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, June 12, 2006

1979 Ballot Discussion

1979 (June 26)—elect 2
WS W3 Rookie Name-Pos (Died)

642 208.1 1951 Willie Mays-CF
293 85.0 1956 Luis Aparicio-SS
297 73.9 1960 Frank Howard-LF/RF
241 70.6 1959 Johnny Callison-RF
241 59.2 1958 Felipe Alou-RF/CF
210 71.1 1958 Milt Pappas-P
142 54.1 1960 Chris Short-P (1991)
125 47.1 1961 Ron Perranoski-RP
139 38.5 1966 Tommie Agee-CF (2001)
115 43.1 1964 Gene Alley-SS
139 33.1 1962 Joe Pepitone-1B
103 35.9 1961 Bobby Bolin-P
106 33.1 1960 Eddie Fisher-RP
098 32.2 1963 Ray Culp-P
105 28.2 1965 Rick Reichardt-LF
108 25.7 1967 Mike Andrews-2B

Players Passing Away in 1978
HoMers
Age Elected

74 1945 Bill Foster-P
63 1956 Joe Gordon-2B

Candidates
Age Eligible

91 1928 Jack Graney-LF
90——Joe McCarthy-P/HOF Mgr
86 1935 George Harper-RF
85 1943 Jesse Haines-P
84 1935 George H. Burns-1B
83——Ford Frick-HOF Commissioner
82 1943 Rube Walberg-P
75 1945 Carl Reynolds-RF/CF
69 1947 Monte Pearson-P
68 1951 Gene Moore-RF
68 1953 Bill Dietrich-P
68 1954 George McQuinn-1B
58 1961 Billy Cox-3B
49 1972 Jim Gilliam-2B/3B

Upcoming Candidate
27 1984 Lyman Bostock-CF/RF

As always, thanks to Dan!

John (You Can Call Me Grandma) Murphy Posted: June 12, 2006 at 12:09 AM | 268 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. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 06:11 PM (#2077776)
By the mid 1950's, you are back down to 16 teams. It only makes sense that an era with 30 or so major league equivalent teams is going to produce more players with impressive statistics than an era with 16 teams.

I think a part of the "problem" is that, I think, some of the white guys were only compared to their white contemporaries from the major leagues by some members of the electorate, ignoring the NeLers in the analysis. If you place domination at your position and overall place in the ML as a criteria in your system, the absence of NeLers will distort things.
   202. karlmagnus Posted: June 27, 2006 at 06:11 PM (#2077778)
My 9 worst of the 1930s list would be Averill, Bell, Dickey, Dihigo, Ferrell, Gehringer, Herman, Suttles, Wells. Still only 4 negro league players, so we seem to have been consistent in our overrating of the 1930s.
   203. Jim Sp Posted: June 27, 2006 at 06:13 PM (#2077783)
How about if we make the rule this:

New voters need to post a ballot during discussion week in order to be eligible for that year's ballot.

That makes sure there's time for discussion rather than trying to figure out if a ballot should count at the last minute.

IMHO a new voter should demonstrate familiarity with Duffy et. al and it's ok to come to the conclusion that we've already elected the people from that era that they would have voted for. After almost 80 years we haven't elected Duffy yet so it shouldn't be too shocking that a new voter thinks he doesn't belong in the top 15. But a new voter be capable of saying something like "well, I would have voted for Brouthers and Nichols, but Duffy, Van Haltren and Ryan don't quite cut it for me."

It seems to me that B Williams said something more like "I don't know much about 19th century baseball so I'm leaving Duffy off." If that's what he meant, I wouldn't count the ballot. On the other hand if he meant "I'm finding it hard to analyze 19th century baseball but right now I agree with those not voting for Duffy", then that's probably ok. It requires discussion to figure out what was meant, so let's not try to figure this out at the last minute. New voters can participate in the discussion and post sample ballots while they are catching up on 19th century and Negro league issues, then their ballot can start counting. I don't think the bar needs to be very high, but fairness to all eras would seem to require that a bar exists, and we make sure time for discussion with the new voters is built into the process.

I would say that a mild timeline is not unconstitutional, then again the constitution says what the judges say that it says :).
   204. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 06:23 PM (#2077792)
My 9 worst of the 1930s list would be Averill, Bell, Dickey, Dihigo, Ferrell, Gehringer, Herman, Suttles, Wells. Still only 4 negro league players, so we seem to have been consistent in our overrating of the 1930s.


It should be noted that all four of those NeLers are also in the HOF. I doubt that they were influenced by Chris and Doc's MLEs. :-)

As for Suttles and Bell, I wasn't that crazy about them myself. But I definitely would not say that they weren't near the borderline of electability.

BTW, Dickey? Gehringer?
   205. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 06:26 PM (#2077800)
New voters need to post a ballot during discussion week in order to be eligible for that year's ballot.

I think that's an excellent idea, Jim. Joe?

