348), third in SB (13) [no data on doubles and triples]
-Playoff 7/29
-P.R. Winter league: 99/124 (.390) to lead league in average.

1947: a) KC, CF. hit .336.
b) MLB, you know the story.
c) P.R. Winter League, 101/234 (.432) to lead league, 27 HR to lead league.

1948: KC, CF. Led league in HR (18), HR/550 (101!!!); second in SB (13); third in doubles (20) and AVG (.374).
-P.R. Winter League. .323 AVG. and 18 HR (per Riley).

Holway numbers end, we're in Riley territory now.

1949: .371 AVG
-P.R. Winter League: 16 HR, .353 AVG.

1950: Border League, Ottowa. .352 AVG, 1 HR, 18 RBI*
-P.R. Winter League. 14 HR, .325 AVG.

1951: P.R. Winter League. .295.

1952: P.R. Winter League. .342.

1953: Texas League, Dallas. Hit .310 with 23 HR, 108 RBI.
-P.R. Winter League. .265.

1954: Texas League, Houston. Hit .314, 25 HR (another source, http://www.attheplate.com/wcbl/majorleaguers.htm, says 35 HR), 120 RBI.


1955: Texas League. Hit .301, 19 HR, 104.

1956: a) Texas League, Austin, San Antonio, Tulsa. Hit .299, 14 HR, 73 RBI.
b) Western League, Topeka. Hit .294 with 3 HR, 14 RBI.
-P.R. Winter League, Santurce. .261, 2 HR, 23 AB, 5 RBI.

1957: Mandak League, Minot. .307 AVG, 9 HR, 29 RBI

*per http://www.attheplate.com/wcbl/majorleaguers.htm

Can anyone fill in the gaps with more info? His Texas League numbers, his other Minor League numbers would be great, as would his P.R. numbers.
   12. Devin has a deep burning passion for fuzzy socks Posted: July 26, 2005 at 03:03 AM (#1498069)
Correcting Dr. C, Bill Veeck wasn't repsonsible for Brown and Thompson's indignities. He was running the show in Cleveland and signing Larry Doby. (Although I'm pretty sure he did bring Doby right to the majors, I don't recall any notable problems with his fellow players.) The guilty party here was....Bill Dewitt? I know he ran the club for a while before Veeck, but I'm not entirely certain he was the guy in charge at that point.

And since Veeck did wind up taking over in St. Louis in 1951, there probably was someone there who was unhappy they'd let Thompson go. Veeck brought in Satchel Paige, and claimed he would have signed Ernie Banks if he'd had the cash on hand.
   13. sunnyday2 Posted: July 26, 2005 at 04:35 AM (#1498377)
Bill Dewitt is the guy who Bill James mentions in connection with Roy Cullenbine. Dewitt traded Cullenbine because he (Cullenbine) was a "lazy bastard, always trying to get a walk." James referred to Dewitt as being "high octane ignorant."
   14. Gary A Posted: July 26, 2005 at 01:33 PM (#1498786)
I believe I've seen three different birth years for Brown--1911, 1913, and 1915 (all with the same date, June 26). Not sure which is currently accepted. It would make sense that he lied about his age when signing with the Browns, especially if he was really 36 or thereabouts.
   15. Dr. Chaleeko Posted: July 26, 2005 at 06:08 PM (#1499383)
Does anyone have a sense of how strong the Puerto Rican Winter League was during the 1940s and 1950s? Brown played in it extensively during his career and became quite popular there while doing so. In fact, he set the single-season PRWL HR record (27).

Anyway, his PRWL batting AVG declines precipitously as he ages, while his NgL AVG remains pretty high. I think this means that the NgLs were declining in quality over the late 1940s, while the PRWL was either staying roughly the same (so that aging appeared to take its normal observed toll on Brown's AVG) or it was becoming a tougher league over time.

Since I don't know its composition at all, I was curious if anyone had any sense of it.
   16. Gadfly Posted: July 26, 2005 at 08:31 PM (#1499744)
WILLARD BROWN

Willard Brown was definitely born in 1915 (confirmed by birth certificate, social security application, and census research). When he came up to the Majors, Brown listed his birth year as 1921. Years later, when reporters and writers came around to interview him, Brown took a page from the Satchel playbook and started saying how he was just too old when he got his chance (i.e. born in 1911 or 1913); but it's definitively 1915.

Brown's Puerto Rican Record:
YEAR-AB-R-H-2B-3B-HR-RBI-BA-SA
41-42 122 22 050 13 04 04 26 .409 .680
46-47 254 44 099 25 04 09 50 .390 .626
47-48 234 79 101 20 05 27 86 .432 .906
48-49 294 59 095 20 03 18 69 .323 .595
49-50 331 65 117 21 06 16 97 .354 .598
50-51 305 57 099 19 03 14 76 .325 .544
51-52 112 14 033 04 01 04 22 .295 .455
52-53 114 20 039 07 00 03 20 .342 .482
56-57 023 02 006 00 00 02 05 .261 .522
TOT 1940 378 679 135 27 101 473 .350 .604

NOTES
Played with Santurce every year except the first.
1941-42 Finished 2nd in BA (behind Josh Gibson)
1946-47 Finished 1st in R, H, BA.
1947-48 Finished 1st in R, HR, RBI, BA (Triple Crown) also finished just 1 behind the league leaders in H (101-102) and 2B (20-21).
1948-49 Finished tied for 1st in HR.
1949-50 Finished 1st in H, HR, RBI, BA (Triple Crown) and second in 2B.
1950-51 Finished 1st in RBI, 2nd in HR, 3rd in H

Of course, the most interesting thing about his PRWL career is the 1947-48 season which is pretty much entirely out of context with the rest of his PR career. He won the Triple Crown counting stats by enormous margins [27 HR to 13 (2nd), 86 RBI to 58 (2nd)] with a ridiculous .906 SA which was bettered only by the great Josh Gibson in a much shorter season.

Brown had great motivation for this season. Major League scouts were all over the Carribean and a good performance would possibly earn him another shot at the Majors (and Major League money) and Brown was almost assuredly pissed off over his treatment in the Majors.

[In every interview that I have seen where Brown discussed his 1947 time with the St. Louis Browns, Willard, who was an easy-going man, had nothing but contempt for his namesake team. He generally always mentioned that the KC Monarchs of the 1940s could have killed the Browns in a series without even half trying.]

To understand the statistics of the PRWL, you have to keep in mind the fact that, from 1946 to 1955, the Major Leagues were busily expanding their monopoly into Latin America. The Majors began providing subsidies, support, and protection to Latin America Leagues that accepted Major League rules, regulations, and players.

In other words, the Latin Leagues, that accepted the Major's offers, sacrificed their independence and became training grounds for the Majors. By the early 1950s, the quality of the League was in sharp decline as the veteran Negro League players (like Brown) where driven out of the League in favor of players being developed by the Major Leagues.

In my opinion, the PRWL was of about 3A strength from its founding in 1938-39 until the early fifties (with the typical dip for WW2) when it began to decline to about 2A.

Hank Aaron 1953-54 PRWL
274 37 84 16 3 9 42 .307 .485 7
Willie Mays 1954-55 PRWL
172 63 68 15 7 12 33 .395 .773 10

In 1954, Aaron went on to hit .280-.447 in his rookie season for the Milwaukee Braves. However, it should be pointed out that these Major League statistics badly underestimate how good a hitter Aaron was at this point. Milwaukee was a very good pitcher's park (Aaron hit 13 HR in 1954. One at home and 12 on the road).

On the other hand, Mays, in 1954, had just hit .345-.667 in the Majors before coming to Puerto Rico. But the important thing is that Aaron was typical of the players being sent to Puerto Rico at this time, i.e. a rookie. Mays, on the other hand, had to get a waiver of the rules to play and also arrived late and left early so as not to interfere with the rights of his Major League club.

In my opinion that .350-.604 BA-SA career line for Willard Brown line converts to an about .315-.335 BA/.525-.575 SA in the Majors.
   17. Dr. Chaleeko Posted: July 27, 2005 at 02:35 AM (#1500938)
Thanks, Gad, that's great stuff about Brown and the PRWL. And nailing down his birthdate is pretty key too. It completely changes how I'm looking at him because he goes from a 24-year old entrant into the KC lineup to a 20-year old, and it also raises interesting questions about how to credit play later in his career in various non-major leagues around baseball.

Speaking of which... I'm hoping that Gadfly or anyone else out there can offer a little further help with Brown. There's three apsects of his later career which need some further clarification, and I'm much obliged to anyone whoe can help.

1) Brown played 1948, 1949, and 1951 in the NAL for KC. Most experts regard 1948 as the NgL's last year of major-league type play. So to what level did the leagues fall in 1949-1951? Were they as good as AA-level leagues?

2) In 1951 and 1952, Brown spent some time in the Dominican Summer League. Does anyone have any numbers on that time, and does anyone know about the level of play there? I'm assuming AA-level or lower, but I'd love some help.

3) Brown played in several integrated minor leagues. In 1950 he was in the Canadian Border League. In 1953-1956 he was in the Texas League. In 1956 he was also in the Western League. In 1957 he was in the ManDak League. Can anyone tell me what level of play these leagues were at in this time?

Thanks everyone. In Chris's absense, I'm going to try to piece as much of Brown's career together as I possibly can. It's a pretty wild ride though, so I'm probably going to ask for help a lot.
   18. Gary A Posted: July 27, 2005 at 11:21 PM (#1503611)
The Texas League in the mid-50s was one of two AA leagues, the other being the Southern Association.

The high minors looked like this:

PCL (special "open" classification)
IL (AAA)
AA (AAA)
Texas League (AA)
So. Assoc. (AA)
Class A leagues included the Eastern League, South Atlantic League, Western League.

The Mexican League was classified AA in 1956.
   19. Dr. Chaleeko Posted: July 28, 2005 at 05:06 PM (#1505020)
Gary, thanks for that classification information, that's helpful stuff.

Let me ask another question. In 1939, 1941, and 1942, Brown's rates take a nose dive, but it appears that it's either the result of the league playing as an extreme pitcher's league or as the Monarchs' home field being tough on homer hitters. During those seasons, Brown led the league one time with two homers, so I'm thinking this is a leaguewide phenomenon or a reporting problem, rather than a park issue.

So two questions then:
1) Can anyone tell me what the heck happened to the NAL in the early 1940s?
2) Does anyone have any information about how the Monarchs' home field played from about 1935-1950? (hitter's park? pitcher's park? neutral?)

Thanks!
   20. Dr. Chaleeko Posted: July 28, 2005 at 05:57 PM (#1505204)
Incidentally, once we're able to resolve this question about the early 1940s, should be able to post up some unregressed conversions. Given how many places Brown played, they're kind of complex and will need some 'xpalainin', but they're almost ready to roll.
   21. Gary A Posted: July 28, 2005 at 07:23 PM (#1505539)
Doc, check out Gadfly's post #1112361 in the Beckwith thread--he's got some league offensive totals for NAL and NNL in the 1940s, though I think he only goes back to 1944.

The NAL's reputation was of a low-offense (in particular low-HR), small ball league with lots of base stealing. The Monarchs' park was still Muehlebach, and in the 1920s it had suppressed runs by a little bit and home runs by a lot, while increasing batting average a little and extra base hits (especially triples) quite a bit. Kansas City Blues' stats in the late 1930s and early 40s aren't inconsistent with the park retaining these characteristics (though Vince DiMaggio did hit 46 HR for KC one year).

