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, March 22, 2004

1923 Ballot Discussion

Quite a new class coming on the ballot and only one electee this year. Wonder who it’ll be . . .

 WS   W3  Rookie Name-Pos (Died) 
655 180.1  1897  Honus Wagner-SS (1955)
446 115.2  1899  Sam Crawford-RF (1968)
361 100.1  1901  Ed Plank-P (1926)
268  70.8  1903  Johnny Evers-2b (1947)
231  58.6  1903  Chief Bender-P (1954)
157  28.7  1906  Hans Lobert-3b (1968)
124  35.0  1909  Jim Scott-P (1957)
129  32.9  1909  Chief Meyers-C (1971)
139  25.1  1907  Mike Mowrey-3b (1947)
137  23.4  1907  Bobby Byrne-3b (1964)
128  26.8  1903  Cy Falkenberg-P (1961)
113  26.0  1905  George Gibson-C (1967)
Negro Lg   1901  Rube Foster-P (1930)

Thanks to DanG for compiling this list:

HoMers and candidates who died in the past year:

HoMers:
Age Elected
 69 1903 Cap Anson

Candidates:
Age Eligible
 81 1882 Wes Fisler
 70 1896 Orator Shaffer
 67 1896 Joe Gerhardt
 62 1902 Sam Thompson
 59 1902 Tommy McCarthy
 53 1907 Nig Cuppy
 52 1905 Billy Rhines
 43 1918 Jake Stahl

Future Candidate:
Age Eligible
 27 1928 Austin McHenry
Joey Numbaz (Scruff) Posted: March 22, 2004 at 04:11 PM | 221 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 2 of 3 pages  < 1 2 3 > 
   101. Chris Cobb Posted: March 25, 2004 at 06:21 PM (#523231)
Again, I?d like to see the Win Shares analysis, but f the true universe were WARPed, this doesn?t speak well for Brown?s candidacy. Evers had about as much value as Mordecai in this brief period, and much more in his career, and we aren?t talking much about Evers as a HoMer.

Tom, I'll try to post WS data tonight to match what you've given for WARP.

I'd like to see some of Brown's strong supporters -- that large group ranking him in the top 5 -- put forward a clear case for why Brown's _value_ is really as high as they claim it is. The WARP numbers fit with my rankings of the Cubs stars pretty well, though my rankings are not based on it.
   102. John (You Can Call Me Grandma) Murphy Posted: March 25, 2004 at 06:33 PM (#523232)
Given that sabemetrics seems unable to explain the Cubs' success, is it possible that the sabermetric discipline's dismissal of the importance of baserunning and SB is all a ghastly mistake, at least in the deadball era when so much revolved around it?

Sabermetrics doesn't dismiss the importance of stolen bases. This cannot be stressed enough. James created a formula specifically for the Deadball Era.

What sabermetrics does account for that other orthodox statheads don't account for is caught stealing.

Yes, it drives me crazy when I hear this about sabermetrics and stolen bases. :-)
   103. OCF Posted: March 25, 2004 at 06:51 PM (#523233)
I don't know what, if anything, this has to do with baseball, but it seems worth mentioning. In #104, Chris Cobb mentions "players of Indian heritage from Oklahoma." What may not be so well known is that there were black Indians.

In the 1820's and 1830's, the Indian tribes that had occupied the southeastern US, in particular, the Cherokee, Chickasaw, Choctaw, Creek, and Seminole (collectively known as the "Five Civilized Tribes") were forcibly relocated westward into Indian Territory. But those tribes, and in particular a landed aristocracy within those tribes, picked up some cultural traits from their southern white neighbors. One thing they picked up was the owning of slaves. African slaves.

The nations of Indian Territory faced a three-way choice with the Civil War: side with the Union, side with the Confederacy, or try to stay out of it. All three positions had their advocates, and Indians from Indian Territory fought on both sides. The last Confederate general to surrender was a Cherokee named Stand Watie.

Several things happened with the settlements after the Civil War. One is that the "Five Civilized Tribes" lost land. Some of that lost land remained unassigned, which led evetually to the opening of parts of Indian Territory to white settlement starting in 1889. Another part of the settlements: the slaves were not only freed, they were given full tribal membership. (Well, they were supposed to. I'm not sure how that worked out.)

The paragraph I was quoting from Chis's post is mostly about the years 1900-1910. At the beginning of that decade, what would become Oklahoma was split between Indian Territory in the east (still the territory of the "Five Civilized Tribes") and Oklahoma Territory in the west, dominated by white settlers but including western Indian tribes. The two territories were merged into one state in 1907. Since there never were that many slaves of the Indians, I would assume that by 1900 they and their descendents were outnumbered even in Indian Territory by African-Americans who had migrated from the South. But they did exist.

Indians were allowed in major league baseball, as demonstrated by Albert Bender's listing among our candidates. African-Indians (either descendents of Indian slaves or products of intermarriage) would not have been, no matter what their tribal membership might have been.
   104. Marc Posted: March 25, 2004 at 06:56 PM (#523235)
1. Re. Tom's Cubs data: First, Tom, when you say that Evers has almost as much "prime" (team prime) value and more career value than Brown, you forgot to mention that Tinker has even more of both! But re. Brown, I think the simple fact that he did have more value during that team prime period and that comparatively speaking he was more more-valuable than other pitchers than the others were vs. their peers at SS, 2B, etc., I think explains Brown's vs. Evers and Tinker's ranking. It's about position differentials.

2. The Tigers thought the Cubs base running and stealing was an important part of their success. Before I accept that at face value, I'd want to know how many they stole, how many CS, etc. Not that it's your assignment to tell me, and not that I have time to look it up right now.... But even then, the question would be: whose candidacy would it affect? Maybe Sheckard? So many questions, so little time.
   105. Max Parkinson Posted: March 25, 2004 at 07:20 PM (#523236)
My plea to all who are currently planning to vote for Roger Bresnahan:

The main argument in his favour thus far seems to be, ?he was the best Catcher in baseball between (Ewing or Bennett) and Cochrane (or ?30s glut)." To this I respond, 1. That?s not necessarily true, and 2. Even if it was, so what?

For this exercise, please allow me to use EqR as a substitute for actual runs, and EqA as a substitute for BA. Regardless of your predispositions towards Davenport?s WARP system, I hope that we can all accept that EqR is simply a Linear Weights system that holds up to generally accepted SABR principles. As such a LW system, I think that it is superior to a simple counting statistic, and in all likelihood benefits Bresnahan, as it definitely favours players that get on base, Roger?s strength on offense. The data that I am using is WARP 1, adjusted at the 2/3 power for season length, based on Paul Wendt?s list of games played by year. This is so to assign credit as equitably as possible between eras, while maintaining priority of besting your actual competition, rather than players that came 20 years ahead or behind.

Let?s compare Roger to every other hitter that we?ve had a chance to see in this process. Suspend your disbelief and put yourself in 1923 ? so you?ve witnessed professional baseball since at least 1871. Of every professional ballplayer to date, we?ve deemed 31 hitters worthy of election to the HoM.

Bresnahan has 895 career EqRs, and a career EqA of 0.302. 19 of the 31 honorees have better career EqAs, and of the remaining 12, all have more EqRs than Bresnahan except Charlie Bennett, who has 871, and George Wright, who has 854. Note that this only includes Wright?s career from 1871 forward. These 12 average 45% more runs than Bresnahan.

EQR EqA
   106. Paul Wendt Posted: March 25, 2004 at 07:21 PM (#523237)
John Murphy #96
   107. John (You Can Call Me Grandma) Murphy Posted: March 25, 2004 at 07:45 PM (#523238)
Let?s compare Roger to every other hitter that we?ve had a chance to see in this process.

Max, how does Roger compare to other catchers as a hitter?

The main argument in his favour thus far seems to be, ?he was the best Catcher in baseball between (Ewing or Bennett) and Cochrane (or ?30s glut)."

I actually thing Schang was better, though he straddles the Deadball and Lively Ball Eras.
   108. John (You Can Call Me Grandma) Murphy Posted: March 25, 2004 at 07:50 PM (#523239)
<i>John Murphy #96
   109. Max Parkinson Posted: March 25, 2004 at 08:22 PM (#523240)
John,

On the same list of 30 catchers, he's 5th in career EqR and 4th in EqA. Unfortunately I don't have is numbers strictly as a catcher handy. IIRC, someone made a great post in the Catchers thread that compared many of these guys as catchers only. I think that it used OPS+...
   110. Daryn Posted: March 25, 2004 at 09:07 PM (#523241)
Gentleman,

I find Max's post convincing. The only reason I vote for Bresnahan is because I do think he is the best catcher between Bennett and Cochrane. I don't have a quota, but that impresses me. But if his defense was so awful and he wasn't the best catcher, I think I have to drop him out of my top 15 (I've already dropped him from 8 or so to 13). Is Max missing something obvious? The one thing I do have to say is that I don't find Win Shares' catcher defense points to be that compelling, particularly in Bresnahan's era. That being said, I don't know if Wshares underrates or overrates Bresnahan.
   111. Daryn Posted: March 25, 2004 at 09:10 PM (#523242)
That should be addressed to "Gentlemen". And by the way, are there any females here? Are there any non-whites? Or is this just an exercise conducted by affluent white males? Just curious. Please don't feel you have to "out" yourself.
   112. John (You Can Call Me Grandma) Murphy Posted: March 25, 2004 at 09:19 PM (#523243)
re: Bresnahan

I don't think he's an overwhelming choice for the HoM. On my next ballot, he will only be #11. Mediocre defense and lack of playing time behind the plate does hurt him in my eyes.