I personally don't like the idea of someone new popping in at the last second. That voter could be trying to sway the election a certain way instead of voting his or her conscience (BTW, Bernie's ballot indicates that he's not that type of voter).
   206. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 06:31 PM (#2077803)
I would say that a mild timeline is not unconstitutional, then again the constitution says what the judges say that it says :).

I agree with you, Jim, in this sense: I think there are more great players than there were 120 years ago. However, that doesn't mean that, on average, the greatest back then were not comparable to the greatest of today. IOW, if your system indicates that nobody from the 19th century could compete with the elite of today, then that's a very questionable system to me.
   207. Dizzypaco Posted: June 27, 2006 at 06:34 PM (#2077807)
I think a part of the "problem" is that, I think, some of the white guys were only compared to their white contemporaries from the major leagues by some members of the electorate, ignoring the NeLers in the analysis. If you place domination at your position and overall place in the ML as a criteria in your system, the absence of NeLers will distort things.

I may be opening up a can of worms here, but I have been wanting to put up a post about this for some time. I have followed the HOM threads pretty closely for some time, but have never voted. And a major reason I have never voted is that I have no idea how good individual Negro League players are. I am also very skeptical about other people's evaluations of how good they really were. I just don't know how the level of play of the Negro Leagues compared with the Leagues, given that the teams never played against each other in a meaningful game. I'm sure a lot of work has been done to establish MLE's for Negro League players, but my confidence in those numbers is not particularly high.

I'm sure that the best Negro League players were as good as the best Major League players, but if I don't trust the MLE's, how can I rate the players? And if I can't, how could I vote? I have a real good idea how to compare, say, Red Ruffing and Wes Ferrell. I don't have any idea how to compare them with a comparable Negro League pitcher. My instinct is that nearly everyone voting has overestimated the level of confidence in comparing the numbers, although I'm sure people can point me to the study supposedly showing me wrong.

The problem isn't that bad for the 1940's, given that many of these players eventually played in the majors, giving us a basis for comparison. The early the period, the more problematic it gets.

Every period has its biases - the absence of Black players, the absence of Latin players, the smaller number of people alive to play, the smaller number of people with access to baseball fields growing up, the number of teams, and on, and on. The only way to deal with it is just lay some assumptions, and go with them. For example, you can assume that all periods are relatively equal, and we will take about the same number of players from each era. But you have to make some assumptions, and just go with it, and be consistent. And anything you assume will lead to some distortion.
   208. sunnyday2 Posted: June 27, 2006 at 07:39 PM (#2077856)
>then that finger should be pointed at the Caucasian players.

Let's just say for the sake of discussion that we wanted to delete ten of them, to get back into balance with the other decades. It's not a foregone conclusion to me that it wouldn't be some of each. I sure hate to say karl was right, so I won't say it. I'll just say we maybe got a bit carried away.

Of course we were bound to elect X number of players. For me, a few more among Browning and Jones and and Waddell, etc., would have been good options. And more recently, I still think the WWII generation has suffered a bit, that we haven't really "adjusted" much for the guys who missed some time.

Also the black players whose primes coincided with early integration (1945-1955) have gotten a very very rough ride from us. I think the number of blacks elected drops from about 15-16 for 1935-1945 to about 5 for 1945-1955. It's not too late to look at those guys again though obviously it's too late to elect them ahead of Earl Averill or CP Bell.
   209. sunnyday2 Posted: June 27, 2006 at 07:42 PM (#2077859)
>>New voters need to post a ballot during discussion week in order to be eligible for that year's ballot.

>I think that's an excellent idea, Jim. Joe?

Absolutely yes. This solves pretty much any real problem and reduces the likelihood of hard feelings a lot.
   210. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 07:44 PM (#2077862)
Let's just say for the sake of discussion that we wanted to delete ten of them, to get back into balance with the other decades. It's not a foregone conclusion to me that it wouldn't be some of each.

I don't disagree, Marc. But the majority of them would be White players, IMO.
   211. Dr. Chaleeko Posted: June 27, 2006 at 08:10 PM (#2077885)
My instinct is that nearly everyone voting has overestimated the level of confidence in comparing the numbers.

I'm a creator of some of those numbers, an advocate for them, and believer in the mission to include NgL players in the fullest...but I do agree on this point with Diz. I think as a group we probably do overestimate the numbers we've generated. Partly this is a kind of flase-normalization (if that's the right use of the term) required to make the kind of imaginative leap necessary to compare two wholly independent cohorts of players. Partly, however, it's also a simple willingness to trust certain assumptions that came well before our own deliberations. James' MLE methods are 25 years old, and we have trusted the assumptions within them because he proved them well enough to us back then that we think the can work for us. If they work for the minors, they must work for the NgLs, or so goes the reasoning, I suspect. That is, if you could only figure out the NgL correctly, you're ready to plug players in. Once you take that step, it's not hard to take another and another and another until you're pretty sure that the logic makes sense. And to some degree it does.