I guess the crucial question would be what effect the KC park had relative to other NAL parks, as I believe the league was full of large parks. I'm not sure what the answer is.
   22. Gary A Posted: July 28, 2005 at 07:56 PM (#1505645)
Here are Brown's stats in the Texas League, along with league totals:

Yr age g ab h d t hr r rbi bb so sb ave oba slg
53 Brown 38 138 522 162 36 2 23 91 108 35 52 3 310 354 519
53 TexLg 620 41197 10635 1947 258 855 5486 5042 5120 5389 474 258 340 380
54 Brown 39 144 583 183 36 5 35 92 120 35 56 2 314 353 573
54 TexLg 644 42963 11209 1928 331 957 6006 5498 5231 6321 455 261 341 388
55 Brown 40 149 544 164 34 4 19 73 104 39 42 3 301 348 483
55 TexLg 646 42403 10790 1928 243 947 5594 5218 5191 6031 411 254 336 378
56 Brown 41 104 351 105 17 0 14 50 73 36 36 2 299 364 467
56 TexLg 618 41714 11049 1848 270 1027 5972 5464 5028 5727 468 265 344 396

These are his teams and games played with each:
1953: Dallas
1954: Dallas 108, Houston 36
1955: Houston
1956: Tulsa 28, San Antonio 54, Austin 22

Looking at their team totals during these years, Dallas may have had a pitcher's park, but not dramatically so; they twice led the league in ERA, and won two pennants without coming near the league lead in runs scored.

Houston's park was probably neutral, maybe slight tendency toward pitcher; had both good offensive and good defensive years (as well as some awful offensive years).

Since Brown played for three teams in 1956, it would probably be all right to consider his park context neutral.
   23. Dr. Chaleeko Posted: July 28, 2005 at 09:11 PM (#1505909)
Thanks Gary for all this information, it's really helpful.
   24. Dr. Chaleeko Posted: July 28, 2005 at 09:16 PM (#1505923)
That K/BB data is especially helpful. The only other reports I had on his strike zone judgment was from Mexico, and it was about 300 at-bats worth. This will really help flesh things out. Interestingly, it appears that Brown was walking a lot more at the end of his career than he did in Mexico, as per the "old-players skills" theory. I do wonder if perhaps some were intentional, but then again, he's walking at roughly half the league's rate, so it's not like turned into Eddie Yost over the years....
   25. Gadfly Posted: July 28, 2005 at 09:47 PM (#1506011)
Doc-

Got your e-mail about fluctuations in Brown's batting record but am not sure if you meant from 1947-48 to 1948-49 in the Puerto Rican Winter League or in the early 1940s as mentioned above.

In any event, each period has sample size issues (i.e. unless you've got over 500 at bats, you're going to get some strange fluctuations which don't fairly represent a player's talent, being either too high or too low). I must admit that I can't wait to see the final product of the Hall of Fame's Negro League research project since it should even things out a bit.

As Gary A reports, the KC Monarchs played in a pitcher's park that surpressed home runs. On top of that, the Negro American League, in general, had a ton of really good pitching parks that surpressed both offense and home runs.

Taking a quick look at Willard Brown's stats in 'The Negro Leagues Book' (and ignoring Holway, who I feel is statistically unreliable) gives Brown:

AGE YEAR AB H 2B 3B HR BA SA SB
KANSAS CITY
22 1937 143 053 10 03 08 .371 .650 04
23 1938 104 037 03 03 06 .356 .615 10
24 1939 119 040 09 01 01 .336 .454 04
TOTAL-3 366 130 22 07 15 .355 .577 18
MEXICO
25 1940 294 104 18 04 08 .354 .524 13
KANSAS CITY
26 1941 089 030 05 04 03 .337 .584 02
27 1942 181 066 07 02 07 .365 .541 04
28 1943 113 039 06 02 06 .345 .593 ??
TOTAL-3 383 135 18 08 16 .352 .567 ??

[Note- The 'Negro Leagues Book' and Holway credit Brown with playing for Pueblo in the Mexican League in 1941. This is an error. Willard Brown did not play in Mexico in 1941 and the stats are actually those of Barney Brown. This may be responsible for a lot of your confusion.]

A few notes on these stats:

1) Brown did hit better in 1937 and 1938 than he did later. Of course, 1937 and 1938 are the first and second years of the Negro American League. The League was very unstable during those first two years and probably of a lesser overall quality. This could account for Brown's good 37 and 38 season or it may just be random fluctation.

2) Brown's 1940 Mexican League season fits right into the overall pattern of his statistics except for one thing:

1937-39 366 130 22 07 15 .355 .577
1940-40 294 104 18 04 08 .354 .524
1941-43 383 135 18 08 16 .352 .567

His home runs were down. But oddly enough, this is a pattern of Brown's career. In his first exposure to a new baseball League, Brown would either hit very poorly (for neither BA or power) such as in the Majors (.179 BA) or Cuba (.145) OR he would hit ok but without the expected power.

A good example of this is Brown's time in the Texas League. In 1953, the 38-year-old Brown played 138 games, hit .310, and knocked 23 HRs. In 1954, the 39-year-old Brown played 144 games, hit .314, and crunched 35 HRs. A quite amazing increase for a player of his age.

Brown had a reputation as a bad ball hitter, but I always like to think of his type of hitter as a 'recognition' hitter. In other words, Brown was a much better hitter against a pitcher once he got comfortable with the pitcher's delivery, style, and assortment of pitches [a modern guy like this would be Alfonso Soriano or Shea Hillenbrand].

In other words, Brown was probably a better hitter the more he saw of a pitcher.

3) The Negro Leagues, like the Major Leagues, began to have trouble with the quality of their baseballs in late 1942. Brown's 1942 and 1943 stats may or may not be effected by this and the short WW2 deadball era that lasted from 1942 and into 1946.

WORLD WAR 2

Brown spent 1944 and 1945 in the U.S Army. Of course, he got to play some ball as both he and Leon Day ended up playing in the U.S. Army's 1945 European Baseball Tournament on a integrated team against various Major League players.

In 1946, the 31-year-old Willard Brown got out of the Army and had what I think is one of the greatest unknown years of all time. Brown was credited with hitting .348 (80 hits in 230 at bats over 58 games for the season). He barely lost the batting championship to his teammate, Buck O'Neil (O'Neil hit .350).

The final statistics also stated that Willard Brown lead the NAL in 2B with 28 and credited him with 5 HR. His teammate, Ted Strong, was crowned HR king with either 7 or 9 HR, depending upon the source.

But here is where things get weird (I mentioned this somewhere in the HOM before). If you go through the game accounts, it becomes quickly apparent that Brown hit more than 5 homers. In fact, it becomes quickly apparent that Brown was the great HR hitter in the 1946 NAL.

[The 'Negro League Book' credits Brown with 8 2B, 4 3B, and 13 HR which are derived from game accounts in 1946 and inserts them into his published 1946 record (58-230-80). This, of course, is the type of sloppy stats that make Holway so unreliable. Holway often credits players with every HR he can find and then sticks it into their batting line, irregardless.]

In my opinion, Brown probably hit 15 HRs in 1946, but they dropped the "1" and just credited him with 5. Of course, that's just a supposition, but it's evident that Brown was the 1946 NAL HR champion.

I think his actual 1946 line was actually something like this:

G-AB-H-2B-3B-HR-BA-SA
58 230 80 28 04 15 .348 .700

with the great possibility that he hit more than 4 triples.

In 1946, the NAL was still in a World War 2 type deadball offense. In that context, this would make Willard Brown's 1946 season a totally forgotten monster year. Of course, it should be said that Brown, who was playing regularly in 1945, had a jump on the other 1946 NAL players.

In any event, Brown's 1946 season is, in my opinion, his greatest Summer Season (he hit 3 HRs in 6 games in the Negro World Series too); just as 1947-48 is his greatest Winter Season.

As for his PRWL stats, they don't really fluctuate much, ignoring the bizarre 'better than Babe Ruth' year in 1947-48.

In fact, his slugging averages conform nicely to the usual aging pattern:

AGE-YEAR-SA
26 1941-42 .680
31 1946-47 .626
32 1947-48 .906
33 1948-49 .595
34 1949-50 .598
35 1950-51 .544
36 1951-52 .455
37 1952-53 .482
41 1956-57 .522

Though it should be pointed out that his last three seasons are all simply partial seasons and have sample size problems.

Hope that helps, I'll try to get to his career after 1946 this wekend.
   26. Dr. Chaleeko Posted: July 28, 2005 at 09:57 PM (#1506034)
Gadfly,

Thanks for your help, that's great information. Between your info and Gary's I think I've got enough to build a really robust translation. It still won't be as good as Chris Cobb's work (and I don't know anything about regression), but I should at least be able to present some data later tonight with appropriate explanations for the methodology so that someone else can point out where I've made errors.
   27. Dr. Chaleeko Posted: July 28, 2005 at 10:37 PM (#1506121)
Gad,

I'd bet that park factors have something to do with Brown's 1940 Mexican HR totals. I've got no proof, just intuition, but in looking over several years of the leagues ERA, Nuevo Laredo's W-L record, and their team ERAs (esp relative to the league), it looks like they either always had a terrible offense, or they played in a place that suppressed offense a little bit and so kept their ERA down despite what you might expect from their records.
   28. Dr. Chaleeko Posted: July 29, 2005 at 01:48 AM (#1506513)
WILLARD BROWN TRANSLATION, ATTEMPT #1

This is my first attempt at a complete translation of Willard Brown’s career. I am initially viewing Brown as having a 15-year career that ends after 1950. I don’t know that this is absolutely correct. And I need a little help from the group to see if I’ve made the best decision (see below when I talk about MLE conversions for AA-level teams). I’m presenting these first without any military service credit. The last column, SFBWS means short-form batting win shares and uses the basic short-form formula but without any team-based reckoning. That formula is

[ RC – ( outs / 12 ) ] / 3

Therefore, the SFBWS column should be considered a VERY general stab at his batting value. They will have to be redone in the more meticulous OPS-comps method that Chris Cobb uses.

YEAR AGE POS  PA   AB    H  2B 3B  HR  BB   K   TB  AVG  OBP  SLG   RC SFBWS
----------------------------------------------------------------------------
1935  20  SS 350  343  100  20  3  14   7  27  169 .293 .307 .492   53  10.9
1936  21  SS 685  672  221  45 23  24  13  54  385 .329 .342 .574  135  32.3
1937  22  SS 650  637  174  29  6  20  13  51  275 .274 .288 .432   81  14.0
1938  23  CF 628  615  197  18 12  24  12  49  311 .320 .334 .506  106  23.6
1939  24  CF 575  564  170  25  0   5  11  45  210 .302 .316 .372   68  11.6
1940  25  CF 611  594  189  26  2  12  16  22  255 .319 .337 .430   88  18.2
1941  26  CF 625  613  203  33 17  12  13  49  306 .331 .345 .500  108  24.5
1942  27  LF 650  637  210  13  3  18  13  51  283 .330 .343 .444   99  21.1
1943  28  CF 365  353  110  13  5  13  12  28  172 .312 .334 .486   59  13.0
1944  29  CF                            
1945  30  CF                            
1946  31  LF 620  600  196  39  6  23  20  48  324 .327 .349 .541  117  27.8
1947  32  CF 650  629  203  37  7  35  21  50  357 .324 .346 .569  128  30.8
1948  33  CF 620  591  171  37  6  32  29  47  312 .290 .323 .529  106  23.7
1949  34  OF 620  591  180  22  6  18  29  47  260 .305 .337 .441   92  19.3
1950  35  OF 250  238   67   8  0   5  12  19   91 .279 .313 .380   30   5.1
============================================================================
TOTALS      7899 7675 2392 366 96 255 222 558 3711 .312 .331 .483 1269 276.1


Next we have a quick-and-dirty look at what military service might have taken away from Brown’s career. Each of the two seasons is an average of the nearest surrounding three seasons. The totals line shows the sum of these two seasons and the career line above.
YEAR AGE POS   PA  AB     H  2B  3B  HR  BB   K   TB  AVG  OBP  SLG   RC SFBWS
----------------------------------------------------------------------------
1944  29  CF  545  530  172  22   5  18  15  42  260 .325 .343 .490   92  20.6
1945  30  CF  545  527  170  30   6  24  18  42  284 .322 .345 .540  101  23.9
TOTAL        8989 8732 2734 417 107 297 255 673 4255 .313 .333 .487 1462 320.6

METHODOLOGY
-Sources for this section.
Gadfly provided seasonal data for NAL seasons 1937–1946.
Holway is the source for NAL seasons 1936, 1947, and 1948.
Basic NAL data for 1949 and basic Canadian Border League data was skimmed off the Western Canadian Baseball website mentioned earlier in this thread. It only gives HR, AVG, and RBI, but it was a start.
Gadfly also supplied the Puerto Rican Winter League data for 1941-1950.
Cisneros is the source for Brown’s Mexican sojourn in 1940.
Gary A provided his Texas League data for 1953-1956 as well as the leaguewide data.
Thanks guys.