With that said, the man was an exceptional hitter considering his era and the amount of abuse catchers took behind the plate. While I don't think his offense was Piazza-like during that time (obviously), I feel it's more impressive than his OPS+ indicates. Would Bench, Campanella, Cochrane, etc., been able to sustain their OPS+ during the Duke's era (or earlier)? I'm doubtful.
   113. Max Parkinson Posted: March 25, 2004 at 10:30 PM (#523244)
John,

There's no doubt that Bresnahan is one of the best couple of catchers (hitting only) between Ewing and Cochrane. What I'm saying is that he was bad enough defensively to overcome that.

Let me compare him to Bill Joyce, a 3rd baseman and sometime shortstop for Brooklyn, Washington and the Giants during the 1890s. 3rd base, like catcher is a primarily defensive position. If you can do the job adequately and hit a little, so much the better.

Now, according to the same methods as my much-too-long post, Joyce was essentially tied with John McGraw for the best hitting infielder (2nd, 3rd or SS) between Ross Barnes and Rogers Hornsby - basically 50 years. He had a career EqA of .327, and had 827 career EqR - 92% of Bresnahan's total in about 2/3 the playing time. Where are the calls for his election?

Now, why did he have only 2/3 the career Bresnahan did? Someone who has better access to history books can correct me, but until I hear otherwise, I'll believe that it's because he was a complete butcher at 3rd, and so he wasn't wanted, even when he played on pretty bad teams. If you play a premium defensive position, you better pick it at least a little or no one will care what you hit.

Joyce (and to some extent McGraw) is simply an example of a player who was the best hitter at his position between X and Y over a large time period, and isn't going to get a sniff here. That seems to be Bresnahan's stump speech, and if it won't work for Joyce (or McGraw), I'm not sure why Bresnahan is getting 2nd and 3rd place votes.
   114. Rick A. Posted: March 25, 2004 at 10:44 PM (#523245)
No one has mentioned this yet, but I was very interested in OCF's post #125, about the "Five Civilized Tribes" and the Oklahoma territory. Good stuff. Sometimes it's nice to find some off-topic discussions here. (Although, I don't want to descend to Clutch Hits depths of off-topic conversations. We do have a job to do here, and need to stick to the subject, more or less. But one or two off-topic conversations wouldn't hurt.) Is that a topic of interest to you? Where did you get your information? You seem to be pretty knowledgable about that era and it was very interesting to me.
   115. John (You Can Call Me Grandma) Murphy Posted: March 25, 2004 at 10:52 PM (#523246)
Now, according to the same methods as my much-too-long post, Joyce was essentially tied with John McGraw for the best hitting infielder (2nd, 3rd or SS) between Ross Barnes and Rogers Hornsby - basically 50 years. He had a career EqA of .327, and had 827 career EqR - 92% of Bresnahan's total in about 2/3 the playing time. Where are the calls for his election?

It's a little different since we can point to McGraw, Cross and Collins (the latter is not a perfect contemporary admittedly) than Joyce, while the worst you can say about Bresnahan, for his seasons as a catcher, is he was the equal of Johnny Kling (and Bresnahan also had some impressive years as a centerfielder).

I'm not sure why Bresnahan is getting 2nd and 3rd place votes.

No disagreement with me there, Max.
   116. Chris Cobb Posted: March 26, 2004 at 02:07 AM (#523247)
Here's the WS follow-up to TomH's excellent post at #117 above on the distribution of value on the great Cubs teams. For ease of comparison, I've followed Tom's formatting.

Player? 5yrs career
   117. User unknown in local recipient table (Craig B) Posted: March 26, 2004 at 02:59 AM (#523248)
Sorry I missed a ballot. I'll have a prelim up shortly. Like a lot of others, I have Wagner at 1 and Crawford probably at 2, though I may need to slip him down to 4. Rube Foster will probably be on my ballot as my first blackball player ranked on peak value.
   118. Marc Posted: March 26, 2004 at 03:51 AM (#523249)
I have one word for Chris' post.

WOW!

I am strongly inclined to believe that WS undervalues solid middle infielders based on this analysis. But...is this chronic, or just deadball?

What to do!?
   119. OCF Posted: March 26, 2004 at 04:37 AM (#523250)
There's a problem we've always had, and always will have - a problem that we've argued from many different directions and mostly not resolved. For the purposes of this project, how do we compare a pitcher to a position player? The career trajectories of pitchers are different and more varied than those of position players. The evaluation-clouding issues of different environment and different usage loom larger with pitchers than with position players.

WARP and Win Shares both attempt to compare pitchers to position players. As we can see in posts #117 and 139, the results may be sharply different.

If Chris Cobb had the only vote, we would elect McGinnity almost as soon as our "shoo-in" candidates clear out of the way, and we would never elect Brown at all. One thing is clear - we can't ignore Chris and his arguments. We have to deal with what he's saying.

I haven't yet run the sort of numbers on both Brown and McGinnity that I'd like to - I've been too busy for that, so far. But the one part of Chris's argument that I'm having a hard time buying is the ranking of Brown far below McGinnity. Maybe they're near-equals. Maybe McGinnity is also being far overrated by the electorate. If that's the case, maybe we should elect Johnson and Grant and Wallace and Sheckard before we take any more pitchers, even Plank.

Maybe I could pose this question for Chris: why do your arguments for ranking Brown low not also apply to McGinnity, at least to some degree?
   120. Chris Cobb Posted: March 26, 2004 at 04:48 AM (#523251)
I am strongly inclined to believe that WS undervalues solid middle infielders based on this analysis. But...is this chronic, or just deadball?

What to do!?


I'm not sure that WS is undervaluing "solid" middle infielders; I am sure that, before the lively ball (including pre-1900) it is undervaluing _great_ middle infielders, esp. those who played on teams that were excellent in fielding across the board.

What to do first, I'd suggest, is examine the problem more fully.

As we discussed a couple of elections ago, the WS formula normalizes pitching and fielding values to a ratio appropriate to the modern game. However, that sort of distortion should lead to a fairly consistent, fairly small undervaluing of defense in the deadball era.

jimd, in a brilliant reading of the WS fine print, observed at that time that, on the top of p. 33 of win shares, is the obscure little rule "which says that the Defensive Win Shares of a team cannot be less than .16375 per game played, nor more than .32375 per game played."

It strikes me that the significant undervaluing of great defensive players is more likely to take place on the teams where this little rule has kicked in. I've started to survey teams to find the instances where this "maxing out" of defensive value occurs, but I haven't gotten very far yet. Once a great defensive team over a period of several years was identified, a comparison on the same model as the one Tom and I have run could be performed to check for significant undervaluing.

Voters who use WS could then decide how far to adjust the value of players they believe are "shorted" by WS.

Or they could just trust WARP1. That would be an easier solution . . .

It's also worth keeping in mind that as the middle infielders are being undervalued, the pitchers are being overvalued, though not to the same extent. With the very strong ballot we have right now, Three Finger Brown won't reach the top at his current valuation until 1925 or 1926, so his election isn't imminent, but I'd recommend a more cautious assessment of him while the group continues to study the fielding issue.
   121. Chris Cobb Posted: March 26, 2004 at 05:27 AM (#523252)
I haven't yet run the sort of numbers on both Brown and McGinnity that I'd like to - I've been too busy for that, so far. But the one part of Chris's argument that I'm having a hard time buying is the ranking of Brown far below McGinnity. Maybe they're near-equals. Maybe McGinnity is also being far overrated by the electorate. If that's the case, maybe we should elect Johnson and Grant and Wallace and Sheckard before we take any more pitchers, even Plank.

Maybe I could pose this question for Chris: why do your arguments for ranking Brown low not also apply to McGinnity, at least to some degree?


There are several reasons why McGinnity rates so much higher in my evaluation than Brown does, all of which should be subjected to scrutiny.

