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, April 19, 2004

1925 Ballot Discussion

1925 (May 2)?elect 2

 WS  W3 Rookie Name-Pos (Died)
354 77.5 1904 Sherry Magee-LF (1929)
231 43.0 1905 Hal Chase-1b (1947)
214 42.1 1908 Heinie Zimmerman-3b (1969)
171 50.7 1904 Terry Turner-SS/3b (1960)
198 41.9 1904 Red Ames-P (1936)
161 31.8 1911 Red Smith-3b (1966)
132 34.5 1908 Jack Barry-SS (1961)
130 32.2 1910 Chick Gandil-1b (1970)
132 30.6 1905 Lefty Leifield-P (1970)
125 26.6 1912 Larry Cheney-P (1969)
104 27.6 1912 Lee Magee-CF/2b (1966)
110 24.0 1911 Vic Saier-1b (1967)

DanG sent me the death list, but due to email difficulties, I can’t post it now. Will edit this later today or tomorrow with that info.

I don’t believe any new Negro Leaguers are eligible, let me know if that’s mistaken . . .

Joey Numbaz (Scruff) Posted: April 19, 2004 at 10:15 AM | 243 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 3 of 3 pages  < 1 2 3
   201. Brad G. Posted: April 28, 2004 at 04:22 PM (#524030)
Chris Cobb said: "The white stars you mention, yest -- Thompson, Duffy, Waddell, McCarthy -- all had much shorter careers than Johnson"

Correct me if I'm wrong, but isn't it reasonable to assume that many of these major leaguers played several years in undocumented minor leagues that may (or may not) have been comparable to the Negro Leagues of the time? And therefore, couldn't we argue that the full careers of these players may have, in fact, been closer to the length of Johnson's career?
   202. Daryn Posted: April 28, 2004 at 05:13 PM (#524032)
chris cobb,

You wrote this about Andrew Foster, placing him 24th:

Argument that Foster should receive MLE credit for 1897-1901 play isn't persuasive. Foster looks like major-league average pitcher at best in 1902, a dominant pitcher 1903-1908 (esp. 1903-1905), with an excellent half season in 1909, a good year in 1910, and effective spot-starting 1911-1914. That looks like a bit more quality that Rube Waddell, and a bit less than Mordecai Brown, so Foster goes here.

You are also a strong advocate of blackball players and do a great job of piecing together underdocumented careers. To me (using your characterizations, which I think understate the quality of 1900-1902 and 1911-1912) this looks like a career with 6 dominant years and 4 to 6 years of good pitching. He also was, if I am correct, considered to be the best black pitcher in the world during his peak. Isn't this close to Walsh; better than Koufax?

I know you don't like pitchers too much (I have Foster between Brown and Waddell too, I just have Brown at #2) but if this guy is the best black pitcher of the aughts, couldn't he find a place on your ballot?
   203. Jim Sp Posted: April 28, 2004 at 05:55 PM (#524033)
I don't know that I buy this line of reasoning, but I think Gavy Cravath is your test case.

If he played well in the PCL, and if the PCL was comparable to the Negro Leagues in quality, then it could be argued that his career is not short, and he deserves a fair bit of credit for his PCL work, which could bump him into the HoM...

A lot of ifs in there, but I can see the argument.
   204. John (You Can Call Me Grandma) Murphy Posted: April 28, 2004 at 06:04 PM (#524034)
I think minor league seasons should only be included if they can be reasonably shown to have been of major league quality.
   205. Chris Cobb Posted: April 28, 2004 at 06:43 PM (#524035)
Dan G wrote:

Correct me if I'm wrong, but isn't it reasonable to assume that many of these major leaguers played several years in undocumented minor leagues that may (or may not) have been comparable to the Negro Leagues of the time? And therefore, couldn't we argue that the full careers of these players may have, in fact, been closer to the length of Johnson's career?

There are two sides to this question, one that applies to how we interpret negro-league careers, and one that applies to white major-league players.

On the major-league side, I agree with John Murphy's answer above.

On the negro-league side, we need to consider carefully whether there are seasons early or late in the player's career during which he would not have been able to be a starter for a major-league team.

In Johnson's particular case, there's good evidence that he was a better than major-league average hitter in his age-21 season in 1895 and in his age 39 season in 1913, so I give him credit for the equivalent of a nineteen-year career at the major-league level. His play at below-major league level, for which I don't give him credit, includes his 1894 season in semipro ball and his play from 1914 and later on lower-tier black teams.
   206. Chris Cobb Posted: April 28, 2004 at 07:02 PM (#524036)
daryn wrote:

To me (using your characterizations, which I think understate the quality of 1900-1902 and 1911-1912) this looks like a career with 6 dominant years and 4 to 6 years of good pitching. He also was, if I am correct, considered to be the best black pitcher in the world during his peak. Isn't this close to Walsh; better than Koufax?