-General method
I used every scrap of info I could find on Brown. When I had summer and winter league data, I used them in combination to help bolster the seasonal and career sample. When I had leaguewide data to work with, I first placed Brown’s AVG and SLG within the National League environment by comparing his league’s AVG and SLG to the NL’s and recalibrating his AVG and SLG so that he retained his performance relative to the league average. Then I converted his AVG and SLG in the usual way, reckoned playing time, and built up the translated numbers around these pieces of information.

-Specifics
Holway, Cisneros, Gadfly, and Gary A provided lots of info on Brown’s extra-base hits so that I felt comfortable projecting doubles, triples, and homers when data was not available by using simple career averages. Walk and strikeout data was a little harder to come by. Cisneros offers a little, but Gary A’s data offered enough data to start really building a working model. With leaguewide walk and strikeout data available in the Re-evaluating NgL pitchers thread and with Gary’s Texas League data, I adjusted Brown’s in-season walk rates and strikeouts in the same way I did his AVG and SLG. In 1940, Brown walked only about 2% of the time, but he also didn’t strikeout much. In the Texas League he did more of both. Therefore, I gave him a gently upward sloping walk rate: 2 percent for the first several years, then climbing to 3.3% for a few years, 4.7% for a few more years, then leveling off at 6% during his denouement.

1935, Brown’s rookie year is made up from whole fabric. Holway and Riley seem to intimate that he broke in that year as a regular or semi-regular, but there’s no stats. So I just tried to bluff my way through.

In terms of conversion rates, for the above translation, I think they are pretty typical. I used Chris Cobb’s standard procedure of determining the rate for batting average, then using its square for converting SLG. For the NAL the standard rates are .90/.82 rate until 1948, then I took it down to .80/.64. Chris had also recently posted the exchange rate for Cuba 1937 on the Oms thread, it was .96/.92. For the 1940 Mexican season, I used the same rate as for the NAL: 1940 and 1941 were the years when the most Negro Leaguers played south of the border. Based on Gadfly’s description of the level of play in the Cuban Winter League I hedged my bet, applying a .87/.76 conversion rate until 1948, then declining to .85/.72 in 1949 and .80/.64 in 1950. In addition the Canadian Border League is a bit of a mystery. I pegged it at .80/.64 for 1950, remembering that it was accepting numerous refuges from the Negro Leagues.

Essentially, I pegged AAA-level as .9 and AA-level as .8 for average. Could someone let me know if this is correct? Thanks.

-Playing Time
Brown appears to have been a fairly healthy player. I gave him some time out during his rookie campaign, and I think he may have missed some of the 1943 season too. Other than that, however, he played all the time. In fact, I declined his playing time in 1949 and 1950 to account for rapid aging. Using the Sabrmetric Baseball Encyclopedia, I look for players with a similar profile to Brown’s and came up with Jim Rice, Cecil Cooper, Dante Bichette, Al Oliver, Tony Oliva, and Orlando Cepeda. Few of these players had much hang-on time at the end of their careers and exited as soon as their skills diminished, so I figured the same might hold true for Brown as well.

-Other adjustments?
I have made no other adjustments to these numbers, however several could be made, and probably should be. The level of competition for KC in 1935-1939 was probably somewhat uneven and may need re-assessment. In addition, Brown’s homefields are reputed to suppress homers but encourage other types of hits; some of his triples should therefore probably become homers. I have already speculated that Nuevo Laredo was a pitcher’s park and an adjustment might be helpful for 1940. On the other hand, all this might well come out in the wash due to the uncertainty I have about when the NAL become uncompetitive. It seems that it became increasingly less competitive from 1946 onward. And having league and home-park adjustments for the PRWL might be helpful, though probably not crucial.

-My own questions and concerns for the 1935-1950 translation
1) Is the .80/.64 the right conversion rate for AA-level leagues?
2) Should there be further adjustments for league quality or home park?
3) 1935 is made up from whole cloth.
4) For a guy who led the NAL in everything but walks all the time, Brown doesn’t look all that dominant here, especially where the early 1940s, his salad days, are concerned. This is both a process and a context problem, but one I’m not yet able to answer.
5) Is the playing time reasonable?
6) If Gadfly, Gary A, or anyone else were able to supply any other leaguewide data for the NAL of 1937-1943, it would be immensely helpful in solving some of these issues because we could at least retain the dominance over league aspect of his performance.
7) Fielding???
8) Post-1950 Translations: Should I bother? Well, the answer to that hinges on what everyone thinks about the AA-level translations. Brown played in no league higher than AA or its equivalent after 1948. On the other hand, even if the AA conversion rate is .85/.72, Brown’s 1953 Texas League season comes out to .272/.404 average and slugging. Given that he was already in decline by this point, and it doesn’t seem like a 38-year old who never walked could survive in MLB after a season with a .404 SLG.
   29. Dr. Chaleeko Posted: July 29, 2005 at 01:47 PM (#1507242)
Forgot to mention, these are based on a 162-game schedule.
   30. andrew siegel Posted: July 29, 2005 at 02:12 PM (#1507283)
The good doctor did a great job and those stats look like an accurate translation of what we have. However, I find it impossible to believe that someone with that few walks would have put up such amazing offensive numbers in the white majors. I'm fairly sure that he either would have learned how to take 40-50 walks per year or had the rest of his numbers decline precipitously when pitchers figured out how much of a hacker he really was. The problem, of course, if figuring out which.
   31. Dr. Chaleeko Posted: July 29, 2005 at 02:45 PM (#1507355)
Thanks, Andrew.

I'm feeling very unsettled about the 1937-1943 era. If the NAL was playing as tough as Gadfly's numbers suggest that it was in 1944 and 1945, Brown's pre-war translations could be overly conservative.

The numbers I've posted work out to a 119 OPS+ (with pitcher batting removed this time!)---all of it on the slugging side, of course. Actually, he was a 124 for SLG+, so the man had some serious power. Anyway, it all suggests that SFBWS (or more accurately basic RC) may be overstating his production a bit.

Once we put together the story of 1937-1943 and get a read on his defense, we'll be in good shape.
   32. Mark Shirk (jsch) Posted: August 02, 2005 at 02:07 PM (#1516789)
How good of a fielder was Brown? Could we expect to add 3-4 WS onto those batting WS numbers or not? His candidacy may rest on how good of a fielder he was, if he was a great fielder his peak could look good enough to get my support.
   33. Dr. Chaleeko Posted: August 02, 2005 at 03:00 PM (#1516883)
Purely a guess here, but I'd imagine his fielding was above average. He was athletic as evidenced by his playing SS for three years before moving to the OF. And he had plenty of speed, as evidenced by his frequent placement among team and league leaders in steals.

Brown mostly played CF until his mid to late 30s. So as a SS, I'm guessing he was at best an average SS, but more likely slightly below. As a CF, however, he was probably a little above-average. As a corner outfielder, he was probably well above average.

I had a little email exchange with Gary A. about Brown earlier this summer. Gary suggested that Brown would not likely have played in the infield in the major leagues, but that he would have been a very good corner outfielder in the majors. I have personally inferred from Gary's email that Brown would probably have been an average to slightly-above-average CF.

The names that come to mind are someone like Dwight Evans, Reggie Smith, Andre Dawson, Barry Bonds, Dale Murphy: guys with enough speed and fielding ability to play CF reasonably well, but who were outstanding corner outfielders instead.
   34. Mark Shirk (jsch) Posted: August 02, 2005 at 05:02 PM (#1517151)
So in terms of quantifying his value are we looking at a B+/A- outfielder in the Win Shares grades? That may seem low but remember that James lumps all outfielders together so most of the A grades are CFers.

If this is true I would guess an average of 3 WS per 1000 Defensive innings seems reasonable though maybe a little conservative. A peak of 3.4 or 3.5 would make some sense.

If I am correct I will put Brown behind Oms and at #29. That seems low but right now I can't justify having him above Pete Browning and Gavvy Cravath, who are #'s 22 and 27.
   35. Dr. Chaleeko Posted: August 02, 2005 at 05:15 PM (#1517168)
jschmeagol,

i think you're probably right about brown's outfielding.

as for his placement, however, i'm personally still waiting to see if Gadfly might have league batting information for 1937-1943. At this point, Brown's pre-War seasons look a little less consistently above the league average because I used a straight translation of .9 without any adjustment for NAL offensive levels.

The effect is to dampen the impact of that group of seasons, when in reality, Brown was all over the leader boards in all of them.

Short of having league data, I may well simply use the 1944-1945 data (which shows the league averages as lower than the post-war period) as a league reference and recalculate with that baseline. See post 21 and 25 in this thread for the basis of why I'm considering that adjustment.
   36. Jim Sp Posted: August 02, 2005 at 10:39 PM (#1518077)
Can someone give a summary of why Brown's trial with the Browns went so badly? Thanks.
   37. Dr. Chaleeko Posted: August 03, 2005 at 12:24 AM (#1518483)
JimSp, there's a little summation of what happened starting around post six and going through around post 12-15.
   38. Jim Sp Posted: August 03, 2005 at 02:16 AM (#1518972)
Yes, I don't feel like that really explains it. Bats aren't expensive or hard to find...
   39. Dr. Chaleeko Posted: August 03, 2005 at 02:51 AM (#1519098)
CONTEXT FOR WILLARD BROWN'S SEASON IN MEXICO

Using Cisneros, I've estimated team and league batting and slugging for the Mexican League in 1940, Brown's lone season there. I say estimated because, as always, Cisneros does not provide splits, and several players with a lot of PAs are split between teams.

This data is with pitchers removed.

The league hit .299 and slugged .420. Compare that to the NL (sans pitchers) for 1940, when it hit .272 with a .391 slg.

Brown hit .354 and slugged .524.

I would translate this to roughly .299/.400.

However, check out this team-by-team breakdown (remembering these are estimates)
NAME          AVG  SLG   AB
----------------------------
Veracruz     .315 .476  2445
Mexico City  .308 .442  2479
Torreon      .304 .426  2512
Santa Rosa   .295 .442   760
Monterrey    .287 .407  3136
Tampico      .276 .413   909
Nuevo Laredo .259 .386  2654
Chihuahua    .253 .343  1074
============================
LEAGUE       .289 .420 15942


Veracruz (where Willie Wells, Josh Gibson, Ray Dandridge, and Martin Dihigo all had great seasons) boosted offense to amazing levels, while Brown's home park in Nuevo Laredo appears to have tamped down offense.

Here's the same chart, but each team's batting average and slugging are expressed as a percentage of the league's AVg and SLG. In the fourth column is the team's winning pct.

NAME          AVG  SLG   AB   PCT.
----------------------------------
Veracruz     1.09 1.13  2445  .569
Mexico City  1.06 1.05  2479  .692
Torreon      1.05 1.01  2512  .500
Santa Rosa   1.02 1.05   760  .293
Monterrey     .99  .97  3136  .533
Tampico       .95  .98   909  .412
Nuevo Laredo  .89  .92  2654  .451
Chihuahua     .87  .82  1074  .268
==================================
LEAGUE       .290 .420 15942

The winning percentage should help us see whether a team's average and slugging are being influenced by team quality and/or park. In the cases of Veracruz and Mexico City, it seems obvious that team quality is a heavy influence on their average and slugging being well above the league norm. However, the rest of the list is a mixed bag, and Brown's Nuevo Laredo team seems to be a victim of park illusions in some measure.

I think that I will adjust Brown's percentages to reflect his home park, using a 95 park factor.