The first reason, which relates directly to the current discussion of fielding, is that, while in a few seasons McGinnity played before an excellent corps of fielders, his defensive support was _never_ as good as what the Cubs gave Brown, season in, season out. They have a very good argument as the best defensive team of all time; the Giants of the aughts don't.

The second reason, also directly relevant to the fielding issue, is that McGinnity appears to have been very good at hit suppression; he consistently gave up many fewer hits than his team's defensive averages suggest he should have allowed (I use WARP's numbers for this, which match a few trials I calculated for myself pretty closely). Brown was usually pretty close to his team average. My system gives McGinnity significant credit for this. WARP, if I understand it correctly, uses a standard 70-30 division of responsibility for hit suppression on balls in play between fielders and pitchers (the WS division is 50-50), so it probably gives McGinnity less credit for hit suppression than I do and gives Brown more credit. It could be argued that the WARP system is fairer than mine; it could be argued the other way. This is a point worth debating. If I accepted the WARP division, I expect McGinnity would drop a bit in my rankings and Brown would rise. This is probably one place where my conclusions differ from WARP's a good deal, because Brown is getting a 30% share of all that great defensive value in the WARP system.

The third reason, which is not related to the fielding discussion, has to do with the way I calculate the value of durability within a single season. For all the innings that a pitcher throws up to the league average, I compare his value to the value of an average pitcher. For all the innings that a pitcher throws above league average, I compare his value to that of a replacement pitcher, whose winning percentage equals that of an average pitcher divived by 1.2. My reading of the durability factor is that the extra innings a great pitcher throws keep the team from having to rely on bad pitchers. I did a study in which I compared the winning percentage and DERA of starting pitchers in this era to other pitchers used sparingly, and 1.2 was the typical quotient I found when the starters wining percentages were divided by the spare pitchers' percentages. It's an estimate, as replacement level always is. McGinnity threw tons of innings, and this enhances his value; Brown tended to be close to league average. One might argue that the overall strength of the Cubs staff and the wisdom of not overworking pitchers means that Brown should not be penalized for the way he was used. I see being able to throw a lot of innings as a significant talent in this, or any era for pitchers, so my system assesses value this way. Take this element out, and McGinnity drops and Brown rises. I don't know how WARP figures value above replacement level, but this is probably another point where my system differs significantly from WARP.

Finally, my system sees pitchers as having more value tied up in their peaks than is true of position players. An average pitcher throwing an average number of innings on an average team I see as earning about 15 WS in this era, while average position players in average playing time were earning more like 19 WS. So pitchers accumulate less value by being average than position players do, which, when coupled with their typically shorter careers, means their career value lags significantly behind that of position players. The flip side of this is that, at their best, they tend to have very high value, and their value above average (which is my chief peak measure) tends to be quite a bit higher than that of position players. McGinnity has a fabulous peak, as my system sees it; Brown doesn't. He has a lot of good years, but only one hugely valuable season. If you look at peak and career value for pitchers somewhat differently from this, McGinnity could drop and Brown could rise.

These are the reasons my system prefers McGinnity to Brown by such a substantial margin. It also prefers Walsh to Brown. I'm still working on Plank, but from the half-career study I have, I think he will rank quite highly, certainly ahead of McGinnity. He doesn't have any lights-out, blow-em-away unhittable seasons, but he has a lot of excellent seasons and was above average in many more. I can see ranking Brown and McGinnity fairly equally in mid-to-low ballot positions. I can see ranking Brown higher than McGinnity, I suppose. But a ranking of Brown in the number 3 or 4 spot in this election seems to me not to take sufficient account of the fielding support he received for most of his career.

Comments?
   122. Dag Nabbit is part of the zombie horde Posted: March 26, 2004 at 06:40 AM (#523253)
(note: though this is also about those Cubs teams, it's not really a response to any of the other comments on those teams. Oh vey!)

Well, for the last few elections I've done some bellyaching on those old Cubs teams. My main starting point has been how could a team win as many games as they did in a 1,2,3. . . 10 year period as they did unless they had more than one great player. Sure they were solid, but there have been other solid teams that haven't done that. So I just assumed that they weren't the solidest team that's ever been & from that starting point been trying to prove it ever since without really examing the question were they really THAT solid or not.

So I finally checked.

Took the WS book & charted out how many WS the Cubs got at each hitting position (I'll let the pitching debate hang fire for a second, besides there's already plenty of looks of Brown going on & no one else from the staff's on anyone's ballot) from 1903-12. Then, for comparison, I looked at the Yanks from 1934-43 (the best 10 year win span I can find) & compare to see what happens.

First off, Cubs got less overall value from their starting 8 than the Yanks did & more from their pitching with an equal value from the bench. 53.3% of the Cubs value came from their starting 8, while the Yanks got 57.6%.

But the real question here is depth. So let's chart it out. Of the 80 seasons on both teams, how many times did a Cub get from X-Y win shares & how did the Yanks do in comparison. The chart:

......Cubs......Satan
   123. Paul Wendt Posted: March 26, 2004 at 07:09 AM (#523254)
John Murphy #138 in reply to Max P's analogy Bresnahan:Joyce
   124. John (You Can Call Me Grandma) Murphy Posted: March 26, 2004 at 07:25 AM (#523255)
Paul:

<i>Johnny Kling, regular catcher 1902-1910 (eight seasons; 3/4 of his career measured by Batting Outs):
   125. Philip Posted: March 26, 2004 at 01:24 PM (#523262)
Prelim ballot

Top 10 all-time:
   126. Yardape Posted: March 26, 2004 at 06:41 PM (#523264)
The main argument in his favour thus far seems to be, ?he was the best Catcher in baseball between (Ewing or Bennett) and Cochrane (or ?30s glut)."

I think another problem with this argument is that it doesn't take into consideration Negro Leaguers. I would guess Louis Santop and probably Bruce Petway were as good or better than Bresnahan, which should cut down that time frame. Even a player like Chappie Johnson, pretty much a direct contemporary of Bresnahan's, probably would have been close in value. (I see Johnson as Kling-like). Bill James wrote that catcher was probably the strongest position in the Negro Leagues. To me, that brings down Bresnahan's positional argument enough that he's off my ballot (though he might make it during the drought).

But a ranking of Brown in the number 3 or 4 spot in this election seems to me not to take sufficient account of the fielding support he received for most of his career.

Comments?


Wow. I had Brown at #3 last time, just ahead of McGinnity, but I don't think I'll do that again. Thanks, Chris, that was a really persuasive argument. Now, I still don't think the difference between McGinnity and Brown is as great as you say it is, but I think 3-Finger will drop on my ballot. (Tinker and Evers will pick up most of the resulting credit, although it may not be enough to move them onto my ballot).

Not trying to find another way to knock Caruthers here -

Sure, Joe, we all know your new "job" is just surfing the Internet, looking for anti-Caruthers arguments. :)
   127. Chris Cobb Posted: March 26, 2004 at 06:50 PM (#523265)
An addendum on Brown vs. McGinnity in my system (put this one in the D'oh! category):

I had forgotten to give Brown credit for saves. Given the tightness of the rankings, with his 49 saves accounted for, he moves up a half a dozen spots. Still not on my ballot, but now he's much closer.

Very Preliminary ballot

1. Wagner
   128. jimd Posted: March 26, 2004 at 07:13 PM (#523266)
Wow, too much good stuff to digest in one lunch period.

Once a great defensive team over a period of several years was identified,

Look at the Beaneaters of the 1890's. If anything, they were capped more often than the Cubs were (defensively by Win Shares, that is).

This results in a radical discrepancy in the assessment of Kid Nichols between the two integrated-evaluation systems. Win Shares has Nichols as a top-25 career (23rd). WARP-3 has him just missing the top 100 (I estimated 108, though I easily could have missed somebody), due to all that great defense behind him. He's still an upper-half HOMer, though his inner-circle candidacy could be debated.
   129. Chris Cobb Posted: March 26, 2004 at 08:28 PM (#523267)
Look at the Beaneaters of the 1890's. If anything, they were capped more often than the Cubs were (defensively by Win Shares, that is).

Indeed. The FOHL might see if the case for Tinker provides a model for the case for Long.
   130. John (You Can Call Me Grandma) Murphy Posted: March 26, 2004 at 08:46 PM (#523268)
Isn't the big reason for the prolonged success of the Cubs (or Beaneaters) the handling of their pitching staffs?
   131. Yardape Posted: March 26, 2004 at 08:49 PM (#523269)
Chris or jimd (or anyone else who wants to answer),
   132. Al Peterson Posted: March 26, 2004 at 09:02 PM (#523270)
Continuing with regards to Bresnahan on some of the things Max P and Paul Wendt have written.