I know you don't like pitchers too much (I have Foster between Brown and Waddell too, I just have Brown at #2) but if this guy is the best black pitcher of the aughts, couldn't he find a place on your ballot?


daryn, I think we'll have to disagree about Foster's value 1900-2; the biography I've read is pretty clear that he was not a successful pitcher in his first season with a top black team and in a predominantly white minor league in 1902. I'm open to the idea that Foster was better than I have claimed in 1911-12. If he was a regular member of the pitching staff during those years and not a spot starter, then he would deserve to rank higher, if not move onto my ballot. I'll revisit the evidence for 1910-12 and see if I should revise my view of those seasons.
   207. Daryn Posted: April 28, 2004 at 07:19 PM (#524037)
I think the evidence will show that he was a player/manager those years who chose to spot start himself in the big games. Hard to assess the value of that, but I count them as analogous to seasons in relief (like Smoltz' first two years as a reliever, but not as good). Still, he has ten to twelve good to great seasons, highlighted by 6 dominant seasons. Try him at 14th -- I think it'll feel good.
   208. Daryn Posted: April 28, 2004 at 07:37 PM (#524038)
In searching for evidence of Rube's early career I did find a few nuggets on the net that suggested he had moved on from the Waco Yellow Jackets to the Leland Giants by 1900 or 1901. I've also seen that he beat Waddell head to head in 1902, not at a later date.

Most interestingly, in May 2004 Robert Charles Cottrell is releasing The Best Pitcher in Baseball: The Life of Rube Foster, Negro League Giant. It'll be interesting to see if that book has any more details.

Lastly, I found on the net an article Rube wrote that was reprinted in the Chicago papers in 1921 about black umpires. Quite fascinating. If you just Google Rube Foster you can find it on one of the first three pages.
   209. David Concepcion de la Desviacion Estandar (Dan R) Posted: April 29, 2004 at 12:15 AM (#524039)
Where do you find Win Shares for old time players?
   210. jimd Posted: April 29, 2004 at 12:31 AM (#524040)
Assuming that you don't have the book, I'd look under 'Positional Threads' from the top page for a lot of 19th century guys from the first prolonged discussion. Most of the Ballot Discussion threads will have info about new eligibles. I hope that helps.
   211. Marc Posted: April 29, 2004 at 12:31 AM (#524041)
I agree with Chris. When faced with the following dilemma, I would opt for the higher standard for both. ie.:

? A major league player racks up some nice numbers in the minor leagues before and/or after his major league career. I would count that only in extraordinary cases, and almost only before the major leagues had become clearly defined, i.e. pre-1880 or so. Beyond that it would have to be a really extraordinary case, though I would never say never.
   212. jimd Posted: April 29, 2004 at 12:33 AM (#524042)
Almost forgot. The 'New Eligibles by Year' thread contains a lot of the newly eligible players in a more concentrated form.
   213. Chris Cobb Posted: April 29, 2004 at 02:39 AM (#524043)
Most interestingly, in May 2004 Robert Charles Cottrell is releasing The Best Pitcher in Baseball: The Life of Rube Foster, Negro League Giant. It'll be interesting to see if that book has any more details.

The Cottrell biography came into print in 2001; it's been my primary source on Foster's career, in fact, and yes, it has more details.

I think the evidence will show that he was a player/manager those years who chose to spot start himself in the big games. Hard to assess the value of that, but I count them as analogous to seasons in relief (like Smoltz' first two years as a reliever, but not as good). Still, he has ten to twelve good to great seasons, highlighted by 6 dominant seasons.

The evidence, drawn from Cottrell's biography, Holway's _Complete Book of Baseball's Negro Leagues_, and Riley's _Biographical Encyclopedia_, is contradictory, and I'm not at all sure how to interpret it, so I'll just lay out what I have found, for now.

First, here are some relevant passages on Foster's play 1911-14 from Cottrell.

On Foster in 1911: "Foster, for his part, was now pitching less frequently than in the past, although his outings drew large crowds and considerable publicity. Left-hander Pat Dougherty and righties Frank Wickware and Bill Lindsay led the American Giants' pitching staff."

On Foster in 1912: "In 1912, the American Giants repeated as champions of the Chicago semipro circuit, winning 112 of 132 contests. . . . Scattered box scores and news accounts suggest that Foster remained, at least on occasion, a superb pitcher, with such incomplete records displaying a 6-0-1 record for the year."

Game summaries show him beating the Indianapolis ABCs (a black team), the Plutos (no indication of what sort of team this was), the West Baden Sprudels (a major black team under the management of C. I. Taylor, who would soon move to the ABCs), and the St. Louis Giants (a major black team). His tie came against Jose Mendez, pitching for the Cuban Stars.

On Foster in 1913: "his traceable mark for the regular season stood at 6-1." No indication whether he pitched on not in championship series against New York Lincoln Giants, which the Lincolns won. Smokey Joe Williams dominated the series, winning 7 of the 15 scheduled games.

On Foser in 1914: "Foster produced a 6-3 mark in games whose news accounts have been uncovered" (the team went 126-16). These nine games include a win against St. Joseph's (an independent Michigan team), a loss in relief to the Cuban Stars, a 1-0 shutout win over the Cuban Stars, a 2-0 shutout victory over the ABCs, victories over Benton Harbor and the French Lick Plutos, a victory over the Cuban Stars, a victory over "a group of top City League players" (I'm guessing this game, as an exhibition is not counted in Foster's record), a loss to the ABCs, and a loss to the Cuban Stars.

When the American Giants played the Brooklyn Royal Giants in a championship series in September, Foster did not pitch: the Chicago pitchers in the American Giants' four game sweep were Lee Wade, Frank Wickware, and Horace Jenkins, who pitched twice.

Passages from Riley

On Foster in 1910: "Incomplete statistics show Foster contributing a 13-2 record" (to a team record of 128-6).