I have no other years of leaguewide information for the Mexican League, so I can't offer any specific data backing whether this is a repeated pattern with Nuevo Laredo, however, just looking at the team totals and the winning pcts for Nuevo Laredo teams in the Re-evaluation NgL Pitchers thread, the low ERAs and medicore winning pcts of Nuevo teams support the hypothesis that it was a tough place to hit and generate runs.
   40. Chris Cobb Posted: August 03, 2005 at 03:33 AM (#1519231)
This is great work on Willard Brown! Thanks, everyone!

I did as much as I could on Brown's MLEs before I went on vacation in the absence of any walks data and absent the Mexican, Cuban, and Puerto Rican data, so I should be able to factor in the additional data in the next couple of days.

My initial projections agree pretty well with Dr. Chaleeko's, though, so I don't believe that the full projections will alter the view of him.

There are two big questions that I'm going to have to consider in evaluating Brown:

1) Did the large NAL ballparks suppress Brown's power? If so, how much?

2) What to make of the lack of walks. We're dealing here with a contextual value issue that a formulaic translation simply can't address. No upper-echelon major-league hitter in the modern game has had walk rates that low over an extended career. Does that mean Brown wouldn't have been an upper-echelon hitter in the majors, or does that mean that he would have raised his walk rates to, say, the Joe Medwick level? He clearly didn't have an incentive to develop plate discipline in his league context: he was the best hitter just as he was. So his dominance in context needs to be weighed against the presence of a hole in his game that might have limited his value to certain contexts.

It's a similar problem to evaluating Cool Papa Bell's speed or Martin Dihigo's two-way play . . .
   41. Dr. Chaleeko Posted: August 03, 2005 at 12:58 PM (#1519550)
I wonder if Brown's walk rate in the Texas League is a little more indicitive of what he would have done at the MLB level: about six walks per every 100 ABs, or about 30-40 a year.

I just don't know.
   42. Mark Shirk (jsch) Posted: August 03, 2005 at 01:11 PM (#1519563)
One reason top echelon hitters start to walk more is that they get pitched around. Take Vlad Guerrero, he really isn't any less aggressive than he used to be but pitchers are much more careful to him. Even he can hold back on a ball two to three feet outside on occasion.

Also, how often would you expect Brown to be intentionally walked? Is this something we can calculate?
   43. Dr. Chaleeko Posted: August 03, 2005 at 07:09 PM (#1520219)
Here's some information I have no idea what to do with regarding Brown.

I went through Holway's 1936 batting average numbers for the rosters of the Western teams. I assumed that these rosters represent the bulk or close to all the ABs of the players on the six independent teams he gives stats for. Might be a bad assumption.

Anyway, I went through and took an educated guess at the actual number of ABs and Hs for each player by using the number of decisions as a guideline as well as Holway's AB/H totals for the top-five AVG leaders.

The point of all this was to figure out what kind of league context Brown was compiling all these .300 averages in.

So here's my guesstimate, with the cavaet that there's two or three guys for whom he gives no average, just an asterisk.
<pre>
TEAM RECORD AB H AVG
--------------------------------
K.C. 7-0 201 90 .448
Chicago 5-4 302 69 .228
Cincinnati 3-5 372 110 .296
Birmingham 0-2 36 6 .167
Memphis 2-0 65 23 .354
St. Louis 1-3 136 22 .162
================================
TOTAL 18-14 1112 320 .288

The non-KC teams batted .252.

I have no idea how to use this information, or whether to use this information at all, other than to damp down Brown's seemingly outstanding 1936 season. His .367 average, for instance, doesn't seem so impressive in light of his teammates' averages. But how much should I damp it down?
   44. Dr. Chaleeko Posted: August 03, 2005 at 07:44 PM (#1520299)
Eeeoowwwww, forgot the un-pre...

Here's that table again

TEAM     RECORD   AB   H  AVG
--------------------------------
K.C.       7-0   201  90 .448 
Chicago    5-4   302  69 .228
Cincinnati 3-5   372 110 .296
Birmingham 0-2    36   6 .167
Memphis    2-0    65  23 .354
St. Louis  1-3   136  22 .162
================================
TOTAL     18-14 1112 320 .288
   45. DavidFoss Posted: August 04, 2005 at 09:48 PM (#1523889)
Is Chris going to do MLE's for Willard here?

I'm going on vacation tomorrow. Maybe I'll just post pitcher-removed context information for the NL here tonight. I'm not sure if I've posted any as last as 1950 yet.
   46. EricC Posted: August 05, 2005 at 01:24 AM (#1524274)
The HoF favored NeL East stars from the late 1930s-1940s over NeL West stars. Was this a bias, or was the East substantially stronger than the West during this era? As much as I like to include league strength in my ratings, I haven't dared to attempt it for the NeL leagues.
   47. DavidFoss Posted: August 05, 2005 at 03:35 AM (#1524539)
1940s and 1950s National League pitcher-removed contexts:

1940 0.272 0.337 0.391
1941 0.266 0.337 0.375
1942 0.256 0.328 0.356
1943 0.265 0.334 0.360
1944 0.268 0.335 0.377
1945 0.273 0.343 0.377
1946 0.263 0.338 0.368
1947 0.274 0.349 0.407
1948 0.269 0.343 0.398
1949 0.271 0.344 0.405

1950 0.270 0.347 0.418
1951 0.268 0.342 0.406
1952 0.262 0.335 0.389
1953 0.274 0.345 0.428
1954 0.274 0.346 0.424
1955 0.266 0.337 0.421
1956 0.264 0.331 0.417
1957 0.269 0.332 0.416
1958 0.270 0.338 0.422
1959 0.270 0.336 0.418
   48. Dr. Chaleeko Posted: August 05, 2005 at 03:22 PM (#1524994)
David, thank you! Fortunately, I've just finally figured out how to get the pitcher-removed totals from the SBE....
   49. Dr. Chaleeko Posted: August 05, 2005 at 03:27 PM (#1525005)
WILLARD BROWN MLEs v 2.0

This time with input from Chris Cobb, so these have more of his official stamp of approval than the last batch.
YEAR AGE POS  PA   AB    H  BB   K  2b  3B  HR   TB  AVG  OBP  SLG OPS+
-----------------------------------------------------------------------
1935  20  SS 350  343  100   7  27  20   3  14  169 .293 .307 .492 109
1936  21  SS 685  664  220  21  53  35  16  33  386 .331 .351 .581 145
1937  22  SS 685  664  222  21  53  32   8  28  354 .334 .354 .533 136
1938  23  CF 628  603  193  25  48  18  12  23  304 .320 .348 .505 130
1939  24  CF 575  552  183  23  44  35   4   7  247 .332 .359 .448 113
1940  25  CF 611  586  184  24  22  24   4  10  246 .314 .341 .420 107
1941  26  CF 625  600  208  25  48  40  13  18  328 .346 .372 .546 154
1942  27  LF 650  618  204  33  49  17   9  18  293 .330 .364 .475 142
1943  28  CF 365  347  112  18  28  16   3  16  182 .323 .357 .524 151
1944  29  CF 545  516  170  29  41  28   6  19  267 .315 .337 .491 129
1945  30  CF 545  514  169  31  41  34   6  25  290 .313 .336 .536 138
1946  31  LF 620  583  194  37  47  51   6  23  325 .333 .373 .558 160
1947  32  CF 650  611  202  39  49  34   7  37  364 .331 .371 .596 151
1948  33  CF 620  583  179  37  47  33   5  35  327 .307 .349 .561 141
1949  34  LF 620  583  183  37  47  24   5  20  277 .314 .355 .476 119
1950  35  LF 500  475  140  25  38  19   3  14  207 .294 .327 .435  96
1951  36  LF 250  240   73  10  19   6   2   3   90 .304 .332 .376  88
=======================================================================
TOTAL       9523 9081 2936 443 702 465 112 344 4566 .323 .355 .503 131

NOTES
Same basic premise as last time, only this MLE looks a little different for a few reasons:
1) 1937: I’ve decided to eliminate the CWL season from my calculations. It’s so incongruous with the rest of Brown’s career that it hardly feels like part of the same career., and it makes 1937 look like his worst season instead of a season well within the context of the surrounding seasons.

2) 1939–1949: These years have been adjusted to reflect Brown’s NgL stats relative to his own league before conversion to the MLB setting. Thanks to Gadfly for the numbers covering 1944–1949. Based on a combination of his characterization of the NAL as using a dead ball from 1942 through parts of 1946, I decided that 1942 and 1943 would be similar enough to 1944–1946 that I could extrapolate league AVG and SLG backwards. Looking at Brown’s stats and his placements on the leaderboards, it appears that 1939 and 1941 also conform to this same deadened-ball pattern (whether or not they actually used a deadened ball, the numbers come out looking that way). For 1940, I just took an evening and blitzed through Cisneros to get the league AVG and SLG for the MxL that year.

3) Park Factors: These have not been adjusted except for in 1940, in which I guessed that Nueveo Laredo played as a about a 95 PF for hitters based on not only that season’s numbers but other seasons as well.

4) Conversion rates: On Chris Cobb’s suggestion, I’ve revised my AA-level conversion rates. .80/.64 was simply too low. He recommended .87/.75, and so that’s what I’ve gone with for any league that might have been at a AA level.

5) 1951: After revising my conversion rates it became obvious that Brown’s terminal season was more likely to be 1951 than 1950. That makes his last three seasons 119, 96, 88 for OPS+ which doesn’t seem unreasonable given the swift denouement of similar hitters like Medwick and Rice who both retired at age 36. I have no reason to believe that he would have continued in the majors thereafter. Since this falls under the same general heading, Brown played full time through 1955 or so, but I declined his PAs over his final two seasons as a matter of mirroring his skills erosion.

6) Walk rates: I took a tip from Chris Cobb and pegged Brown’s MLE walk rate to Joe Medwick’s. Not exactly the same, actually a little lower, but so that it climbs from 2% of all PAs in his rookie season to 6% around age 30–34, then going back down at the end to 5% and 4%. He doesn’t walk as much as Medwick, nor as little as in the previous MLE.

7) Military Years: I just included them in there this time. Brown saw no skills erosion after his war years because he was playing ball during that time. I have simply averaged his nearest three surrounding seasons to generate 1944 and 1945. There is no wartime deduction, and you may apply any discounts you want to.

REMAINING ISSUES
1) The biggest thing is home park. I have made no systematic adjustment for home park. It appears that this is probably DEFLATING Brown’s HR totals in the early 1940s while also possibly inflating his batting averages and doubles and triples. Based on Gary’s information, I’d say some of those triples or doubles should be shifted into the HR column, though I have no idea how many as a percentage.

2) League contexts for 1935–1938, 1950, 1951, and the PRWL. League contexts for these seasons could drastically change how we view the beginning and end of Brown’s career. If the Canadian Border League and PRWL were pitcher’s leagues in the early 1950s, then his terminal year could be pushed backward.

3) Win Shares: I’m hoping to post these later today, if not, I’ll defer to Chris Cobb.

4) Of course the 2B, 3B, and HR totals are estimates, best fits for the Total Bases generated by the SLG conversion rates and based on what we know about his 2B, 3B, and HR rates from Holway's Complete Book and the numbers Gadfly presented earlier in the thread.

ONE BIG CAVAET
It is my belief that as these MLEs stand, they UNDERRATE Brown in general. I think what you’ve got here is probably a baseline, and that there’s much room for more information to change how we think about Brown. If Gadfly can offer up league contexts for the PRWL and Chris has his 1937-1938 NAL median-average based conversions handy, we’ll see a more accurate picture. For now, however, I think this expresses an excellent player, consistently valuable, multifaceted, and well positioned to have his case examined very closely for the HOM.
   50. karlmagnus Posted: August 05, 2005 at 06:14 PM (#1525314)
Sorry, I don't buy this one, very far from it.

1) You don't get to eliminate the 1937 bad news and smooth over it.

2) I don't believe Brown would have had anything like 2935 hits, or an OPS+ anything like 131.
3) Also, you've added to his walk rate without subtracting anywhere else.

4) Also, his career after 1949 was in increasingly minor leagues, which got further and further away from ML.

5)Also, I don't see how he can possibly have an OPS+ for 1947 when he spent a month of it in the actual majors with an OPS+ of 22.