<I>WARP1 rates Kling +16 fielding runs above average in 3104 Batouts as a catcher;
   133. karlmagnus Posted: March 26, 2004 at 09:13 PM (#523271)
What about McGuire, who looks to me better than Bresnahan, Bennett, Kling or anyone till the 30s. Why isn't he? (wasn't on Max Parkinson's list, I notice.)
   134. Max Parkinson Posted: March 26, 2004 at 09:26 PM (#523273)
Karl,

You're right - he has totally slipped my list. 400-odd players and I've missed him and Pop Snyder. I have to say I'm a little embarrased.

I would rank McGuire ahead of Bresnahan at a cursory look. His rate was 96 to Roger's 95 as a catcher - bad, but a little better than Bresnahan. His much longer career gives him more runs than Bresnahan, although not quite at Bresnahan's .302 rate (McGuire's at a .270)

Hope that helps...
   135. Dag Nabbit is part of the zombie horde Posted: March 26, 2004 at 09:41 PM (#523275)
Re: Those Boston teams of the 1890s.

Given that James set a cap of .32375 Def WS per game, it appears that those old teams were capped out on Def WS in: 1890-3, & 1897-1901. That's 9 years in a 12 year period. It ended when Jimmy Collins left. Long played in 1178 of his 1874 career games in those years - 62.9% of his overall total. In comparison, jimd mentioned that the Cubs were maxed out in 1904-10, during which years Tinker played a little over half of his games.

1897-1901 were when Long & Collins were side-by-side in the IF, which is interesting given that many said that Collins may have been overrated as a player.

This is also interesting news for Hugh Duffy supporters, not to mention the legion of Tommy McCarthy fans here.
   136. Chris Cobb Posted: March 26, 2004 at 09:47 PM (#523276)
I would rank McGuire ahead of Bresnahan at a cursory look. His rate was 96 to Roger's 95 as a catcher - bad, but a little better than Bresnahan. His much longer career gives him more runs than Bresnahan, although not quite at Bresnahan's .302 rate (McGuire's at a .270)

If one is using WARP numbers, Bresnahan has to rank ahead of McGuire. In many fewer games, he accumulated _slightly more_ wins above replacement than did McGuire, 78.4 to 78.0.

Karl, I know your system has some adjustment in it for changing playing conditions, but are you adjusting enough when comparing McGuire to Bresnahan? McGuire played a long time, but 40% of his at bats came 1893-99, the highest-offense years we've seen, while Bresnahan's big seasons were 1903-08, in the heart of the deadball era.

I no longer see Bresnahan as a leading candidate, but he's well ahead of McGuire in my view.
   137. John (You Can Call Me Grandma) Murphy Posted: March 26, 2004 at 10:01 PM (#523277)
I have no problem with taking defense into account when analyzing a player, even if they are a catcher. Questions about Sam Thompson and Pete Browning have played a part in suppressed support for those players since we started. If it hurts a right fielder, how can it not hurt a catcher?

Who is arguing that defense shouldn't be part of the discussion for catchers?
   138. karlmagnus Posted: March 26, 2004 at 10:12 PM (#523278)
I wouldn't dignify my selection methods as a "system" in the way some people have one, but McGuire to me although not as good a hitter as Bresnahan or Bennett, gets lots of extra points for career length; when you adjust his hitting to 130 game seasons he gets over 2800 hits, which given decent defensive value makes him quite Meritorious, in a bottom-of-ballot kind of way.

I actually don't think we should elect any of these guys, and Bennett will probably not make my personal HOM (though he won't miss by much, unlike Walsh or Flick.) But if we collectively get into a "let's have another catcher" mood, I'd rather McGuire than Bresnahan.

Schalk I haven't yet looked at.

I have Duffy quite high (well, #10 this year, but higher when we're not in a candidate glut). The Beaneaters error on fielding WS suggests he should be looked at more seriously -- he was supposed to be a WONDERFUL CF, so may well deserve a high allocation of extra fielding WS, even though he wasn't an infielder. I may put him back above Thompson -- we should certainly elect at least 1 or maybe 2 more of these 90s OF, in my view, before we're finished. Even allowing for era, Duffy and Thompson seem to me significantly ahead of Sheckard and indeed Magee (who I've just been looking at -- better than Sheckard, but not much.)
   139. karlmagnus Posted: March 26, 2004 at 10:15 PM (#523279)
Incidentally, on Sheckard, the Cobb book is interesting in that the Tigers don't seem to have ascribed much credit to Sheckard for the Cubs' '07-08 strength (mind you, they had Cobb and Crawford, so their standards for OF may have been high.) Pitching, fielding, baserunning and the T/E/C trio were what drove the Cubs, in their view.
   140. Carl Goetz Posted: March 26, 2004 at 10:25 PM (#523280)
Joe,
   141. Chris Cobb Posted: March 26, 2004 at 10:27 PM (#523282)
I have Duffy quite high (well, #10 this year, but higher when we're not in a candidate glut). The Beaneaters error on fielding WS suggests he should be looked at more seriously -- he was supposed to be a WONDERFUL CF, so may well deserve a high allocation of extra fielding WS, even though he wasn't an infielder. I may put him back above Thompson -- we should certainly elect at least 1 or maybe 2 more of these 90s OF, in my view, before we're finished. Even allowing for era, Duffy and Thompson seem to me significantly ahead of Sheckard and indeed Magee (who I've just been looking at -- better than Sheckard, but not much.)

Duffy is a peculiar case. Given the WS cap on fielding WS affected the Beaneaters regularly, one would expect that WARP would like Duffy better than WS, but the opposite is true. I have no idea why. I think team breakdowns for the Beaneaters along the lines of the ones TomH and I did for the Cubs might be illuminating.
   142. Al Peterson Posted: March 26, 2004 at 10:39 PM (#523283)
Who is arguing that defense shouldn't be part of the discussion for catchers?

Not anybody, I just feel it's being emphasized to a smaller degree than it should. Probably just my perception is skewed but if a catcher is poor at half the game of baseball that should be recognized.

Vote for Bresnahan how you like; I'm just saying buyer beware...
   143. Carl Goetz Posted: March 26, 2004 at 11:11 PM (#523286)
I see what your saying, Joe, but I don't see how setting a Peak level is any more arbitrary than setting a Peak Time. As I understand it, most people measure peak by something like Best 3 years, Best 5 years, Best 3 Consecutive years, and so on. I would rather know how long a player played at a certain level and how far above it, which is partly why I use 2 different levels. A player who gets 25 WS every year will score very well against the Average Baseline. I set replacement level at 45 wins, so Star level could also be called the Anti-Replacement player. You're probably right about the .200 Replacement level built into WS. I'll have to come up with a way to adjust my 3 levels to account for this.
   144. Carl Goetz Posted: March 26, 2004 at 11:15 PM (#523287)
Joe,
   145. Jeff M Posted: March 26, 2004 at 11:28 PM (#523288)
I just feel it's being emphasized to a smaller degree than it should. Probably just my perception is skewed but if a catcher is poor at half the game of baseball that should be recognized.

But defense does not represent 1/2 of a player's value. It might be 1/2 of the game of baseball, but of that 1/2, a huge chunk is pitching, not fielding, and certainly not fielding behind the plate.

Any player whose runs saved on defense approximates his runs produced at the plate -- in other words 50% hitting/50% fielding -- is a defensive "specialist."
   146. Marc Posted: March 26, 2004 at 11:35 PM (#523289)
The Tigers saw the Cubs play 10 games in 07 and 08. The Cubs W & L 306 regular season games those 2 years plus a few ties.

During those 2 regular seasons the Cubs avg. about 1.6 XBH and 1.4 SB. In the 10 Series games (including 1 tie game) the Cubs avg. 1.3 XBH and 3 SB.

So whatever the math is for the Series wins in terms of SB contributions to runs scoring does not at a glance appear to be terribly relevant to their overall success. A few 2B and 3B go a long way in my experience.

Also over those 2 seasons (300+) Scheckard hit almost exactly the team BA (17 points over and then 18 points under. Those 2 seasons not being a big part of Sheckard's HoM case.) In the Series he hit 19 and then 55 pts under. He was not a career .238 hitter no matter what Ty Cobb thought.
   147. Chris Cobb Posted: March 26, 2004 at 11:35 PM (#523290)
I don't agree with disregarding the team defensive WS cap. James put it there for a reason - he knows the system isn't perfect, and he didn't want to overrate anyone because of that.

Well, James doesn't actually tell us what his reasons are, but whatever they are, it can't be a concern not to overrate _anyone_. Since WS are assigned from a fixed pool, if one player doesn't get them, another one will. It's just a question of who, and how much. Given that the system is already overrating deadball pitching, the cap on fielding WS seems to me to exacerbate the system's imperfections rather than mitigating their effects.

Even if they bumped up against the cap - how far could they be over it? Then divy that up amongst 8 players and how much of an adjustment can you make?