On Foster in later years: "That year, 1911, Foster renamed the team the Chicago American Giants. . . . Foster himself was phasing out of his active career, with incomplete data showing records of 5-4 in 1914 adn 1-0 in 1917, two of his best teams of the decade. Enormously popular at the time, he still took the mound on special occasions to enhance attendance."

Statistics from Holway

1910: Foster 3-0, 2.17 RA (Frank Wickware leads team with 6-0 mark, posts 2.88 RA)
   214. Daryn Posted: April 29, 2004 at 12:39 PM (#524044)
Great work Chris. I hope your review of all the great info you have collected convinces you that there is enough of a great career there to put him on your ballot.

I wonder what the May 2004 release of the Cottrelle book is if the book is already three years old.
   215. Marc Posted: April 29, 2004 at 01:01 PM (#524045)
It doesn't look like anybody responded to Dan's question--where to find Win Shares for the "ancients." I take that to mean pre-NL. I mean, 1876 is ancient enough, but they're in the book.

Chris Cobb computed NA Win Shares, Dan.

But I have to ask Chris...where are they? Can you provide Dan with a link to where they are?
   216. DavidFoss Posted: May 03, 2004 at 06:17 AM (#616049)
Testing...

Back online here... seems to be a wee bit quicker tonight.
   217. andrew siegel Posted: May 03, 2004 at 10:38 AM (#616069)
I was Andrew Siegel on the old Primer, but thought I'd shorten it here. Hope we can get the kinks worked out quickly.
   218. Michael Bass Posted: May 03, 2004 at 02:14 PM (#616109)
Alright, my first ever ballot. I'm posting it here first to make sure it is in compliance with all of the rules. If it's OK, I'll then post it to the main thread.

I am mainly using WARP3, Win Shares, OPS+, ERA+. I am more a career voter than anything, but that's hardly an exclusive description.

1. Frank Grant - A lot has been written on him here, and I've read most of it (I think). The weight of the evidence seems to be that he was the best player of the 19th century Negro Leagues, and while measures the quality of the leagues at this time are speculative at best, my gut is that the best Negro Leaguer from any significant period of time would almost certainly have been a sure-fire HOMer if not banned.

2. Grant Johnson - The stats we have are interesting on him, but they are so amazingly incomplete that I hesitate to give them much weight. What we do know is that he was considered an excellent defensive player for a very long time at a key position, and he appears to have had a very strong bat. I have to believe that would have been at worst a better Bobby Wallace (see: next entry).

3. Bobby Wallace - Amazing career longetivity (at least as compared to most of the others eligible for this ballot). Had an above average bat, and carried it at shortstop. Was by all the measures I can find a really strong fielder.

4. Bob Caruthers - Wonder if you'll find any other ballots with these two back to back. Seemingly an odd placement for Caruthers on a career ballot, but I've discounted for the AA as much as I can, and I still think he packed more value into his few productive years than any other pitcher on the ballot did in more seasons.

5. Jimmy Sheckard - His bat was rarely outstanding, but consistently very good over a long career. I hesitate to give too much credit to even the best defensive corner outfielders, but every measure of his defense is awesome. James even has him as an "A" mixed in with center fielders.

6. George Van Haltren - Like Sheckard, rarely a great hitter, but very good so many seasons that his value adds up very nicely over time. Not as outstanding of a defensive player as Sheckhard, but at a more important position in center. I understand why peak/prime voters don't like him, but I'm a career guy, and I do.

7. Jimmy Ryan - Hey, another OF with a long career! Little more inconsistant than the above two. Slightly more top 10 OPS+ seasons, but also more mediocre seasons. Defense is mixed; seems to be rated well, but that he was moved off center at age 31 worries me. Further investigation seems to indicate he was moved not because of his deficiencies, but because the new CF was amazing, so I'm inclined to give him strong credit for his D.

8. Mordecai Brown - Though a relatively late start, a very good career, with a number of outstanding seasons. Would be higher, but like everyone else I feel ERA+ overstates his contributions a bit because of the defense behind him. It boggles the mind a little that we have not yet elected anyone from the great mid-1900s Cubs teams, but they simply lacked surefire HOMers. I think, though, that Sheckard and Brown will finally give them some representation over the next few elections.

9. Fielder Jones - Not a great hitter. Comparable to Wallace with the bat, but a little better. OPS+ does understate him a bit, as he is OBP heavy. What gets him on my ballot this high is his fielding. All the measures I look at have him as one of baseball's greatest centerfielders. I'm less confident on him than I am on those above him, so I'd listen to arguments why I should be moving him down on future ballots.

10. Sam Thompson - Defense seems a mixed bag of reputation vs. stats. I'm going with stats here. Will listen to reasons why he was a better defender than I'm rating him. Seems as though he had a nice arm, something that nearly always ends up in an OF being overrated defensively. Bat's not in question, though, he was a better pure hitter than any of the OFs above him. Career shorter than theirs, though, and that combined with his defense drops him to 10th.

11. Sherry Magee - Offense like Thompson, though a little less. Defense is in question, but probably okay-ish. Being cautious on him for now, hope to read more in future years to get a better grip on him.