ENRON ACCOUNTING!!! If Chris signs off on these, I shall re-evealuate all his other projections and knock 20% off them, because it will prove that we are being hopelessly rosy-eyed in our methodology.
   51. karlmagnus Posted: August 05, 2005 at 06:15 PM (#1525316)
5) should read "possibly have an OPS+ of 151 for 1947. Sorry.
   52. Dr. Chaleeko Posted: August 05, 2005 at 07:09 PM (#1525436)
Karl,

1) I think everyone's going to have to take 1937 on their own and decide for themselves. He was 8/55 on the island. This and his MLB sample are the only instances where he failed to hit at all.

2) The reason Brown gets so many hits is that he doesn't walk very much. He just has more opportunity to get hits.

3) I maintained his PAs and subtracted his walks to get his ABs all along, so the number of PAs, hits, walks, TBs, xBHs, and Ks should be accurate. Please let me know if you see any specific instances where there's a discrpancy between v1 and v2 in terms of PAs and ABs.

4) I adjusted the NAL down to AA-level in all seasons after 1948, and took the CBL as a AA-level league because leagues like this one and others were havens for NgL refugees as the NAL fractured in the 1950s. All PRWL seasons are at NAL levels, per Gadfly's recommendation, until the late 1940s when they slide down to AA leagues and a lower conversion rate.

5) The 1947 OPS+ includes his MLB time which was a paltry 67 at bats. We know from Gadfly that he had about 235 ABs in the PRWL that season. In addition, I used an extremely conservative number of ABs for his NAL play that year, just 107, representing about one-half to two-thirds of the ABs he accumulated in the surrounding NAL seasons. There is no xBH information available for 1947 in the NAL, so I based my estimates on his xBH distribution rates for seasons we know about. Combining that with the xBH distributions, AVG, and SLG for his AL and PRWL seasons, you get a line of .331/.596 AVG/SLG. This is in line with his surrounding seasons of .333/.558 and .307/.561.

ENRON: You should as always feel free to discount any posted projections as you see fit. But I think this is capturing the kind of player that the information above in this thread suggests. Look at his placements on the leaderboards. He was top five or higher in everything every year.... The OPS+ of 131 career and the seasonal OPS+s suggest that he would be close to the top five OPS+ scores each year, though he would only be among them a few times.

Also keep in mind that a 131 OPS+ is not exactly rarified air. Rafael Palmeiro's career OPS+ going into 2005 was 132, and 132 is not within the top-100 OPS+s ever (per bb-ref). In fact, the top 100 OPS+s end around 136.
   53. karlmagnus Posted: August 05, 2005 at 07:30 PM (#1525493)
Sorry, now being somewhat more temperate:-)).

Did you not add walks without subtracting any hits? If so, doesn't that artificially inflate his OPS+? I can't accept that he could just have increased his walks without losing out somewhere else, presumably hits, or possibly power. There's a reason Nomar and Shea Hillenbrand play the way they do, it's not just to annoy wimpy intellectual sabermetricians.

What does is his OPS+ look like if you just assume the walk rate he actually had? Surely that would be more accurate -- the other must surely be a distorted view of his value to the team.

We're probably talking 128 rather than 131, and the figures aren't accurate enough to take account of this, but still...
   54. Dr. Chaleeko Posted: August 05, 2005 at 08:00 PM (#1525555)
Karl,

There's enough adjustments and changes since v1.0 of his MLEs that although it may appear superficially as though I've not removed any hits or ABs in favor of walks, I have. In particular, the raising of my arbitrarily created AA conversion rate to conform with Chris's and Gadfly's previously established rate (a 7% hike), raised Brown's batting average considerably.

It's entirely possible that what you are reacting to, however, is really his playing time. I only found evidence of his missing extended time in 1943 (and of course the war years). He was extremely healthy otherwise. That said, I think it's possible that some PA estimates could be lowered. I'm open to suggestions for which seasons should be tempered.
   55. Dr. Chaleeko Posted: August 05, 2005 at 08:13 PM (#1525588)
WIN SHARES FOR WILLARD BROWN BASED ON MLE V 2.0
YEAR     bws    fws     ws
-----------------------------
1935     8.8    3.3    12.1
1936    24.0    6.5    30.5
1937    21.9    6.5    28.4
1938    19.5    3.8    23.3
1939    14.7    3.5    18.2
1940    14.3    3.7    18.1
1941    29.5    3.8    33.3
1942    21.5    3.9    25.4
1943    13.7    2.2    15.9
1944    16.4    3.3    19.7
1945    17.4    3.3    20.7
1946    31.0    3.8    34.8
1947    24.1    3.9    28.0
1948    20.3    3.6    23.9
1949    17.1    3.5    20.5
1950     9.3    2.5    11.8
1951     3.7    1.0     4.7
============================
TOTAL  306.9   62.3   369.2

Notes: For the 1935-1937, Holway has Brown at SS, so I just left him there and used SFWS to generate his WS (one FWS for every 25 games). Then I moved Brown to the outfield and gave him 3.0 WS/1000 innings until the last four seasons, after which I declined him year by year to 2.9, 2.75, 2.5, and 2.0.

These are all unregressed figures, by the way.
   56. DavidFoss Posted: August 05, 2005 at 08:14 PM (#1525589)
1) The biggest thing is home park. I have made no systematic adjustment for home park. It appears that this is probably DEFLATING Brown’s HR totals in the early 1940s while also possibly inflating his batting averages and doubles and triples. Based on Gary’s information, I’d say some of those triples or doubles should be shifted into the HR column, though I have no idea how many as a percentage.

You have to be careful, here. Power numbers were down during the war (and in '42 and '46 as well so its not just that the power hitters were in the service). If I recall correctly, they even switched to a balata ball in MLB one of those years. Anyhow, the point being that if you think Brown's HR totals are a bit low in the early forties, remember that the NL context's HR totals are low as well.
   57. Dr. Chaleeko Posted: August 05, 2005 at 08:20 PM (#1525601)
David,

That's one of the reasons I haven't made any park adjustments: i just don't know how to account for it all.

Just an FYI, right now Brown would be just behind Earl Averill for me in the CF group.
   58. Chris Cobb Posted: August 05, 2005 at 08:29 PM (#1525634)
Hmm. It's hard to know how to comment here. I see Dr. Chaleeko's projections for Willard Brown as plausible. They are not the same as mine at this point (they are a bit higher), but 1) mine are not finished and 2) I'm not even sure yet what I will do with the Puerto Rican data. So I can't say how mine will look relative to his. I need to re-check my league offense and competition factors as part of my fine-tuning of my projections and decide what to do with Puerto Rico.

I do not endorse his projections but neither do I reject them as wildly inaccurate. Voters should use their own judgment.

Brown's case is very difficult for the translator because of two features of his career that have no direct parallel in a major-league career: the walk rate issue and the multi-league issue.

I need to do more study before I comment fully on the walk rate issue, but here's a start. On the one hand, I think Karlmagnus is right to want to see what Brown's numbers would look like with a straight projection of his walk rates, and I aim to provide those numbers eventually. On the other hand, I think that either a) Brown would have been _much_ less successful in the majors than he was in the NeL or b) Brown would have learned to take a few walks. Learning to take walks would not, in the longer term, cost him batting average or power. Take a look at the dance of power, plate discipline, and batting average in Sammy Sosa's career. There are short-term losses as plate discipline rises, but long-term gains. I believe that, as the game was played in the NeL, Brown had little incentive to develop plate discipline. In the majors, he would have had a strong incentive, so I am inclined to think that the best projection of him will show him developing at least the minimal level of plate discipline necessary to be a successful major-league hitter. I need to study more career patterns before I will be satisfied with this view or decide I have to modify it.

Did you not add walks without subtracting any hits? If so, doesn't that artificially inflate his OPS+?

I believe that Dr. Chaleeko added walks without lowering Brown's _batting average_, which is not the same as adding walks without subtracting any hits. His hit total will drop because he has fewer at bats, but his OBP would rise.


The multi-league issue is a bit clearer. Brown's hitting appears to have dropped off very significantly in most cases in which he was switching to a new and more difficult league: his numbers are terrible in Cuba in 1937 and in the AL in 1947, and they are significantly below his career norms in Mexico in 1940, and there is a drop in his power numbers when he enters the Texas League. Gadfly has pointed this out and suggested that it tells us something about Brown's hitting style that he needed a fairly lengthy period of adjustment--longer than he got in Cuban or in the Majors--to learn the pitchers in the league.

We would expect that a major-leaguer might have to adjust like this twice in a career: as a rookie and when traded from one league to another. Brown, however, has four adjustment periods in his record (and it would be five if we had data for his rookie season with KC). Those adjustment periods are part of what he did, but should they all figure fully into major-league equivalents? Looking at the issue as a translator, I doubt that they should, so I am ok with Dr. Chaleeko deciding to leave Cuba 1937 out of Brown's record.

As a voter, however, Brown's difficulty in making adjustments quickly gives me some concern, so I'm inclined to rank him on the low side of what the projections that I ultimately accept show his MLEs to be.

I'm very confident at this point that Brown's batting MLEs should lead to an OPS+ of somewhere between 120 and 135, but that's not a narrow enough range yet to provide a satisfactory basis for ranking him.
   59. Dr. Chaleeko Posted: August 05, 2005 at 10:24 PM (#1525937)
Chris, the thing that keeps jumping out at me is Brown's total dominance of the leaderboards. There's just no other player in the west from 1937-1948 who can claim to be as dominant over the NAL as Brown. Maybe that's somewhat subjective, but it leads me to believe that the projection I've ended up at is more likely than my first one.

Also, the first projection essentially does take his walk rates and estimate them based on a straight line from 1942 (when he walked in 2% of ABs) and the Texas League (where he walked in 5-6% of ABs). The main difference is that in the second projections, I made his walk rate increase more quickly over time to coincide with his power peak.
   60. Chris Cobb Posted: August 06, 2005 at 03:32 AM (#1526731)
Chris, the thing that keeps jumping out at me is Brown's total dominance of the leaderboards.

I don't think your MLEs for him are necessarily too high. It may be right on target. Since folks seem interested in whether or not I "endorse" the estimate, I'm being cautious until I understand my own estimates better that I do now.

But the leaderboard evidence cuts two ways. Who was Brown's competition? Let's look at 1937-1942, Brown's early period of stardom. In the NNL during these years, Josh Gibson and Buck Leonard were in their primes, Mule Suttles was still going strong through 1939, as was Willie Wells. Wild Bill Wright had some great seasons in there also. Monte Irvin burst onto the scene in 1939, Larry Doby in 1942. That's 4 HoMers, and 5 HoFers, and Gibson, Suttles, Leonard, Irvin, and Doby are all power hitters with very strong bona fides.

In the NAL during these years, there was only one HoM or HoF hitter in the NAL: Turkey Stearnes, who was in the final 3 years of his career.

If one is concerned that the NAL might be unfairly overlooked, consider this: during these seasons, Brown appeared on the NAL leaderboards 5 out of 6 years: he missed in 1940 because he was in Mexico. Only 3 other players appeared in more than two seasons: Neil Robinson (5), Buck O'Neil (4), and Turkey Stearnes (3). Robinson is an unfairly neglected star, though almost certainly not a HoMer. O'Neil was a good hitter, but no one ever called him a power hitter. Stearnes was a great power hitter, reaching the leaderboards in each of his last three seasons, 1937-1939.

In the east, there's much more consistency in the leaderboards. Even with Mexican sojourns by some, Gibson, Leonard, Wright, Jim West, Ed Stone, Sammy T. Hughes, Suttles, Wells, Lenny Pearson, Vic Harris, and Irvin all appear on the leaderboards in at least 3 seasons in this six-year stretch. There's just much stronger evidence of a substantial pool of highly accomplished hitters for the NNL than there is for the NAL.