Well, with the 1904-1910 Cubs, pretty far; this is a defense that is saving up to 100 runs above _average_ on the basis of hit prevention on balls in play at its best. In a low run environment, that's a lot of wins. As far as the divvying up goes, it's not clear that the division would be even: WARP certainly suggests that the WS system is denying value disproportionately to the top defensive players on the team. Now, I'm not saying that WARP is correct -- nobody's made a case one way or the other yet (and I know that objections [whose particulars I don't recall] have been raised to WARP's handling of Fielding Runs Above Replacement that would be relevant to the current discussion) but the possibility needs to be examined. And given that this is not a rare but a consistent application of the cap right through the defensive peaks of both Tinker and Evers, it's possible that we're looking at a systematic underrating of the two players by 3-5 WS a year for 7 years. If that turns out to be the case, that would not be at all an insignificant underrating. It wouldn't make either of them an obvious HoMer, but it would be the difference between their being viable candidates and their being immediately consigned to the Hall of Very Good.
   148. karlmagnus Posted: March 26, 2004 at 11:55 PM (#523291)
There's no question that Cobb, while one of my heroes (among others are Williams and Durocher -- I like ballplayers salty!) had opinions on other ballplayers that weren't always sound -- he didn't think much of that young Ruth feller, for one thing. But for those like me without a universally solid "system," contemporary opinion, now that we've got to an era where it's easily available, is a useful reality check -- we shouldn't be too quick to elect people whom contemporaries didn't rate highly (of my own favorites, I've seen less than I'd like about Beckley, for example.) Equally, I'd be disappointed if we didn't elect 3-finger Brown, whom contemporaries did rate very highly.

Sheckard' mediocre 07-08 does explain why Cobb didn't rate him, I agree.
   149. jimd Posted: March 27, 2004 at 12:14 AM (#523292)
Scattered comments.

<i>Chris or jimd (or anyone else who wants to answer),
   150. Dag Nabbit is part of the zombie horde Posted: March 27, 2004 at 12:31 AM (#523293)
Duffy is a peculiar case. Given the WS cap on fielding WS affected the Beaneaters regularly, one would expect that WARP would like Duffy better than WS, but the opposite is true. I have no idea why.

Didn't the Bostonians consistently win more than they were pythag'd at? I know the Cubs won 11 more than they were pythag'd to win over that 10 year period.

<i>Posted 5:55 p.m., March 26, 2004 (#183) - karlmagnus
   151. Paul Wendt Posted: March 27, 2004 at 12:54 AM (#523294)
Marc #141
   152. Chris Cobb Posted: March 27, 2004 at 01:54 AM (#523295)
Brown's a guy that got about 20-22 wins through offensive & defensive help in his career. And he still ended up with about 220 adjusted wins & an adjusted winning percentage around .600. He'll stay high on my ballot.

Chris J., that estimate matches mine very closely: I think Brown got 23 wins through offensive and defensive help, giving him a support-neutral record 216-153. There we agree. Nevertheless you place him high on your ballot, while he falls just short of mine. The difference is not in the record we see, but in how we're valuing pitchers. How, given Brown's record, do you decide where to place him with respect to position players? How do you turn a support-neutral career (or season-by-season) pitching record into an assessment of merit or value?
   153. jimd Posted: March 27, 2004 at 02:04 AM (#523296)
More comments.

Chris J. WOW. Possibly the greatest team ever in the sense of the best "worst position player" (maximum-minimum sense). I'd love to see a Win Shares analysis of the Red Stockings of the NA to see how the weakest links compare. Too bad that has to be done from scratch.

I am strongly inclined to believe that WS undervalues solid middle infielders based on this analysis. But...is this chronic, or just deadball?

My overall impression of the two is:
   154. Paul Wendt Posted: March 27, 2004 at 02:13 AM (#523297)
How many plays are affected by catcher defense? What's the variation on range in pop-ups? It's basically throwing arm, and an occasional bunt - but even there it's usually a routine play.

That changed a lot in the 1870s-80s-90s. In this respect, I can't judge whether conditions for Deacon McGuire 1890s were closer to conditions for Doug Allison 186x-187x or for modern catchers.

jimd #184
   155. jimd Posted: March 27, 2004 at 03:05 AM (#523298)
That I don't know.

Its linear assumption has a zero-point below which no credit is given.

What I mean by this is as follows. As the team's offense gets worse and worse, it gets less credit, and would receive no credit if it fell below half of the league average run level (according to the main theoretical outline that is not always true because of some fine print for handling really bad teams). Defense follows the same principle, with less and less credit reaching zero at 150% of the league average run level (again, there is a fine print override for really bad teams). These same zero-points are built into the individual calculations also (though not on fielding as far as I've noticed).

A team with average defense (pitching/fielding) and a zero-point offense would be a .200 team, according to pythagorus. A team with average offense and a zero-point defense would win at a .308 clip, again according to pythagorus. This is how these "replacement" level equivalents are derived. (This is not my work, but that of others I've read on Win Share discussions; if I've gotten it wrong, that's my fault.) A team that is below zero-point both ways may still win some games, which is one reason for the fine print.
   156. jimd Posted: March 27, 2004 at 03:15 AM (#523299)
Players earn Win Share credit for their contributions above the zero-points (the effective "replacement" levels, though they actually have nothing to do with real-world replacement, per se).
   157. Dag Nabbit is part of the zombie horde Posted: March 27, 2004 at 04:03 AM (#523300)
Nevertheless you place him high on your ballot, while he falls just short of mine. The difference is not in the record we see, but in how we're valuing pitchers. How, given Brown's record, do you decide where to place him with respect to position players?
   158. Paul Wendt Posted: March 27, 2004 at 04:29 AM (#523301)
<i>
   159. Devin has a deep burning passion for fuzzy socks Posted: March 27, 2004 at 05:39 AM (#523302)
Just thought I'd post this from yesterday's Prospectus chat, leaving my incoherent grammar intact:

Devin (Green Brook, NJ): Clay, I know only me and 5 other people care about this, but when are you going to fix that problem with Pete Browning's player card where he's getting -37 runs for pitching 1 inning?

Clay Davenport: I have it fixed on my machine right now ? something I should have done long before, with a reprojection into his own game status (Since he only played one game, he cannot have a win rating worse than negative point 5 above average, right? And if you follow a 10 run/win pattern strictly, it means a player who only had one game cannot range more than +/- 5 runs. Since I tweaked some other things ? I found some biases in my defensive ratings that correlated with park effects ? I?m still looking them over for problems before releasing them.

I'll put in the WARP3 thread as well.
   160. Devin has a deep burning passion for fuzzy socks Posted: March 27, 2004 at 05:45 AM (#523303)
I'll put in the WARP3 thread as well

Well, I would if I could find it.
   161. Jeff M Posted: March 27, 2004 at 06:17 AM (#523304)
Chris and Chris:

I'm not sure I entirely understand your evaluation of the number of hits allowed vs. the number expected. I typed a long detailed post in response to Chris Cobb's question about Brown vs. McGinnity, but the Web lost it, so this will be more of a summary.

Don't the kind of hits matter? I see WARP's numbers show that McGinnity gave up significantly fewer hits than expected in front of an average defense (by about 120 hits). But isn't it relevant that WARP shows Brown gave up significantly fewer <b>runs<b> than expected in front of an average defense than McGinnity did (about 50 runs)? Using NRA and DERA from BP, it looks like McGinnity gained a 5% advantage from his defense and Brown gained a 10% advantage. It's a significant difference, but I'm not sure you can go overboard with that. McGinnity and Brown's primes are very very similar, but Brown played longer.

I still look at Linear Weights too, adjusted to take into account NRA and DERA. Brown is way ahead on this measure, primarily because he had a longer career and very few instances of actually hurting his team. McGinnity was a poor performer in 1905 with 320 IP, a decent performer in 1906 with 340 IP, a very poor performer in 1907 with 320 IP and a neutral performer in 1908 with 180 IP. He went 77-52 in those years, but his Wins Above Team for those years was -1. Overall, he hurt his team during those years more than he helped.

Of course, Brown had a decline phase too, but he didn't have a truly awful year, plus he pitched fewer innings in that decline phase, so his team wasn't hurt as much by his mediocre-to-poor performances during the decline phase, the way McGinnity was. Brown was 49-40 during his decline phase, but actually had about 6.5 Wins Above Team. Overall, he seems to have managed to help his team during the decline phase.

All of this is in light of me viewing their peaks and career numbers as fairly similar.

I don't want to give the impression that I don't like McGinnity. I had him at #7 in the last election and he has been as high as #2 on my ballot. I'm just explaining one of the reasons I have Brown ahead of him by a little bit.
   162. John (You Can Call Me Grandma) Murphy Posted: March 27, 2004 at 06:48 AM (#523305)
Use a quota! (How many pitchers should be in the Hall of Merit?)