12. Lip Pike - No one on this ballot am I less comfortable with. For the recorded stats we have, he hit the living crap out of the ball as long as he played. Pre-1871 ball was so disorganized that I discount it in a major way, but seems as though he was good there, so he gets a minor plus. I don't have the first clue what to make of his defense, though it was seemingly bad. Voted Player Most Likely to Fluctuate On My Future Ballots

13. Clark Griffith - Very good pitcher for a quality innings number. Helps in my mind (though I'm guessing in very few others) that he was useful as a reliever/spot starter after his prime days were over. For the record, this is a pretty big dropoff point on the ballot, any of the next 10 guys could be placed here, as Griffith is ahead by very little.

14. Cupid Childs - Pretty strong hitter, and OPS+ again understates him because he was OBP heavy. Didn't have an exceedingly long career, but it wasn't short, either. Very good defensive player at second, though I believe (correct me if I'm wrong) that second was considered an offensive position at the time along the lines of third base today. I could see moving him up in future years.

15. Tommy Leach - YAO (yet another outfielder). Worse hitter than Jones even. Lotta career. Great defender by most measures. Can't rate him higher, even as a big career guy, though, with that bat in the outfield. At least Jones had his OBP working for him.

Next 5 (and comments for Top 10 players)

16. Jake Beckley - You'd think I'd like him as a career guy, but I don't much. His hitting is un-special for a first baseman, and I'm not giving extra credit for the dearth of HOM 1Bs in this era, unless someone can point to a reason that 1B was considered an important position or a hard position to play at the time. This just seems to be a period where most of the great players were atheltic enough to play elsewhere.

17. Bill Monroe
18. Vic Willis
19. Rube Foster - Initially did not have him nearly this high. Some late reading has convinced me that there may be an overcorrection here for his non-player work, to the degree that we're underrating him as a pitcher. Will read up more, may very well make my next ballot.

20. Rube Waddell

Dickey Pearce - Stat record unimpressive to me. Yes, he was past his prime, but this wasn't exactly the height of competition in the early 1870s, so I feel that a HOMer would have been better. As I said in the Pike comment above, pre-1871 ball was so loosely organized and the level of play so low, that anecdotoal evidence from that era doesn't so much for me. If he'd been good in the NA, I might take the anecdotal evidence more seriously, but he wasn't, so I'm not.

Joe McGinnity - I simply don't see it. His ERA+'s are good, but hardly great. WARP3 hates him, Win Shares is ambivalent. Someone show me what I'm missing here. I'm open to moving him into my top 20 at least on future ballots, but with what I've got now, I don't see it.
   219. John (You Can Call Me Grandma) Murphy Posted: May 03, 2004 at 11:26 PM (#616162)
Dickey Pearce - Stat record unimpressive to me. Yes, he was past his prime, but this wasn't exactly the height of competition in the early 1870s, so I feel that a HOMer would have been better.

What shortstop had a better record after the age 35 during the 19th century than Pearce? His statistical record, instead of showing that he's not a HoMer, is evidence of a player of high quality. Combined with the contemporary reports of his greatness at the position he "created" and that he was also a star catcher, he belongs somewhere on everybody's ballot.

BTW, the ballot looks good, Michael. You can post it anytime you want on the ballot thread.
   220. Jeff M Posted: May 03, 2004 at 11:52 PM (#616176)
Joe McGinnity - I simply don't see it. His ERA+'s are good, but hardly great. WARP3 hates him, Win Shares is ambivalent. Someone show me what I'm missing here.

Those statements might be a bit strong. For WS, his 3 best years, 5 best consecutive years, 7 best years and "per 43 starts" numbers would rate him about a "B" if you are comparing him to WS from pitchers actually in the HOF. His career WS put him more in low "C" territory vis-a-vis HOF pitchers.

As for WARP, I continue to use WARP1 since the WARP2 adjustments on the way to WARP3 are a complete mystery. For WARP1, his 3 best years and 5 best consecutive years rate him about a "B", his 7 best years and "per 43 starts" numbers would rate him about a "C" and his career WARP1 would rate him about a "D", all as compared to HOF pitchers.

In my comparison to HOF pitchers, I eliminate the ridiculously good (like Cy Young) and the terrible selections (like Jesse Haines).

Those comparisons would seem to make him worthy of a Top 15 finish in this consideration set, but that's just my opinion. Not much different than Clark Griffith really.
   221. Zapatero Posted: May 04, 2004 at 02:20 AM (#616212)
So I just posted my first ballot ever, and I'm sure some folks will be annoyed that I not only pulled out Addie Joss, but put him first. You probably thought you were done with him. Well, I do feel strongly about this. Subtracting his first and last seasons (so we're looking at 7 years: 1903-09) you have a pitcher who:

Never had an ERA above 2.50
Never had an ERA+ below 130
Averaged 2.88 K/BB
Pitched a perfect game during one of the greatest pennant races of all time

Here's how Joss stacks up against some contemporaries in IP per Win Share (so obviously, a smaller number is better, like at-bats per homer):

WSIPIP/WS
Mordecai Brown2963172.310.72
Ed Walsh2652964.311.19
Ch. Mathewson4264780.711.22
Cy Young6347354.711.60
Addie Joss1912327.012.18
Rube Waddell2402961.312.34
Eddie Plank3614495.712.45
Joe McGinnity2693441.312.79

The point I'm trying to make is that Joss certainly belongs on this list. If Plank (or McGinnity, for that matter) makes it for throwing so many innings, then Joss should make it for throwing such good innings.