So Brown's domination of his leaderboards is surely helped by the fact that there were fewer true power hitters in the western league. We don't have the capacity to do any quantitative measurement of league strength, but I think Brown's black and gray ink totals have to be interpreted in light of evidence that he was dominating a league that, if not weaker than the NNL in overall level of competition, was certainly less well-stocked with power hitters. He was undoubtedly the best power hitter in the NAL. But how good he needed to be for that to be true remains an open question.
   61. Mark Shirk (jsch) Posted: August 06, 2005 at 04:36 PM (#1527039)
Plus, if you aren't walking much you can rack up some impressive counting stats.
   62. Chris Cobb Posted: August 07, 2005 at 04:32 PM (#1528278)
Well, I've finished my projections for Brown. I'll post them in a moment, but before they appear I want to note that I found that, after being cautious about "endorsing" Dr. Chaleeko's projections above, mine come out quite similar to his when I use similar assumptions about plate discipline. I've also run projections that simply use a straight translation of Brown's walk rates to the majors, Karlmagnus was correct that we're looking at about 3 points of career OPS in the difference between Brown walking like Medwick and Brown walking like Brown. Not a huge difference, but not a meaningless one.

In the end, I think the differences between Dr. Chaleeko's projections and mine are products of his data set including Puerto Rico and minor-league play: our handling of the NeL numbers is quite similar. With that preview, here follows a long post on Willard Brown.
   63. Chris Cobb Posted: August 07, 2005 at 04:37 PM (#1528286)
Willard Brown MLEs for Negro-League Play

My projections for Brown are for his Negro-League and Mexican League play only. I have not included his 1937 Cuban play, his Puerto Rican play, his 1947 major-league play, or his minor-league play. However, for 1947, I have left out a portion of his season equivalent to his major-league stint, so those stats can simply be added into the MLEs: I’ll give the combined 1947 line and its effect on career data later.

First, to show clearly the differences between a straight translation on walks and a translation projecting up to the Joe Medwick level of plate discipline, here are the career lines for each projection

1935-49 1841 g, 7657 PA, 407 walks, 2306 hits, 3574 tb, .318 ba, .354 obp, .493 sa, 132 OPS+
1935-49 1841 g, 7657 PA, 289 walks, 2344 hits, 3637 tb, .318 ba, .344 obp, .493 sa, 129 OPS+

Here’s a full projection for Brown using the Medwick plate-discipline model.
Year Team EqG  PA   BB  Hits TB   BA   OBP  SA   OPS+
1935 KC   135  567  26  194  299 .359 .389 .554  150
1936      148  622  30  200  309 .338 .370 .522  138
1937      154  647  32  207  340 .337 .370 .552  147
1938      150  630  32  192  292 .321 .356 .489  130
1939      139  584  30  182  260 .328 .363 .470  122
1940 MeL  144  605  32  177  243 .309 .346 .424  111
1941      145  609  33  176  262 .305 .342 .454  123
1942      151  634  36  169  266 .282 .323 .445  123
1943      154  647  36  189  300 .310 .348 .491  141
1944 Military Service
1945 Military Service
1946 KC   146  613  34  185  310 .320 .358 .535  151
1947      114  456  26  125  197 .292 .333 .459  108
1948      133  532  32  157  275 .314 .355 .549  141
1949      128  512  27  152  221 .314 .350 .456  114
         1841 7657 407 2306 3574 .318 .354 .493  132


Here’s a full projection for Brown using the straight-translation model.
Year Team EqG  PA   BB  Hits TB   BA  OBP   SA  OPS+
1935 KC   135  567  14  199  306 .359 .375 .554 146
1936      148  622  17  205  316 .338 .356 .522 134
1937      154  647  19  212  347 .337 .357 .552 143
1938      150  630  20  196  299 .321 .342 .490 126
1939      139  584  19  185  266 .328 .350 .471 119
1940 MeL  144  605  21  181  250 .309 .334 .428 109
1941      145  609  23  179  267 .305 .330 .455 119
1942      151  634  25  172  272 .282 .310 .446 120
1943      154  647  27  192  305 .310 .338 .491 138
1944 Military Service
1945 Military Service
1946 KC   146  613  29  187  313 .320 .352 .535 150
1947      114  456  22  127  199 .292 .327 .459 106
1948      133  532  27  158  277 .314 .349 .549 140
1949      128  512  27  152  221 .314 .350 .456 114
         1841 7657 289 2344 3637 .318 .344 .493 129


My projections are not far off from Dr. Chaleeko’s. Some differences:

1) I’ve used Macmillan’s 4-game data set for 1935.
2) My playing time projections are a bit lower throughout.
3) Mine do not include Brown’s strong Puerto Rican play.
4) Mine do not include Brown’s poor ML play in 1947.
5) Mine do not include Brown’s minor-league play after 1949.
6) Mine are regressed.
7) My league offensive-level estimates are probably slightly different.
8) Mine do not include war-credit estimates.

I’ll conclude with a few steps towards integrating the two models.

First, here’s Brown’s 1947 season in my projections with his major-league play included. Medwick-level plate discipline is first, straight translation second.
47 KC/Stl 135  523  26  137  215 .276 .312 .433  96
47 KC/Stl 135  523  22  139  217 .277 .308 .433  95


Second here’s how Brown’s career line looks with this 1947 season and his 1950 and 1951 seasons as projected by Dr. Chaleeko added in. Again, Medwick-level plate discipline is first, then straight translation:
total    2050 8474 442 2531 3889 .315 .351 .484 127
total    2050 8474 324 2569 3952 .315 .341 .485 125


Concluding analysis.

Since Brown was born in 1915, I think it’s right to include at least 1950 and 1951 as part of his MLEs. Dr. Chaleeko’s projections for those seasons look reasonable, so I’ll be using them myself.

Brown’s early peak is problematic. I have lowered the translation factor for 1935 and 1936 to .87/.75: that may not be far enough. It is likely, however, that Brown (like Medwick, I think) was highly successful early in his career before pitchers learned that he really would swing at anything. I wish we knew more about the circumstances of Brown’s play in these early years.

Brown definitely deserves war credit: it looks very much like he missed two years out of his peak during the war: 1943 and 1946 are two of his best mature seasons.

The differences between Dr. Chaleeko’s peak projections and mine shows the effect of regression pretty well.

In the end, the most important difference between my projections and the good Doctor’s is the use of the Puerto Rican stats, since Brown’s monster 1947 performance there turned what was previously a poor season by his standards into a good one. I think the inclusion of Puerto Rican play in one’s assessment of Brown is quite justifiable: it was a significant part of his real value and his real legacy as a player. The question is whether the translations are accurate, since we don’t know exactly what the level of competition and the level of offense were in those leagues. Still, his 1947 season there was the best anyone has ever had in Puerto Rico, so we can be confident that it would boost his MLEs in any case.
   64. karlmagnus Posted: August 07, 2005 at 07:29 PM (#1528674)
Chris, you end up with Oms looking similar but a tad better -- is that a reasonable summary?

Both are just off in my view (which balances Wells, Ray Brown and Beckwith who I regard as just in.) -- high 120s OPS+ and about 2500 hits.

I don't think it valid to use the "Medwick" projections, since he would have had to become a different hitter to achieve them.
   65. Chris Cobb Posted: August 07, 2005 at 07:50 PM (#1528729)
Chris, you end up with Oms looking similar but a tad better -- is that a reasonable summary?

Karlmagnus,

That's a reasonable interpretation of the data. They are close enough as hitters and in career length that where they stand relative to one another will change depending on how a voter handles peak vs. career, WWII credit, credit to Oms for play before 1921, credit to Brown for Puerto Rico, the plate discipline question for Brown, and fielding (I see Oms as better defensively).

I think I will prefer Oms to Brown, but I'm not sure yet. On my ballot, I expect Brown will be in the 8-25 range (Oms to Medwick). I rank this outfielder type higher than you do, generally, so I'd guess that he'd land in the 21-42 range (Oms to Medwick) in your rankings.
   66. karlmagnus Posted: August 07, 2005 at 07:57 PM (#1528753)
Chris, I'm impressed, you can even predict my ballot placements! On my tentative 1958, he's #25 (Oms is 19, Medwick 41)
   67. Chris Cobb Posted: August 08, 2005 at 03:10 AM (#1529452)
I don't think it valid to use the "Medwick" projections, since he would have had to become a different hitter to achieve them.

Not necessarily. He might just have had to play under conditions that would have induced him to take walks.

Here's the problem with doing a straight translation of Brown's walk rates into the context of the National League, 1935-1951:
nobody who was much of a hitter in a career of length walked as infrequently as Brown would under this projection.

I have tried to track down every major-league player who played the majority of his career between 1925 and 1962 (Brown's career plus ten years on each side) and who got 2,000 or more hits. There are 51 players in this group, if I've found them all.

In this group, the lowest walk rate per plate appearance is Lloyd Waner, at .0504. The second lowest is Joe Medwick, at .0537. Both were free swingers with the bulk of their careers in the 1930s NL, which was the lowest walk rate league in this period.

Brown's straight translation walk rate is .0383, 25% lower than Lloyd Waner. When a projection gives a figure on either the high or the low side of things that doesn't fit any extant historical model, that raises a big flag of caution for me in accepting that projection as accurate. I'm not saying it can't be, but it calls for careful scrutiny.

So there are three possibilities here:
1) Brown would have been a unique hitter in his time
2) Brown would not have achieved greatness in the majors due to lack of plate discipline
3) Brown would have increased his plate discipline at least minimally (as per Joe Medwick) in major league playing conditions.

Medwick's walk rates during his first two full seasons were .0415 and .0325. These are down in Brown territory, but his career rate climbed higher. Did Medwick have to become a different hitter to achieve this incremental improvement in his walk rate? I'm doubtful of that.

Lloyd Waner, on the other hand, presents a different picture. His walk rate stayed flat through his career, though this may be influenced by the drop in the NL walk rate after 1930. Still, his best years as a hitter are his first three, at ages 21-23. He achieved 3 of his 4 top OPS+ marks in those three years. This fits the projection we have for Brown, who, it appears, never raised his walk rate very much. He, unlike Little Poison, was a power hitter, who matched his early OPS+ peak in several later seasons when his power production peaked.

The one piece of this picture that I haven't yet investigated is hitters from 1935-1950 who really did walk at the rate Brown did. Who were they? Is there anybody? What did their careers look like? I'll report on that when I have some data. In the meantime, I hope this account explains why it is reasonable to project Brown for more walks than a straight translation would indicate.
   68. Chris Cobb Posted: August 08, 2005 at 09:27 PM (#1530615)
Here’s some more context on Willard Brown’s walk rates. For 1935, I tried to identify all non-pitchers who had a walk rate of less than 5 walks/100 PA in more than 100 PA. Since the AL was a much higher walk league than the NL, it’s not surprising that there were 6 AL players who fit this profile, while 21 NL players did. I then looked at the career profiles of these 27 players to see what they suggested about whether Brown could have been a 125 career OPS+ hitter in the majors with a .0383 rate of walks/PA, the straight translation rate for Brown. (The Medwick-type rate is .0532, slightly lower than Medwick’s .0537).

Of these 27 players, only five had a career walk rate lower than this projection for Brown. Only one of the five was ever a regular: Woody Jensen, who was a starting outfielder for three years for the Pirates. His walk rate, .0241, was the lowest of any player who stuck in the majors for more than a season. His career OPS+ of 84 was also the highest in this group.

Seven other players had walk rates below .05 (which I’ll call the Lloyd Waner line), but higher than Brown’s .0383. Six were ss/2b/3b types, who had good careers despite anemic OPS+ totals: 81, 78, 74, 72, 63, and 51. They have little in common with Brown. The only player at all close to him is the seventh in the group, Gee Walker, who played for 15 years, mostly as a starter, accumulating 1991 hits and a 99 OPS+ despite a walk rate of .0457. His career probably would have been shorter if not for WWII, however.

Four of the 27 players had career OPS+ scores of above 100, and all of them had higher walk rates than Medwick. These five were Buck Jordan (career .0566 walk rate, 104 OPS+), Pie Traynor (.0569, 107 OPS+), Freddie Lindstrom (.0547, 110), and Ernie Lombardi (.0677, 125). Jordan was a mediocre player in the middle of his career who had a down year on walks, Traynor and Lindstrom were declining fast, and Lombardi was just about to take a significant step forward in plate discipline.