I don't use a quota per se, but I did rig my system so that more pitchers would make it to my ballot.
   163. EricC Posted: March 27, 2004 at 03:28 PM (#523306)
April 1, 1923 prelim:
   164. Chris Cobb Posted: March 27, 2004 at 04:44 PM (#523308)
Chris and Chris:

I'm not sure I entirely understand your evaluation of the number of hits allowed vs. the number expected. I typed a long detailed post in response to Chris Cobb's question about Brown vs. McGinnity, but the Web lost it, so this will be more of a summary.

Don't the kind of hits matter? I see WARP's numbers show that McGinnity gave up significantly fewer hits than expected in front of an average defense (by about 120 hits). But isn't it relevant that WARP shows Brown gave up significantly fewer runs than expected in front of an average defense than McGinnity did (about 50 runs)? Using NRA and DERA from BP, it looks like McGinnity gained a 5% advantage from his defense and Brown gained a 10% advantage. It's a significant difference, but I'm not sure you can go overboard with that.


Can't speak for Chris J., but here's how I see it. I look at hits expected because it's the only quantitative measure we have of the contribution of fielders to run prevention. I use defensive efficiency to find hits saved above average, then use the average run value of a hit, calculated using the runs created formula, to find runs saved on the basis of hits prevented. It's not a perfect measure, but it moves the evaluation of the pitcher in the right direction. My system, because it compares the pitcher's record to that of an average pitcher using the pythagorean method, does give the pitcher credit for whatever runs he saved above what would be expected from the number of hits -- any runs saved not credited to the fielders purely on the basis of hit prevention are credited to the pitcher. I don't look at the number of those runs directly, since I'm working from the pitcher's W-L record, but runs saved will be reflected there.

McGinnity and Brown's primes are very very similar, but Brown played longer.

Well, he played more seasons, but he actually pitched fewer innings, so his greater number of seasons doesn't give him an edge in _value_ over McGinnity. As I see it, the reverse is true. McGinnity gave his teams more help when replacing a pitcher is difficult -- during a season -- where Brown's added durability was across seasons, when pitcher replacement is easier. I wouldn't argue strongly that valuing Brown for having a longer career is bad, but I can't see valuing seasons over innings as a measure of career length for a pitcher, at least when comparing pitchers from the same era.

I still look at Linear Weights too, adjusted to take into account NRA and DERA. Brown is way ahead on this measure, primarily because he had a longer career and very few instances of actually hurting his team. McGinnity was a poor performer in 1905 with 320 IP, a decent performer in 1906 with 340 IP, a very poor performer in 1907 with 320 IP and a neutral performer in 1908 with 180 IP. He went 77-52 in those years, but his Wins Above Team for those years was -1. Overall, he hurt his team during those years more than he helped.

I know nothing about Linear Weights, but my system sees these three seasons quite differently. It shows McGinnity as an average pitcher in 1905, a very good pitcher in 1906, and a poor pitcher in 1907. Moreover, as my system sees it, giving your team 320 IP at league average value is nowhere near close to hurting the team. Wins Above Team is probably a misleading measure for McGinnity during those years, since Mathewson was reaching his prime and the team overall was quite good. I have McGinnity, compared to an average pitcher throwing the same number of innings with the same offensive and defensive support, as 3 wins above average for those three seasons, cumulative. That doesn't _add_ much to his case, but it doesn't detract from it, either.

Of course, Brown had a decline phase too, but he didn't have a truly awful year, plus he pitched fewer innings in that decline phase, so his team wasn't hurt as much by his mediocre-to-poor performances during the decline phase, the way McGinnity was. Brown was 49-40 during his decline phase, but actually had about 6.5 Wins Above Team. Overall, he seems to have managed to help his team during the decline phase.

See above on McGinnity's late performance. Brown's WAT in the latter part of his career are helped by the fact that he pitched for a mediocre team in one season and pitched in the Federal League for two years. If McGinnity's level of competion had dropped by about 10% for two of his thee decline-phase seasons, I expect he would have looked better.
   165. Chris Cobb Posted: March 27, 2004 at 04:54 PM (#523310)
Joe wrote:

Chris, I think estimating 19 WS for an average position player is too high - rather, I think you're comparing an average pitcher to an average position player by this method may be off.

Joe, in figuring these average values, I have worked through a process similar to the one you went through in your post (#166). However, I shift the pitching/fielding split from 67/33 to 55/45 for this era, and I use a calculated average for innings pitched based on the top three starters in IP for each team, which shows that an average pitcher was throwing 17-18% of team innings during this period, not 20%. You may disagree with either of these premises, but if you work through the math on this basis, 15 and 19 as average values are what you get (I use 90% of games played to work in average playing time for position players. That hasn't been checked against actual numbers, so it could be a bit high).

Also, you mentioned the 'unhittable' aspect, and how Brown wasn't as unhittable - but what about Brown's control, which was very good - he was top 4 in the league in BB/9 6 times, 2nd three times. McGinnity's wasn't as good, and his strikeout rate wasn't nearly as good, especially considering that his defense was converting few of his BIP into outs (giving him more whiff opportunities).

"Unhittable" was a poor and misleading word choice here. What I should have said was that Brown doesn't have as many dominant seasons, when he was far and away better than an average pitcher would have been, both in terms of quality and in terms of durability. The exceptional pitchers of the era all tend to have a few seasons in which they have more than 40% more value than an average pitcher. Here are the numbers I have, using my methods, for some of the top pitchers in their best four years:

Mathewson 1.77, 1.63, 1.59, 1.57
   166. OCF Posted: March 27, 2004 at 04:54 PM (#523311)
From Chris J.:

<i>So using batter park factor, the league R/G averge, & all the rest - here's Mordecai Brown's Adjusted W/L record from 1904-10 (actual wins in parathesis):

1904..14-11 (15)
   167. Dag Nabbit is part of the zombie horde Posted: March 27, 2004 at 06:30 PM (#523313)
<i>Can't speak for Chris J., but,/i>
   168. Dag Nabbit is part of the zombie horde Posted: March 27, 2004 at 06:42 PM (#523314)
Just to clear my head - new provisional ballot:

1. Honus Wagner.
   169. Chris Cobb Posted: March 27, 2004 at 08:07 PM (#523315)
1923 Prelim Ballot

Strongest ballot we've seen since 1898, and the total pool of talent past the top 15 is amazing! This is the strongest we'll see until 1934, It's an easy year at the top, fortunately. To keep this from being way too long, I've commented only on new arrivals or players who've been under heavy discussion.

Shoo-Ins. Only one goes in this year, but the other will have a good shot at pulling the top vote on 90% of ballots in 1924.

1. Honus Wagner. Best player so far eligible. Greatest shortstop of all time. 740 cws, 293 total peak, peak rate, 00-09 = 51.05 ws/162. I've been astonished by Wagner's numbers, but this last was most astonishing to me. Wagner has all those huge win-share totals 1900-1910, and I knew he never had an off year or a serious injury, so I assumed that he was playing 150 games a year all down that stretch. But he wasn't: he was playing 135-140 games, typically (except for his monster 1908). So his WS rate stat is considerably higher than his league-leading raw stats. Mind-boggling. I haven't gone back and calculated this stat for all the greats of the 1870s & 1880s who were elected before I joined the project, but from the time I started calculating this stat until the 1922 election, Hughie Jennings had the highest total, 41.19 WS/162 over five years. Last year when Nap Lajoie topped Jennings' rate by a tiny margin (but for a ten year run), I was _very_ impressed. Wagner totally eclipses them both.
   170. Marc Posted: March 27, 2004 at 08:40 PM (#523316)
>Sheckard' mediocre 07-08 does explain why Cobb didn't rate him, I agree.

karl, even thius is too much to say, really. 10 mediocre games in 2 World Series are why Cobb didn't attribute much of the Cubs' success to Sheckard. It is true that he was "mediocre" those two years (i.e. about an average OF in '07 and even below in '08). so perhaps Cobb's opinion would have been the same if he had seen 300 games.

But the larger point, also, is that Cobb also attributed much of the Cubs' success to their baserunning. That too was based on 10 games and in this case 10 games that were not typical of the Cubs, in that they stole twice as many bases per game in the Series as they did in either season.
   171. Marc Posted: March 27, 2004 at 09:16 PM (#523317)
Rube Foster is an interesting case. And judging players by analogy is probably even worse than judging them by quota or pool, but there are some interesting analogies.