By the way, this stat makes Brown look golden, doesn't it?
   222. Zapatero Posted: May 04, 2004 at 02:38 AM (#616219)
wow. that table couldn't have come out worse if I'd tried. Here' goes another attempt:

.............................WS...........IP..............IP/WS
Mordecai.Brown....296..........3172.3..........10.72
Ed.Walsh..............265..........2964.3..........11.19
Ch..Mathewson....426..........4780.7..........11.22
Cy.Young..............634..........7354.7..........11.60
Addie.Joss............191..........2327.0..........12.18
Rube.Waddell.......240..........2961.3..........12.34
Eddie.Plank...........361..........4495.7..........12.45
Joe.McGinnity........269..........3441.3..........12.79
   223. Chris Cobb Posted: May 05, 2004 at 02:08 AM (#616627)
_Second_ try at posting with the new server; first try this afternoon never made it.

Marc at page 3 post 71 above (formerly 271, I guess) asked about the location of the NA translated win shares. With the site change, I'm afraid I have no idea. They were posted in the thread one reached by clicking on the "positional threads" link on the old HoM home page.

Zapatero, I'm afraid the table you've posted above doesn't do much to advance Addie Joss's case, at least with respect to providing any justification for your own ranking for him. If Brown indeed earned more WS per IP than Joss, and threw 845.3 more innings at that superior rate, how can Joss possibly rank ahead of Brown?

Also, this table shows Joss as being about 5% better than McGinnity on a rate basis, but McGinnity as throwing 48% more innings than Joss? How can a 5% advantage in pure quality outweigh a 48% difference in quantity?

This table does show that on a per inning basis, Joss was in the same class as the pitchers we've been electing or are close to electing (if one believes WS). But it also shows pretty clearly that Joss has to be ranked last out of this group of 8. There's just no way he's the best pitcher eligible right now, let alone the best player overall. That doesn't necessarily mean he isn't a HoMer, but if you want to make the case that he is a HoMer, you'll need to find stronger evidence.
   224. sunnyday2 Posted: May 05, 2004 at 03:03 AM (#616689)
Is anybody else getting dead dialogue boxes, that won't accept a cursor? So I clicked to another page and came back (slowly) and then it took my cursor, and then one of those annoying and totally pointless screen refreshes (I'm not tracking a ballgame here) wiped out my message. Arrgghh.

I don't remember what I wanted to say.
   225. Al Peterson Posted: May 05, 2004 at 02:14 PM (#617004)
Zapatero's table shows you really how long and good Cy Young's career was. You can fit 3 of Joss' stats line in there. But really, not enough playing time for Addie to be pushed to the upper parts of the ballot.

Apparently Addie was also really well liked as a player. They had a famous All-Star game after his death with proceeds going to his widow.
   226. John (You Can Call Me Grandma) Murphy Posted: May 05, 2004 at 03:56 PM (#617129)
Zapatero's table shows you really how long and good Cy Young's career was. You can fit 3 of Joss' stats line in there. But really, not enough playing time for Addie to be pushed to the upper parts of the ballot.

That's my problem with him, too. The quality is certainly there, but not the quantity.
   227. Jeff M Posted: May 05, 2004 at 11:43 PM (#617681)
I think WS/IP*9 is more illustrative than Zapatero's IP/WS, because it emphasizes how much value was lost by Joss not pitching as many games/innings. For every 9 IP these guys had, they earned at least .704 Win Shares, except that most of them kept earning these amounts, and Joss (sadly) did not. Call it an "opportunity cost" of not having Joss and having to replace him with someone else.

Brown.......... .840
Walsh.......... .805
Mathewson...... .802
Young.......... .776
Joss........... .739
Waddell........ .729
Plank.......... .723
McGinnity...... .704

Also, one huge point not reflected in Zapatero's table is that everyone on the list, except Joss, has their numbers adversely affected by a decline phase...particularly those with long careers. I'd like to see the same table with everyone else's 9 best consecutive years.
   228. sunnyday2 Posted: May 06, 2004 at 02:21 AM (#617792)
I don't use rate stats at all. Figures lie and rate stats most of all.
   229. Chris Cobb Posted: May 06, 2004 at 03:02 PM (#618052)
Well, I'm going to try to get back to work on some players. jimd's ballot provides an opening on Rube Foster that I 'd like to follow up on.

He wrote:

7) Rube Foster -- Legendary peak for a short time in the oughts. Very good pitcher for some time afterwards. Those who vote peak should re-examine him.

1902; 51-3
1903; 59-1 (I also saw 54 wins and 55 wins) rumored to have won 44 in a row
1904; 51-4
1905; 50-4

From yest. Taking 56 W for 1903, that's 208-12 for four years. My questions about him are: what was the team record, i.e. what were the records of the other pitchers of his team? What was the quality of the competition?

I calculated the Pythagorean ratio expected to produce such a record and did some figuring of MLE's, modifying the runs by percentages. 20% (high minors) reduces it to 193-27; 33% (worse than UA) reduces it to 170-50 .773; 50% (average dead-ball MLB player expected to post Bondsian 1.300 OPS) 114-106. Essentially, if the competition was so bad that his pitching W/L should be considered only MLB average, then his teammates also need Ruthian batting stats to be considered MLB average.


First question: the percentages you are using, jimd, are changes in _runs_? So you use the Pythagorean formula to calculate a runs scored to runs allowed total that fits the WP, then apply the MLE modifier to the runs totals, then recalculate winning percentage? Is that a correct description of what you are doing?