It’s impossible for this sort of study to be conclusive about anything, though if I were to be as rigorous as possible I would take similar surveys at five-year intervals for the rest of Brown’s career, collecting players in 1940, 1945, and 1950 for further comparisons. But I don’t think these would show anything significantly different.

These data suggest that in this era, 1925-1960:
1) it was very rare for a starting position player to have a walk rate below .04 (1 example), and there is no example of such a player lasting long as a starter.
2) it was rare for a starting position player to have a walk rate below the Lloyd Waner line, and generally it was only good defensive players at high-defense positions that could sustain a career with this sort of rate.
3) it did not happen that a true power hitter would have a walk rate below the Lloyd Waner line over a career. Either the player’s lack of plate discipline prevented his power from developing, or the player’s power led pitchers to pitch around him to the extent that a .05 walk rate or better was practically unavoidable.

I conclude that if we are to try to translate Willard Brown’s play into a major-league context, his walk rate needs to be raised above the rate that he maintained in Mexico and the Texas League, or we need to conclude that he could not have been an HoM player in the majors.

As the best power hitter in his league, playing for a team that consistently dominated its competition, in a league where walk rates were generally low, and being of an easy-going (not to say lazy) disposition, it’s clear that Brown never had a serious incentive to develop plate discipline until it was too late, when he got his chance at the majors at age 32. It’s possible that he couldn’t have done it. But he was so clearly loaded with talent, it’s hard for me to believe that he couldn’t have done it if he had had a serious incentive to do so.
   69. karlmagnus Posted: August 09, 2005 at 12:52 AM (#1531086)
Fine, but if you raise his walk rate, you should do so while maintaining his OPS+ constant. Otherwise you're just adding value that didn't exist.
   70. sunnyday2 Posted: August 09, 2005 at 01:36 AM (#1531277)
Just back from 12 days in the Maritimes.

First thread 1957 ballot--then Willard Brown. Very interested to see what you guys came up with on him.

From 1957, very surprised at the votes Alejandro Oms is getting with his 125 OPS+ in a more or less average length career, not short mind you, but not long.

I need some time, probably a lot of time, to digest the Brown work-up, but knowing that the method is the same as that for Oms, I don't see much chance that Brown wasn't better than Oms.

I do know that that OPS+ 22 in the bigs is a major red herring (or herren, as they say down east). Throw that out, karl.
   71. Dr. Chaleeko Posted: August 09, 2005 at 04:09 PM (#1532352)
One other quick note. Any WS analysis will probably come out a smidgen light because stolen bases were a part of Brown's game. Without his success rates in that department, it's hard to know how much they would influence his WS totals, but I'd guess there's a few WS out there that he might have gained by the steal. Probably not enough, however, to change our picture of him.
   72. Chris Cobb Posted: August 11, 2005 at 12:41 AM (#1537485)
Willard Brown Estimated Win Shares

These are all projected for 154-game seasons.

The batting win shares are based on the Medwick-level plate discipline projections from post 63 above, except for 1950 and 1951, which are based on Dr. Chaleeko’s projections in post 49. If you believe those are too high, I’d suggest docking the bws by about 4%.

The fielding win shares are based on an estimate that Brown was a B- defensive outfielder, capable of playing centerfield at a somewhat below average level while being a good corner outfielder. I treated him as a D shortstop for 1935-7.
Year EqG  PA   bWS   fWS  Total
1935 135  567  23.4  4.3  27.7
1936 148  622  21.9  5.2  27.1
1937 154  647  27.9  4.8  32.7
1938 150  630  19.9  4.4  24.3
1939 139  584  16.4  4.0  20.4
1940 144  605  15.4  3.4  18.8
1941 145  609  18.9  3.8  22.7
1942 151  634  19.3  3.7  23.0
1943 154  647  27.1  4.3  31.4
1944 Military Service    
1945 Military Service    
1946 146  613  27.3  3.2  30.5
1947 114  456  10.7  2.9  13.6
1948 133  532  19.1  3.4  22.5
1949 128  512  13.8  2.7  16.5
1950 125  500   9.5  2.3  11.8
1951  68  250   4.0  1.0   5.0
    2034 8407 274.6 53.4 328.0


Two other projections of Brown's win shares from Dr. Chaleeko appear in post 55 and post 28 above. Although there's quite a bit of variation in how individual seasons get projected, the career projections are quite consistent.

Incidentally, win shares 1950 and 1951 in this post are based off of exactly the same data as Dr. Chaleeko's in post 55. I end up with 9.5 and 4.0, while the good doctor has 9.3 and 4.7. Pretty close, though I am higher in both cases.
   73. Dr. Chaleeko Posted: August 11, 2005 at 01:08 PM (#1538222)
Chris, did you use regression in your WS?
   74. Chris Cobb Posted: August 11, 2005 at 01:33 PM (#1538245)
Well, I used regression in calculating the MLEs from which the win shares were derived. I used the same 5-year rolling base that I use in all the projections, except that I tweaked the formula around the war years: some of the seasons before and after the war have a 4-year base.

Why do you ask?
   75. Dr. Chaleeko Posted: August 11, 2005 at 01:37 PM (#1538249)
Just curious, and then I realized after I'd posted that I just should have looked up at your methodology notes above. Sorry 'bout that.
   76. Carl G Posted: August 11, 2005 at 02:42 PM (#1538363)
Should Willard be considered a CF or Corner outfielder? I'm just trying to get an idea of which group to compare him to.
   77. sunnyday2 Posted: August 11, 2005 at 02:57 PM (#1538390)
His raw record shown in post #11 includes a position designation in 11 different "seasons," meaning a season/league (he played 2 "seasons" in many calendar years). I see 6 "seasons" at CF, 4 in LF or OF, and 1 at SS. There are other "seasons" where no position is designated.

I wonder if Al Simmons is a fair comp in terms of his utilization and maybe his defensive ability (setting aside his play at SS).

Or if you like the SS-CF part, you've got Mickey Mantle, Harvey Kuenn and Robin Yount, with Larry Doby as a less specific comp as well. He stole some bases so apparently he could run in CF.

Then there's the Andre Dawson, Dale Murphy, Kirby Puckett group, too. As a "hacker," Puckett is an interesting comp.
   78. Mark Shirk (jsch) Posted: August 11, 2005 at 10:00 PM (#1539874)
Somewhere above Dr. Chaleeko made mention that Brown woudl either have been an OK CFer or a very good corner guy. I have him pegged as a very good corner outfielder for my position considerations, though I guess it is possible that he would have played CF in the Majors. Lord know the Yankees could use him even now.
   79. KJOK Posted: August 12, 2005 at 06:01 AM (#1541041)
Don't know if it would have made any difference with the MLE's, but I just realized that Brown's stats are also in "The Negro Leagues Book", although unfortunately no walk data.
   80. sunnyday2 Posted: August 13, 2005 at 05:12 PM (#1543836)
Just a quick summary of Willard Brown BA as reported above. I don't know whether seasons reported in one source and not in the others didn't happen, or what? All others are reported in at least two different sources.

1936 (age 25)--NeL .367 (1 source only)
1937--NeL .361 or .371
1938--NeL .362 or .356
1939--NeL .343 or .336
1940--MxL .354
1941 (age 30)--NeL .347 or .337 MxL .256
1942--NeL .310 or .365
1941-42--42--PRL .409
1943--NeL .309 or .345
1944-45 Military service
1946-47--PRL .390 NeL .348
1947-48--PRL .432 NeL .336
1948-49--PRL .323 NeL .374
1949-50--PRL .353 or .354 NeL .371 (1 source only)
1950-51--PRL .325 MiL .352 (1 source only)
1951-52--PRL .295
1952-53--PRL .342 TxL .310
1953-54--PRL .265 (1 source only) TxL .314
1955--TxL .301
1956--TxL .299 MiL .294
1956-57--PRL .261 MiL .307

Mean Avgs. taking all posted avgs. as "real," and taking the average as nec. where a given year has "competing" reports
NeL--age 25-38 .348
MxL--age 29-30 .305
PRL--age 30.5-45.5 .339
MiLs--age 39-46 .307

1941 in Mx is certainly an anomoly (125 ABs).
   81. sunnyday2 Posted: August 13, 2005 at 05:33 PM (#1543853)
Here is Brown's NeL mean average comparisons. I only have Brown's mean average, not his actual average. Where the other players show 2 different numbers, the first one is the comp (the mean average) while the second (and/or the one in parens) is their career batting average.

Brown .348
Beckwith .366 (.352)
Bell .320 (.328)
Charleston (.340)
Leonard (.335)
Gibson (.351)
Moore .365 (.355)
Stearnes .355 (.332)
Suttles .329 (.341)
Torriente .343 (.336)
Wells .316 (.328)
Wilson (.354)

Note that there is an approx. even split (4 to 3) as to whether the actual average or the seasonal mean average is higher, so I can't even infer (much less know) whether Brown's actual career average is above or below .348. And of course even this number is influenced by the big spreads in his 1942 and 1943 NeL numbers from two competing sources--e.g. .310 versus .365 in 1942. I took his "real" average to be .3375. So if .365 and .345 are accurate for 1942 and 1943 his mean would increase to .354. And this is of course still not his real average.

But the evidence suggests he was approx. a .350 hitter which puts him smack dab in the middle of the above list of 12 other players, 10 of whom are HoMers.

As to Hr/550, I again don't have his actual career average so I will be comparing his mean season versus other players' career actuals. Can't say if this is "fair" or more to the point accurate or not, but it is what I have.

Brown 1 HR per 24 AB
Beckwith 1 per 27.5
Bell 1 per 69
Charleston 1 per 25
Leonard 1 per 37
Gibson 1 per 13
Moore 1 per 37
Stearnes 1 per 20
Suttles 1 per 14
Torriente 1 per 48
Wells 1 per 27.5
Wilson 1 per 46

Brown is 4th among the 13 players listed here.

If I rank all the players on these two dimensions just as a quick rule-of-thumb measure of their ability to hit 1) for average and 2) for power, I get the following. Here I've used the actual career average because I have it for everybody but Brown, while I don't have a seasonal mean handy for several others. I use Brown's seasonal mean as a surrogate.

1. Gibson 4 avg. 1 power = 5
2. Suttles 6 + 2 = 8
3. Brown 5 + 4 = 9
4. Beckwith 3 + 6T = 9.5
and Moore 1 + 8T = 9.5
6. Charleston 7 + 5 = 12
and Wilson 2 + 10 = 12
8. Stearnes 10 + 3 = 13
9. Leonard 9 + 8T = 17.5
10. Wells 12 + 6T = 18.5
11. Torriente 8 + 11 = 19
12. Bell 11 + 12 = 23

Just a quick way to try to decide whether Brown belongs in this group. As a hitter, the answer clearly is yes.
   82. Devin has a deep burning passion for fuzzy socks Posted: August 14, 2005 at 05:18 AM (#1544878)
Since several people have made the comparison, how would Oms do on these lists?
   83. sunnyday2 Posted: August 14, 2005 at 01:39 PM (#1545048)
Devin, my list (above) is based on actual NeL performance, so whatever caveats that implies. But anyway, Oms:

At .330, his BA exceed only those of Bell and Wells so he would be 11th out of 13.

And with 1 HR per every 40 AB he would rate aheadof Bell, Torriente and Wilson, so he would be 10th.

10+ 11 = 21. IOW he does not do very well. This of course is not the same as saying he is not a viable candidate since 10 of the 13 are already HoMers. But it does say to me that he is in a different (lower) tier than Willard Brown.

And of course this is only BA and AB/HR. Quick and dirty. And since these are rate stats, they don't do anything to assess career longevity, etc., though as a peak/prime voter, I don't care.