One is Harry Wright. Foster and Wright are perhaps overrated as players (like Frank Chance also) if you are blinded by or add extra credit for their leadership achievements, which in the cases of Wright and Chance at least their peers and immediate descendants did (overrate them on that basis). Wright was not the best but only the third best player of his era, after Start and Pearce (but Wright and NOT Start and Pearce is in the HoF). By analogy, maybe Foster ranks behind Johnson and Monroe and/or Grant, depending on how you define his era, but ahead of the rest (those who are eligible now)? Well, who knows if this is a good analogy or not?

Another interesting analogy is Bob Caruthers, another short career/huge hitting pitcher. By analogy, Foster could rate anywhere from top 3 to top 30, depending on your persuasion. It is hard for me (just speaking for me) to picture Foster as better than Caruthers.

Chris offers the Ed Walsh analogy. I think, yes, Rube Foster was Ed Walsh...against vastly weaker opposition.

I had Rube at #15 on my prelim, as strictly a place-holder. Thanks to Chris for a nice summation. It would take a lot of extrapolation to say, as Rube seems to imply, that Rube would have stayed in playing shape longer if he had been able to play in the bigs and then to give him any X-credit for that. I can't give the X-credit. But I am a peak voter (mostly) and so the Caruthers analogy works for me. But returning to the Harry Wright analogy, just as Wright was no Start or Pearce strictly on the basis of his playing record, and just as Caruthers was neither John Clarkson nor Dan Brouthers, then so too Rube Foster probably did not have the career value of Johnson or Grant.

Net-net. For the moment Rube will remain down around #15, but he will move up between now and '32 and he may also leapfrog some of the backlog as I learn a little more. But OTOH then as Mendez and Brown become eligible, I will be reminded that the "experts" rated Foster behind both. And I do not see him ahead of Torriente or Hill. So longer-term and acknowledging that with imperfect knowledge my rating could change a lot, right now he looks like a guy who might peak in the #6-8 range then start sliding after '32. IOW I agree with much of what Chris says, but what Chris says, to me, adds up to a little better package than a #24 this year.
   172. Paul Wendt Posted: March 27, 2004 at 10:40 PM (#523318)
Perhaps under the influence of Honus Wagner, this fortnight's discussion is not focused narrowly on the contest . . .
   173. Chris Cobb Posted: March 28, 2004 at 03:25 AM (#523319)
OCF wrote:

Incidentally, I do have Vic Willis with an RA+-equivalent record of 262-182, for 235 eq.FWP. His best three seasons, in eq.FWP, are 40, 34, and 31 in 1899, 1906, and 1901. Willis pitched in front of some good defenses, too: does anyone have estimates for how many runs those defenses saved?

Here are my estimates for Vic Willis teams? runs saved derived from hits saved

1898 147
   174. OCF Posted: March 28, 2004 at 04:16 AM (#523320)
Home runs wouldn't be that big a story, even though Boston was leading the league in both HR hit and HR allowed.

In 1898, they hit 53 HR and allowed 37; league average was 22.
   175. Chris Cobb Posted: March 28, 2004 at 05:30 AM (#523321)
So it's a factor, but looks small next to those massive numbers of defensive runs saved you mention (147, 184, 132).

Thanks, OCF for those home run numbers! Using them to make some very rough estimates on their role in the increase of scoring in the Boston ballpark, I would suggest dropping the runs saved by fielding numbers I've provided by 15%. Even with that reduction, they are still better than outstanding, but the reduction is large enough to have a discernable effect on an evaluation of Willis, and so it's worth including.
   176. Dag Nabbit is part of the zombie horde Posted: March 28, 2004 at 05:46 AM (#523322)
Re: Vic Willis & defense.

I'm no help to you in pre-1903 baseball, but I got some good info from then on. As mentioned in the longass post on Tinker from 1921, I've got the team H% of every team that retrosheet & b-ref list team TBF for (which begins in 1903). From there you can compare a team's H% veruss the average of the rest of the league (note: note using actual league average, just average the teams H%s rather than adding up all BIPs & hits off of BIPs). Here's how Vic Willis's teams faired from 1903 onward (another note: this approach completely ignore park factor which can be very significant. I didn't bother to check it for Willis's teams, ah well):

1903..+58
   177. OCF Posted: March 28, 2004 at 06:32 PM (#523323)
I re-did Willis. I used the defensive numbers from Chris Cobb's post #211, discounted by 15% in 1898-1900 and by 10% in 1901. I used the same process described for Brown in post #203. In what follows, the first set of numbers is an RA+-equivalent W-L record and eq.FWP without the defensive adjustment. The second set of numbers is the same with the defensive adjustment.

Willis Unadjusted Adjusted
   178. Marc Posted: March 28, 2004 at 08:25 PM (#523325)
Tom, I'm assuming you picked VH and Sheckard more or less at random, not that it matters. But in any event the obvious question from your post is:

Are we overrating Sheckard or underrating VH or both?

And what about VH vs. his alter ego Ryan and contemporary CF Duffy? We started out with Ryan above VH, then VH surged a bit when the new WARP came out and now I think Ryan has moved back ahead.
   179. John (You Can Call Me Grandma) Murphy Posted: March 28, 2004 at 08:27 PM (#523326)
It seems that in the modern game, he would still have been a 1Bman, or possibly a LFer in a small ballpark.

Tom, I don't disagree with any of your points, but if it is fair to downgrade Beckley because his value would have possibly been less today than in his own time, isn't it also fair to do the same thing with contemporary players compared to earlier eras.

IOW, why I love today's' game, I don't see that it is necessarily better or worse than the games played at other times. Would some of today's first baseman be able to handle the responsibilities that Beckley had to endure?
   180. Howie Menckel Posted: March 29, 2004 at 02:18 PM (#523328)
Why wait for the results?
   181. Paul Wendt Posted: March 29, 2004 at 06:03 PM (#523331)
tidbit:
   182. Brad Harris Posted: March 29, 2004 at 06:11 PM (#523332)
It's been a few "years", but here's who I'm considering this time...

(1) Honus Wagner ? one of 3 best players in history
   183. Dag Nabbit is part of the zombie horde Posted: March 29, 2004 at 08:24 PM (#523333)
4) Three-Finger Brown ? not fooled by his detractors

No one's trying to "fool" anyone here. If you disagree that's fine, but it's a bit much to argue others are trying to fool you.
   184. OCF Posted: March 29, 2004 at 11:23 PM (#523334)
Chris, when you try to write one-line comments, it's sometimes hard to pick the right word. The debate that you, Chris Cobb, I, and several others have been having over the last two ballot cycles resists compression into one line. I wouldn't read too much into Brad's choice of the word "fooled".

--

Who asked if there were any friends of Johnny Evers? Maybe I'll be his friend. If you look for runs created above average per out, scaled to a neutral run environment, Evers pops out as a surprisingly good offensive player. His flaw is that he missed time during the seasons - quite a few 125-130 game seasons. For starters, I'll definitely rank him ahead of Cupid Childs.
   185. Marc Posted: March 30, 2004 at 03:49 AM (#523336)
I think I asked this very question based on KJOK's spreadsheets. They certainly help rank Negro Leaguers relative to one another. But how do you use that data to rank them compared to major leaguers? Well, I think you discount them somewhat, just like the AA. AA discounts vary from 0-30% on my chart and average about 15%. I haven't seen any data that justifies it, but I've heard it said that the FL discount is/should be about 15%. Were the Negro Leagues prior to 1920 about comparable with the FL? Well, I would guess (and it is only a guess) that the replacement level would be lower than that, but that the very best players would probably be within 15% of the very best white players in terms of value.

Now, Mark, didn't say 15%. 15 OPS+ points would actually be 7.5%, right? But that's a quibble right now relative to the larger questions of 1) whether we should accept the kinds of figures that KJOK and i9 provide, and 2) to discount them. Only if you answer yes and yes does the question of 3) how much to discount them, matter.

But with Mark's proposed discount OC still rates ahead of TS! So it seems like a worthy proposal to discuss. I mean, really, everybody probably has a visceral response to that question. Whose better--OC or TS?!?
   186. Marc Posted: March 30, 2004 at 03:49 AM (#523337)
I think I asked this very question based on KJOK's spreadsheets. They certainly help rank Negro Leaguers relative to one another. But how do you use that data to rank them compared to major leaguers? Well, I think you discount them somewhat, just like the AA. AA discounts vary from 0-30% on my chart and average about 15%. I haven't seen any data that justifies it, but I've heard it said that the FL discount is/should be about 15%. Were the Negro Leagues prior to 1920 about comparable with the FL? Well, I would guess (and it is only a guess) that the replacement level would be lower than that, but that the very best players would probably be within 15% of the very best white players in terms of value.

Now, Mark, didn't say 15%. 15 OPS+ points would actually be 7.5%, right? But that's a quibble right now relative to the larger questions of 1) whether we should accept the kinds of figures that KJOK and i9 provide, and 2) to discount them. Only if you answer yes and yes does the question of 3) how much to discount them, matter.