Second question: if I'm right in understanding what you are doing in question one, if one were to shift from runs to other individual stats, like, say, batting average and ERA when calculating MLEs, how does the percentage by which the stat should be modified vary in relation to the percentage by which runs are modified?

Since hits are (more or less) raised to the second power in the runs created formula, does that mean that batting average, when calculating MLEs changes by a percentage that is (more or less) the square root of the change in runs? Or would it vary by the same percentage?

Since ERA and RA are derived directly from runs, would they vary by the same percentage as runs do?

I've done some playing around with translations for the negro-leaguers statistics, but I don't understand translations very well, so I haven't been very confident in my conclusions. Answers to these questions would help greatly.

My guess right now is that statistics generated vs. "all competition" for early negro-league players need to be reduced by 30-40% at the runs level. Statistics generated vs. other negro teams only (the sort Holway has compiled) need to be reduced 10-20% at the runs level, if my understanding of how runs MLEs imply batting average MLEs is correct.
   230. jimd Posted: May 06, 2004 at 06:37 PM (#618251)
Something is eating the beginning of the first quote in this post.

<scription of what you are doing? </i>

Yes.

Second question: if I'm right in understanding what you are doing in question one, if one were to shift from runs to other individual stats, like, say, batting average and ERA when calculating MLEs, how does the percentage by which the stat should be modified vary in relation to the percentage by which runs are modified?

That depends.

One school (Palmer) would say OPS varies linearly with runs. Particularly when evaluating one player in a context that is reasonably close to league average.

Another school (James) would say that RC varies multiplicatively (sometimes). But is close to linear when evaluating most players in ML environments. But maybe the multiplicative applies (my speculation) when evaluating a team that outclasses its opponents at the .900 level like Boston 1875 (or Rube's teams). The combinative nature of run-creation may be on display in both very high and very low run-scoring environments.

I used the linear version in my comment about Bondsian OPS. On reflection, it may not be accurate. I need to do more with this.

The translations of pitching stats are the easy ones, because runs are the base metric. The translations of hitting stats are the thornier ones because of the overarching debates about whether conversion is linear or a power function (with even the exponent in question, see EQA).
   231. Chris Cobb Posted: May 07, 2004 at 03:40 AM (#618607)
Thanks, jimd!

Here's some more data to chew on regarding Foster's peak. The Philadelphia Giants' 1905 season is exceptionally well-documented. Here's the info gathered in Cottrell's biography of Foster. For the regular season, their documented record was

132-21-3

In their 153 wins and losses they scored 1,145 runs, and allowed 467. (This pythags to just about their record, so the two data pieces are consistent.)

Three pitchers only, Danny McClellan, Scotty Bowman, and Foster himself, pitched all the games, and every game but one was a complete game. This certain supports Foster's claim that his complete record for that year was 51-4; he would have been a bit better than the team.

His record, documented from boxscores, was 25-3-2, with a 1.66 ERA.

It seems likely that Foster's games against better teams were more likely to receive box-score coverage, so I can see him going 26-1 against mostly semipro competition.

The 1904-06 Philadelphia Giants squad was undoubtedly a great team, the best black team yet assembled, and the best that would be assembled until the 1910 Leland Giants and perhaps the 1911-13 New York Lincoln Giants.

It featured the following four players, all in the prime of their careers: Rube Foster, Grant Johnson, Bill Monroe, Pete Hill.

In 1905, the New York Giants went 105-48, led by Mathewson, McGinnity, Bresnahan, Donlin, and Dahlen. Dahlen was late in his career, and McGinnity wasn't great that year, but the other 3 were great.

If all four of the Philadelphia Giants stars are HoMers, one could make the argument that the team would have been as good as the Giants, or as the Philadelphia Athletics, who won a more competitive American League with a record of 92-56. They were led by Plank and Waddell on the mound, both of whom had great years, and Harry Davis, Lave Cross, and Socks Seybold, all very fine players, in the field.

There's no way the overall competition the Phil Giants faced was as good as the high minors, so let's start by seeing what happens to their expected record if we reduce their scoring and raise their runs allowed by 25%.

The team's pythagorean record under those conditions would be 105-48: that matches the Giants.

If we reduce by 33% (worse than UA, also an imaginable level of competition), we get 92-61: that's in the Athletics range, also the Pirates and the Cubs in the NL.

If we reduce by 40% (equivalent to what?) we get 80-73, a slightly better than average major-league ballclub.

50%? 61-92, better than St. Louis, Boston, and Brooklyn in the NL, better only than St. Louis in the AL. This would be a club of marginal major-league caliber.

So how do we decide which sort of club we are looking at here? It would be helpful if we could learn more about the clubs they played.

Given the reputations of the Philly Giants Big Four, I'm inclined to see a 30ish% discount as about right.

Thoughts?
   232. Yardape Posted: May 07, 2004 at 11:57 PM (#619511)
From what I know, Chris, I'm inclined to agree that a 30 percent or so seems about right. That's really just a guess, though.

Does anybody here know where I can get a copy of Holway's Complete Book of the Negro Leagues? Amazon doesn't have any in stock, nor do any other online stores I've checked. My local library doesn't have a copy either. I'd like to get my hands on one, if anyone has any tips.
   233. Jeff M Posted: May 08, 2004 at 02:23 AM (#619717)
I've done some playing around with translations for the negro-leaguers statistics, but I don't understand translations very well, so I haven't been very confident in my conclusions. Answers to these questions would help greatly.