At some point, I should do just the candidates--Mackey, Trouppe, Taylor, add in Dandridge, Poles, Heavy Johnson, etc. etc. etc.... I think that Moore, Brown and Oms are perhaps the top 3 eligible today, but that assumption has not been re-tested in recent years.
   84. sunnyday2 Posted: August 14, 2005 at 01:48 PM (#1545050)
Oh, and on the other thread I also did Bill Wright, who has a respectable .341 BA. The numbers are not yet posted re. his career AB/HR, but his bet seasons were 24/550 and 29/550, whereas that is the mean for Brown. So I very very very very arbitrarily set his AB/HR at twice his peak, which works for most players, meaning somewhere 48-58.

IF that is in the ballpark, then Wright comes in at 6th out of 14 for BA and 13th out of 14 for AB/HR for a total of 19.

So strictly as hitters the tiers would appear to be:

A--Gibson

B--Suttles, Brown, Beckwith, Moore

C--Charleston, Wilson, Stearnes

D--Leonard, Wells, Torriente

F--Bell, Oms, Wright

Again, 10 of the 14 here are already HoMers and this is grading on the curve. In a larger pool, the As would probably extend down to Stearnes.

Leonard and Torriente may have been a tad overrated, however, taken as part of a larger pool including players from more recent decades.
   85. Mark Shirk (jsch) Posted: August 14, 2005 at 02:24 PM (#1545060)
Yes, Win Shares takes into account how a team did and some of that is 'dumb luck'. And yes, WARP is what a player woudl have done in a vacuum that he id not play in. To me this is evidence that if you are going to use the uberstats, you should try and find a balance of both.

Or we could just order guys on how many career hits they have...
   86. sunnyday2 Posted: August 14, 2005 at 04:22 PM (#1545130)
Personally I still think good old fashioned LWTS, whether masquerading as TPR or whatever, is very approximately as good as WS and WARP. So when WS and WARP disagree about a player, TPR is an acceptable tie-breaker. Of course OPS+ and some other things are just as good or better, or better yet, consider them all.
   87. sunnyday2 Posted: August 15, 2005 at 01:08 PM (#1546714)
Finally, by way of summary--

Taking the different projections of Brown's BB means that his projected OPS+ ranges from 125 to 127 to 129 to 131 to 132. From low to high, a significant difference as it maybe straddles the line where 1B-OF candidates have typically needed to be. Granting, however, that he is a CF and not purely a corner player, maybe he gets by on the lower end anyway...

But taking his WS, the different BB projections make little or no difference. The much bigger consideration is what you do with WWII.

Batting WS range 273 to 275 with no WWII credit. The WWII credit could range all the way from 0 to 54 bWS.

Fielding WS range 53 to 56, and the WWII credit again could range from 0 to 7.5.

So with no WWII credit, you're looking at a guy with about 326 to 331 career WS. With full WWII credit at the higher estimate, you've got 387.5 to 392.5.

The different BB impacts (meaning also the difference between a 125 and a 132 OPS+) just doesn't have a lot of significance if WS is your uber-stat of choice. At least that's how I understand things right now.

But a question for Chris--why the big difference, all the way from 34 to 55--in projected 1944-1945 bWS for Brown? I know one is Doc's and the other is yours, but do you know what were the different data and/or assumptions?
   88. sunnyday2 Posted: August 15, 2005 at 01:15 PM (#1546722)
Using my own personal WWII adjustment (and basically Chris' WS analysis in #72) here is Brown vs. some others whom I happen to have handy.

Career NeL BA...AB/HR...Career MLE WS...dWS...OPS+
Brown .348 24 356 57 131
Beckwith .352 27.5 318 52 137
Wells .328 27.5 386 115 114

Also for MLE peak WS, with the caveats that go with it

Brown 33-31-30/132
Beckwith 34-30-29/138
Wells 29-28-28/131

All of which is highly relevant at least to me as none of the 3 are in my PHoM yet, but all 3 in my PHoM consideration set (of 10 players) this year.
   89. Dr. Chaleeko Posted: August 31, 2005 at 07:22 PM (#1588016)
Gary A.,

Earlier in this thread you generously supplied leaguewide stats for the Texas League c. 1953-1956. I'm doing some preliminary leg work on Bus Clarkson who was in the TXL in 1952. Does your sourcebook have the 1952 leaguewide stats and Clarkson's lines as well? (also position would be wicked helpful.)

Thank you!
   90. Jorge Colon Delgado Posted: September 05, 2005 at 03:36 AM (#1597713)
Willard Brown is the most productive player in the PRWL. He's first in lifetime average (.350) and slugging (.603).

Other highlights of his carrer are:

Batting Leader(3); (1946-47) (1947-48) (1949-50)
MVP (1); (1949-50)
Home Runs (3); (1947-48) (1948-49) (1949-50)
Hits (2); (1946-47) (1949-50)
Runs (1); (1947-48)
RBI's (4); (1946-47) (1947-48) (1949-50) (1950-51)
Slugging (2); (1947-48) (1949-50)
Triple Crown (2); (1947-48) (1949-50)

In addition he connected 27 homers in 1947-48 and 97 rbi's in 1949-50 (both are still record).
   91. Gary A Posted: September 05, 2005 at 04:41 PM (#1598188)
Doc, I have a lot of material. I'll email it to you in an Excel file pretty soon.
   92. Dr. Chaleeko Posted: February 22, 2006 at 08:46 PM (#1871583)
OK, with Gadfly's walk rate information and a few other scattered bits of info on Brown, I've revised his MLEs.

-Here's what's different. The walk rate I used now includes the data from the PRWL Gadfly provided on the 1970 ballot thread. Combining it with AL, Texas League, and Mexican League data, we now have walk data from 10 seasons or partial seasons showing a 266/237 K/BB ratio. Anyway, there's more walks.

-Upon Gad's urging I re-looked at 1951, weighting the NAL performance more heavily than the DMSL and PRWL. As a result, I am now projecting two more years for Brown. His 1951 performance moves to solidly above average, and his final year is now 1952 when he dips way way down to a 76 OPS+

The effect of adding two year to his career is to dampen his career OPS+ by 3-5 points, while increasing his counting stast and career WS.

As always stats are presenting in a 154-game schedule and the same war year process as previously used. No regrssion. Comments, please!

Willard Brown Revised MLEs

YEAR LG AGE POS AVG  OBP  SLG    G   PA   AB    H   TB  BB ops
sfws
--------------------------------------------------------------------
1935 NL 20  of .257 .287 .586   44  174  167   43   98   7 128   7.1
1936 NL 21  of .330 .367 .397  132  531  502  166  199  29 105  18.5
1937 NL 22  of .334 .373 .533  154  622  585  195  312  37 143  32.0
1938 NL 23  of .266 .303 .405   67  268  255   68  103  13  93   7.0
1939 NL 24  of .332 .374 .448  126  509  477  159  214  32 120  21.1
1940 NL 25  of .299 .315 .400  138  538  525  157  210  13  96  15.5
1941 NL 26  of .341 .389 .545  112  460  427  146  233  33 161  25.3
1942 NL 27  of .329 .378 .475  154  634  588  194  279  46 149  28.3
1943 NL 28  of .323 .368 .524   75  305  285   92  149  20 156  14.9
1944 NL 29  of .309 .383 .457  128  525  486  162  240  39 136  25.0
1945 NL 30  of .309 .384 .498  126  517  478  159  257  39 144  27.5
1946 NL 31  of .343 .395 .497  154  635  585  201  291  50 152  31.7
1947 AL 32  of .330 .382 .589  150  610  563  185  331  48 155  35.7
1948 NL 33  of .308 .353 .528  139  564  527  162  278  37 136  26.3
1949 NL 34  of .317 .362 .487  141  573  535  170  261  37 125  25.0
1950 NL 35  of .287 .334 .398   93  380  355  102  141  25  91  11.1
1951 NL 36  of .308 .356 .451  138  560  526  162  237  37 116  21.7
1952 NL 37  of .269 .313 .321   96  387  364   98  117  23  76   7.3
====================================================================
               
.298 .362 .449 2168 8792 8230 2621 3951 565 122 351.9 
   93. karlmagnus Posted: February 22, 2006 at 10:48 PM (#1871780)
2621 hits at 122 still doesn't do it, sunshine. Beckley (for example) is 2930 unadjusted at 125, in a more important defensive position, without the doubts one has about "equivalence."
   94. Dr. Chaleeko Posted: February 22, 2006 at 11:07 PM (#1871811)
Quick note: The career SFWS should read 380.9.

All other column totals should be OK.

Sorry!
   95. AJMcCringleberry Posted: February 23, 2006 at 12:27 AM (#1871906)
Based on the hits and ABs you provided his BA should be .318 not .298.
   96. KJOK Posted: February 23, 2006 at 12:36 AM (#1871914)
I think Brown may have played SS 1935-37...
   97. Gary A Posted: February 23, 2006 at 03:51 AM (#1872048)
Also, his SLG should be higher: 3951/8230 = .480. So his career OPS+ ought to be higher, too -- about 127 if it's proportional to the increase in OPS (.811 to .842).
   98. Dr. Chaleeko Posted: February 23, 2006 at 03:37 PM (#1872355)
Thanks guys. I should just quit while I'm ahead!!!!! I'll repost with corrected totals.
   99. Dr. Chaleeko Posted: February 23, 2006 at 03:46 PM (#1872363)
This time with feeling! Seriously, sorry about the scroo-ups, just bear with me, I'll get my spreadsheet for Brown wholly corrected one of these times.

Anyway, my previous iteration understated his career offensive rates a lot because I'd somehow for his AVG and SLG I'd used PAs as the divisor. Back to first-grade Sabrmetrics for me....

His rates now move from Adnre Dawson land to someplace further northward. To answer KJOK's question, I've made him a CF all the way to keep it simple. Lord knows I apparently need to keep it simple.

Willard Brown

YEAR LG AGE PO  AVG  OBP  SLG    G   PA   AB    H   TB  BB ops
sfws
---------------------------------------------------------------------
1935 NL 20  of .257 .287 .586   44  174  167   43   98   7 128   7.1
1936 NL 21  of .330 .367 .397  132  531  502  166  199  29 105  18.5
1937 NL 22  of .334 .373 .533  154  622  585  195  312  37 143  32.0
1938 NL 23  of .266 .303 .405   67  268  255   68  103  13  93   7.0
1939 NL 24  of .332 .374 .448  126  509  477  159  214  32 120  21.1
1940 NL 25  of .299 .315 .400  138  538  525  157  210  13  96  15.5
1941 NL 26  of .341 .389 .545  112  460  427  146  233  33 161  25.3
1942 NL 27  of .329 .378 .475  155  634  588  194  279  46 149  28.3
1943 NL 28  of .323 .368 .524   75  305  285   92  149  20 156  14.9
1944 NL 29  of .334 .383 .493  128  525  486  162  240  39 145  25.0
1945 NL 30  of .334 .384 .539  126  517  478  159  257  39 155  27.5
1946 NL 31  of .343 .395 .497  154  635  585  201  291  50 152  31.7
1947 AL 32  of .330 .382 .589  150  610  563  185  331  48 155  35.7
1948 NL 33  of .308 .353 .528  139  564  527  162  278  37 136  26.3
1949 NL 34  of .317 .362 .487  141  573  535  170  261  37 125  25.0
1950 NL 35  of .287 .334 .398   93  380  355  102  141  25  91  11.1
1951 NL 36  of .308 .356 .451  138  560  525  162  237  37 116  21.7
1952 NL 37  of .269 .313 .321   96  387  364   98  117  23  76   7.3
====================================================================
               
.318 .362 .480 2168 8792 8230 2621 3951 565 130 380.9 
   100. Joey Numbaz (Scruff) Posted: February 24, 2006 at 06:14 AM (#1873409)
Something seems weird to me - in 1947, he goes 12-for-67 with just 4 extra-base hits, smack dab in the middle of his best season, according to the above translation? He didn't walk once in 67 PA in MLB, but managed an equivalent of 48 BB in 610 PA in his other play?

I'm sure there's a logical explanation, but off the top of my head something doesn't seem right. I know it's only 21 MLB games, but still, that's 13% of his entire season.
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Downtown Bookie
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.0396 seconds
49 querie(s) executed