But with Mark's proposed discount OC still rates ahead of TS! So it seems like a worthy proposal to discuss. I mean, really, everybody probably has a visceral response to that question. Whose better--OC or TS?!?
   187. Chris Cobb Posted: March 30, 2004 at 04:15 AM (#523338)
I haven't read enough on OC yet to have a strong opinion on Marc's specific question, but I think this question of how to benchmark the various sorts of negro-league statistics and translated statistics to major-league statistics is a topic we ought to be discussing.

Mark's discount approach passes my smell test as well; I see the I9s numbers as a bit higher than I am prepared to accept at face value.

However, it's worth noting that Bill James rates Oscar Charleston ahead, not only of Tris Speaker, but of Ty Cobb. He has Charleston at 4 and Cobb at 5 in his top 100 list. So the I9s career numbers are pretty much in keeping with James's view.

Not that his view _must_ be accepted, but one expert whose opinions carry a fair bit of weight around here basically supports the idea that I9s numbers be taken at face value, at least in the case of Charleston. Whether those numbers would be reliable for comparing one negro-league player to another is another question, of course.
   188. KJOK Posted: March 30, 2004 at 06:53 AM (#523339)
I plan on revisiting the Negro League stats in the near future AFTER I do my own studies of the AA vs. NL and FL vs. AL/NL as the "strength assumptions" for the Negro Leagues greatly impact the results.

For now I will only point out that if you increase the discount for Negro League players you quickly find the results for the "2nd tier" guys like Bingo DeMoss, Jelly Gardner, Pelayo Chacon, Dave Malarcher, Doc Wiley, Morten Clark, Luther Farrell, etc. translate to stats that make it appear they couldn't even hit well enough to be starters in the major leagues, much less stars.
   189. Carl Goetz Posted: March 30, 2004 at 03:31 PM (#523340)
Is there a good site on the web for Negro League Stats?
   190. Marc Posted: March 30, 2004 at 04:11 PM (#523341)
KJOK, my comment more or less "endorsing" Mark's 15% discount was specifically for pre-1920, though of course I only speak for myself on that. My hope is that we have more data after about 1920 that will enable us to fix the discount rate with more precision and perhaps on a season-to-season basis. Now, that's not to say that 15% is too much after 1920, either--only that I wouldn't make that as a presumption at this time.

Of course a 15% discount on Rube Foster's 59-1 season may be too little, no?
   191. Dag Nabbit is part of the zombie horde Posted: March 30, 2004 at 04:12 PM (#523342)
OCF, good point. Apologies to Brad for overreacting.
   192. Paul Wendt Posted: March 30, 2004 at 06:24 PM (#523343)
tidbit #223, corrected and extended:
   193. Jim Sp Posted: March 30, 2004 at 08:37 PM (#523344)
What exactly do various people mean by a 15% discount? For example, 85% of Win Shares is a different thing from taking 85% of WARP, since the replacement levels are different. Or discounting runs created by 15% would be more drastic...

Paul's tidbit makes me think Wallace was unusually slow for a shortstop...anyone have other evidence for that?
   194. Chris Cobb Posted: March 30, 2004 at 09:05 PM (#523345)
What exactly do various people mean by a 15% discount? For example, 85% of Win Shares is a different thing from taking 85% of WARP, since the replacement levels are different. Or discounting runs created by 15% would be more drastic...

Good question, Jim, and I hope others will answer!

I would suggest that the discounting that we will have to be concerned about first and foremost (especially for early negro-league players) is batting average. That's often the only statistic we have. Sometimes it comes with at-bat and hit numbers, sometimes not.

It looks to me like i9s is using about a 20% discount to turn the reported batting averages from 1910-1913 against all competition published in Riley (for players like Lloyd and Johnson) into their major-league equivalent statistics on their site.

William McNeil gives a rule-of-thumb of subtracting 50 points from negro-league averages (this applies to real league play, I think) to get MLEs. That looks to me like a percentage discount of about 15%. One could crosscheck the i9s numbers against the numbers Holway provides to see what sort of a discount the i9s folks are using for actual negro-league play. I'm guessing it's a bit smaller than what McNeil uses.

Getting from batting average to projections of a whole offensive output is another matter, one I'm entirely unqualified to speak about, so I'll stop there.
   195. Marc Posted: March 30, 2004 at 10:57 PM (#523346)
Jim Sp, your point is well taken. The original discounting concepts were published in The Hidden Game by Thorn and Palmer though the study cited was actually by Dick Cramer. It discounted BA and SA working backward from a 1980 norm of "0". (Just for perspective, Cramer proposes a discount of about .100 for "now" [1923], making Rogers Hornsby, e.g., about a .285-.320 hitter. Also that means about a 25-33% discount 1920 versus 1980. So you can take or leave his particulars.)

More to the point, the discounts we're talking about have more often been league adjustments--two leagues at the same point in time. Cramer had the AA versus NL adjustments ranging from the AA being better in '86 to 30-35 points at its worst. The Federal League is 30-50 points sorse than the AL-NL, meaning comparable to the AL-NL at the turn of the century.

As an aside, he has the heart of the war years ('44-'45) as being 20 points worse than '41-'42 but essentially equal to '39-'40.

All of that aside, the concept, then, was to discount BA and SA numbers generally, and the 20 points '41-'45 would represent about a 5-7% decline during the war.

Now I use a fairly substantial BS factor (a la Bill James), myself, so I'm not terribly stuck on the numbers--any numbers. So I will take WS or WARP and discount them by some reasonable number. For the AA it was 0-30%, e.g. But since I look at WS, WARP, traditional stats, peak and prime as well as career, even TPR, the discounts just function as a ballpark. But especially if you're a peak voter and you like Fred Dunlap (yes and yes), and assuming you find WS and WARP to be useful, you've got to be able to get those numbers into the realm of a fair comparison (the UA discount is 35%). So, yes, I discount WS and WARP and then compensate for the obvious issues with that by capturing a wide range of corroborating data.
   196. Jim Sp Posted: March 31, 2004 at 03:38 AM (#523349)
Jim, by what measure was Beckley a better hitter? I don't see that.

I could be wrong about that, I just took a quick glance at career OPS+, which Beckley wins 125 to 120. Sheckard has two big years (1901, 1903), on the other hand. Beckley's got about 400 more runs created, 850 more hits, so career is no contest. On the whole I don't see that Sheckard's 2 big years compensate for Beckley's production over his career. I'm out of line with the consensus on him, but it seems to me that his career accomplishments outweigh his lack of big seasons.
   197. Dag Nabbit is part of the zombie horde Posted: March 31, 2004 at 03:42 AM (#523350)
I could be wrong about that, I just took a quick glance at career OPS+, which Beckley wins 125 to 120.
   198. Jim Sp Posted: March 31, 2004 at 06:25 PM (#523351)
Before I get flamed (or at least to redirect the flames), I am aware that Win Shares gives the career lead to Sheckard. My interpretation is that the Win Shares system gives inadequate credit to 1B defense in the dead ball era, and that Sheckard is getting too much credit for the Cubs wins during their great run. The fielding cap and lack of 1B defense adjustment causes wins that should go to Tinker, Evers, and Chance to go to Sheckard. Or maybe it's something else, anyhow, when I look at their career numbers I can't see putting Sheckard's career value ahead of Beckley, it seems like it has to be a fluke of the Win Shares system.

Now of course my argument against Sheckard also works against Brown, who I put ahead of Plank, who pitched for some pretty good defenses as well...I see Brown as much more exceptional than Sheckard, but the more time I spend on this the less I know. Doing my best with the time I have, I guess is what I'll say.

I should also mention that in my system Beckley rates high because he is much better than his contemporaries at 1B, Sheckard doesn't standout compared to other outfielders.
   199. John (You Can Call Me Grandma) Murphy Posted: March 31, 2004 at 06:37 PM (#523352)
I should also mention that in my system Beckley rates high because he is much better than his contemporaries at 1B, Sheckard doesn't standout compared to other outfielders.

That's what my system "says," too (though I like Chance better than Beckley).
   200. MattB Posted: April 01, 2004 at 02:55 PM (#523353)
An interesting look at pitcher's Adjusted Run Support from 24-7 Baseball, giving the Top 100 pitchers in best and worst run support, career, and individual season, adjusted for offensive environments.

Not too many guys we've considered are on the list, but some are, and it's interesting to see the scope of the run support variations.

The best season ever was Chuck Dobson in 1971, whose 165 adjusted run support gave him a 15-5 record despite an 87 ERA+.

What a glance at the list tells me is that there was no single era in which variations in run support had a larger effect on individual pitchers -- the group is pretty well scattered randomly throughout baseball history.
Page 2 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
greenback calls it soccer
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.3319 seconds
49 querie(s) executed