I don't have the answers, of course, but I "back in" to all of my numbers when applying a quality discount. Basically, I apply the quality discount to the resulting metric (Runs Created for example) and reverse engineer the components Doubles, Triples, HR, etc., which proportionately reduces the other stats like Runs and RBI. I use the "Goal Seek" function in Excel to run these calcs. I make some other adjustments, but basically this gets me a set of translated stats...which is really helpful for counting stats (like career RBI, career BA, etc.)

A linear formula like RC or Base Runs (or whatever) is the easiest to do this with. It's nearly impossible with WS because backing into the numbers would take forever. It is actually impossible with WARP, because we don't know how it is calculated. So for WS and WARP, I just apply the discount directly.

If I remember correctly, Bill James has an example of this in the BJHBA in his comment on Willie Davis. That example related to adjusting for run environment (which I do too), but it coordinates well with my discussion above.

For pitchers, I still use LWTS as a measure, and apply the quality discount to the LWTS score, backing into the ERA number and then making some pythagorean adjustments to the wins/losses (with some other adjustments as well). I think the new Baseball Encyclopedia has a new formula for pitching runs (and it's no longer called linear weights), but I haven't dug into it yet.
   234. Arrieta, Gentile Arrieta Posted: May 08, 2004 at 02:57 AM (#619810)
Yardape, have you tried the publisher of the Holway book? It appears to be available, but you'd have to call them, they don't have online ordering. Here's a link:
http://www.hastingshousebooks.com/books/baseball.html
   235. John (You Can Call Me Grandma) Murphy Posted: May 08, 2004 at 04:38 PM (#620459)
Here's something that I picked up on:

Top pitchers for 1886 (AA)

Bob Caruthers:      12.9
Dave Foutz:         11.0
Toad Ramsey:        10.9
Ed Morris:           8.9
Matt Kilroy:         8.5


Ave= 10.44
Sum= 52.2


Top pitchers for 1908 (AL)

Ed Walsh:       16.1
Addie Joss:      9.6
Cy Young:        8.9
Walter Johnson:  6.0
Bob Rhoads:      6.0


Ave=9.32 
Sum=46.6


Top pitchers for 2001 (NL)

Randy Johnson:      9.6
Curt Schilling:     9.0
Javier Vazquez:     9.0
Greg Maddux:        8.7
Darryl Kile:        7.2


Ave=8.7
Sum=43.5

------------------------------------------------
Now for the hitters:

Top hitters for 1886 (AA)

Dave Orr:       8.3
Bid McPhee:     7.5
Tip O'Neill:    6.0
Henry Larkin:   5.6
Hick Carpenter: 5.2


Ave=6.52
Sum=32.6


Top hitters for 1908 (AL)

Nap Lajoie:      13.0
Fielder Jones:    9.7
Ty Cobb:          9.6
Heinie Wagner:    8.4
Bobby Wallace:    8.0


Ave=9.74
Sum=48.7


Top hitters for 2001 (NL)

Barry Bonds:        16.1
Sammy Sosa:         13.9
Luis Gonzalez:      12.5
Todd Helton:        10.6
Rich Aurillia:      10.6


Ave=12.74
Sum=63.7


According to WARP3, there are more good hitters now (no argument from me there), while there was more value from pitchers early on than now and has actually regressed through the years. I know this a small sample, but does this make any sense?

We know of Win Shares' problem with pitching, but WARP3 has one, too (albeit smaller than WS). This is why I take with a grain of salt Caruthers' worthiness as a potential HoMer. There is no way that I'm going to accept that pitchers of today do not have at least equal worth with the 19th century pitchers when the responsibilities of the former are much more greater on a game-to-game basis than the of the latter.
   236. Michael Bass Posted: May 08, 2004 at 05:16 PM (#620470)
Innings pitched for the 1886 players (in the order that you listed them): 387.3/504.0/588.7/555.3/583.0

For the 2001 players: 249.7/256.7/223.7/233.0/227.3


The merits of this method aside, we're supposed to be surprised that a system where value is cumulative (rather than expressed on a rate basis) is going to favor the former, even with some heavy timelining?
   237. John (You Can Call Me Grandma) Murphy Posted: May 08, 2004 at 09:24 PM (#620784)
<former, even with some heavy timelining? </i>

I'm not surprised by it since I've mentioned it countless times before you joined our club (though I never illustrated it exactly this way). My point is that the old-timers get a reward for playing a game that was easier on a season basis. WARP3 is supposed to show all pitchers on a level playing field. In that regard, it fails. Again, there is no way that pitchers were better than they are today.
   238. Michael Bass Posted: May 08, 2004 at 10:24 PM (#620812)
Better? Completely agreed, but that's clearly true for all players due to nutritional/workout/scientific/etc. reasons.

More valuable? Hard to say. I know that pitching was easier back then, but it's not like the huge IP were good for them. Most of these guys aside from the elite flamed out pretty fast.

I guess it's just not as much of an issue for me as a career voter as it is (and will moreso become) for the pure peak/prime voters. Leaving aside Caruthers, who was a bit of a freak for both his low IP totals for the era and his hitting, the other 4 on that 1886 list top out at 45 WARP3 (most well below that). Of the 5 on the 2001 list, 3 are already well past that mark, 1 will pass it in the next few years barring injury, and the 5th was at a little over 42 when he prematurely died.

What I'm saying...most of these guys with the (in retrospect) ridiculous IP totals almost by definition were more valuable than anyone today for a short period of time. They paid for it by having for the most part ridiculously short careers. How pure peak/prime voters will take that into account as we move through the years and are eventually comparing, say, Bob Caruthers to Pedro Martinez will be very interesting. Because I'm not sure it's even possible to do so using the same mathematical system for both in any way that doesn't favor the old guys.
   239. Chris Cobb Posted: May 09, 2004 at 02:50 AM (#621105)
Because I'm not sure it's even possible to do so using the same mathematical system for both in any way that doesn't favor the old guys.

One way that I'm trying to do incorporate pitchers from all eras in the same mathematical rating system is by incorporating a peak rate stat based on a fixed number of innings pitched. I'm using 365, along with a career measure and a total peak measure.

Top early pitchers and later pitchers tend to have fairly comparable career values. Earlier pitchers tend to rate much more highly than later pitchers on peak measures that calculate value on a seasonal basis. This advantages early pitchers. Top later pitchers, however, will tend to rate much more highly on win share rate stats than earlier pitchers. This advantage balances out the peak advantage, making pitchers from different eras scalable against one another. We're only at 1926, so I don't know what will happen with 1990s pitchers, but so far the two groups seem fairly balanced against one another on this basis in my system.

Since unadjusted win shares overrates early pitchers substantially, one would have to use heavily adjusted win shares or WARP for the rate stat to balance the peak stat properly, of course, but there are lots of ways to work it out.
   240. John (You Can Call Me Grandma) Murphy Posted: May 09, 2004 at 05:44 AM (#621538)
I guess it's just not as much of an issue for me as a career voter as it is (and will moreso become) for the pure peak/prime voters.

Agreed.

Because I'm not sure it's even possible to do so using the same mathematical system for both in any way that doesn't favor the old guys.

Like Chris (but in a different way), I compare the pitcher or player to his era. Since I don't believe in timelining for the HoM, my list of pitchers up to 1926 is a good mix from all of eras by peak and career.
   241. Yardape Posted: May 09, 2004 at 06:07 PM (#621666)
Thanks, Don, I hadn't thought of that. I've looked at every retailer I could find, but I hadn't thought to try the publisher directly. I'll give them a call this week.
   242. jimd Posted: May 11, 2004 at 01:43 AM (#623175)
There is no way that I'm going to accept that pitchers of today do not have at least equal worth with the 19th century pitchers when the responsibilities of the former are much more greater on a game-to-game basis than the of the latter.

But John, if you look more closely at the numbers you are quoting, the pitchers of 2001 are worth much more than those of 1886 on an inning-by-inning basis. The five pitchers you cite for 2001 averaged .39 PRAR/IP (Pitching Runs Above Replacemant per Inning Pitched). The five pitchers you cite for 1886 averaged .18 PRAR/IP. Today's pitchers are about 210% better.

I did a similar calculation for the Hitters. I threw Bonds out because he is such an extreme outlier (more than double the average of the others on the 2001 list). The numbers are .21 BRAR/Out for the 2001 guys. Similarly I threw Carpenter out because his inclusion is an error (-2.4 WARP-3 in 1886); their numbers were .11 BRAR/Out. The modern guys check in at about 190% better (excluding Bonds because he'd skew such a small sample).

So both the hitters and the pitchers of today's NL have WARP-3 rates approximately double that of the 1886 AA. I suppose that will please some of us and dismay others. But it is approximately balanced, which might be as expected. (I have to think about that some more.)

WARP3 is supposed to show all pitchers on a level playing field.

No. It's supposed to show all players (hitters, fielders, pitchers) on a level playing field based on how many Runs they improved their team over a borderline player of their own time, then applying a quality adjustment based on the year-to-year evolution of their leagues.

***

So why do the pitchers of the 1880's score so high relative to the position players of then? Or the pitchers of today? Because they pitched so often and so much (as Michael Bass pointed out). They pitched the workload of 2.5 starters today, which allows them to overcome that factor of two quality ratio. The hitters of 1886 did not hit/field the workload of 2.5 starting players today. Plain and simple. The relationship between hitter's seasons and pitcher's seasons was radically different then so that pitchers tended to be the most valuable players in any given season; OTOH, they burned out too quickly (due to abuse or rules changes) to accumulate the most valuable careers.
   243. John (You Can Call Me Grandma) Murphy Posted: May 11, 2004 at 03:16 PM (#623668)
So why do the pitchers of the 1880's score so high relative to the position players of then? Or the pitchers of today? Because they pitched so often and so much (as Michael Bass pointed out). They pitched the workload of 2.5 starters today, which allows them to overcome that factor of two quality ratio. The hitters of 1886 did not hit/field the workload of 2.5 starting players today. Plain and simple.

But it was easier to do so during the 1880s and distorts their peaks when compared to other eras. A Randy Johnson or Curt Schilling would have easily been able to duplicate the innings pitched of that time, but they happen to play the game when the responsibilities of pitchers are much greater than of the 19th century hurlers.

If the average pitcher today was given a legal supplement that would allow them to pitch 300 innings a season, their WARP and WS numbers would dwarf the preceding seasons. However, the average pitcher with the supplements would be no better or worse than the non-supplement pitchers. The rating systems add up the innings pitched without examining how tough it was or not to achieve those numbers.
Page 3 of 3 pages  < 1 2 3

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
rr
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.8134 seconds
68 querie(s) executed