On the other hand, has an overestimation of the MLEs done any harm? A look at the fellows we've elected versus the total population of Negro Leaguers and the total population of MLB+NgLs suggests that we're pretty much on track with them. And the player's we've elected support this: we have almost exclusively hit the biggest stars, by reputation and by our estimations, and not dipped into the secondary stars at all (working off the top o' my head on that one, but Bill Wright, Ted Page, Ted Strong, Chino Smith, Andy Cooper, Bill Byrd, Dandridge, et al aren't in). That's not true of some of the MLB players we've elected: Faber, Lyons, Ashburn, Bunning, probably Terry, Doerr, and McPhee were clearly not the biggest stars of their era.

There's a third possible reason for overstimation of our own numbers: what else is there? Answer: deeply flawed oral histories, overly glowing biographical sources, a mish-mash of far-flung and often incomplete minor- and foreign-league encyclopedias, an overblown annual account of dubious research quality, a seemingly reliable Lester/Clark book, and now SoG + a .pdf from the same study that profiles about 20-40 total players and finally gives us some benchmark leaguewide figures. Given how open various people have been about their sourcework and methodology around the MLEs, it's not difficult to see why these second- or third-generation creations might be considered more reliable than some of what's preceded them. We can see in front of us what the biases are, how the compilers are grappling with the idiosyncracies of the sources, and we can also ourselves contribute critiques of the method and see new iterations. That kind of process-oriented work builds a lot of institutional confidence, I think.

So I guess that the overestimation of confidence is a fundamental to taking the leap. Even Karlmagnus, one of our most skeptical voters has, at times, accepted some aspect of the MLEs as representative of the character of the player's career and/or "true" talent level.

Of course, as Sunny2 is wont to say: "we're dealing with outliers here," and it's possible that the overestimation effect is somewhat mitigated by dealing with a subset of players whose accomplishments and talent are going to go off of most scales geared toward projecting a player in terms of his relationship to an average player.

Though, as always, I could be wrong.
   212. Sean Gilman Posted: June 27, 2006 at 08:11 PM (#2077887)
I'm sure that the best Negro League players were as good as the best Major League players, but if I don't trust the MLE's, how can I rate the players? And if I can't, how could I vote?

This isn't an exact science, despite all the numbers and decimals and acronyms. I think most of us read as much of the available evidence as we could and then took our best guess as to how good these players were.

We do the same thing with the white players, actually. This whole project is far more subjective than we sometimes like to think it is (the debate over timelining is a good example of this, turning as it does on each of our interpretations of the Constitution). The HOM gets its credibility from the combination of all our subjective evaluations: the whole is more likely to be correct than any of our individual ballots.
   213. Dizzypaco Posted: June 27, 2006 at 08:43 PM (#2077908)
A couple of points:

First, I have noticed that when many voters post comparisons between major league players, they put up numbers - comparisons between win shares, WARP, OPS+, that sort of thing. But when NgL players are entered into the picture, I rarely see numbers, even MLEs.

Second, when people are ranking their top 100, it seems like the top ten or fifteen contain some NgL players, but numbers 16 to 100 are only or mostly major league players - they just kind of give up comparing trying to understand NgL players after a while. Not that this harms anything. I could be wrong on this one, but it seems that way.

Third, I think minor league MLEs are a good analogy for NgL MLEs from the 1940s - after all, in both cases, we can see how the players actually performed once they reached the majors, and make inferences regarding the quality of play based on that information. You can't do that for the 1920's, because none of the players ever played in the majors.

Fourth, I very much appreciate the efforts to come up with MLEs, and I realize that you need to do something to account for them. I just wouldn't be comfortable, even after looking at the MLEs, to say that a player should be placed fifth or eleventh on my list. The borderline players are a real problem. Yet nearly everyone who posts on HOM threads says with almost complete certainty, this guy should be ranked fourth, or seventh, or eleventh, or wherever. I have a feeling that even if Bill James had access to the MLE's, he still wouldn't include NgL players on the same lists as the others, because of these problems.
   214. DavidFoss Posted: June 27, 2006 at 08:48 PM (#2077916)
Indian Bob is the only 30s guy left high in the backlog and starting in the mid-80's we are *really* going to be digging into the backlog. The 30s will still be the most represented era, but the margin could lessen a bit.
   215. Dizzypaco Posted: June 27, 2006 at 08:50 PM (#2077917)
One additional point. There are a few people who have spent an enormous amount of time researching the NgL players and the games - Dr. Chaleeko, for example. I can see how they would have a pretty good idea how various players should rate. Yet I don't think this describes the majority of voters, and it is those people who I think are expressing far too much confidence in their opinions of how different NgL players should rate. I should add that most people who vote are clearly very knowledgeable baseball fans in general, so I'm not nearly as concerned in comparisons between Major League players.
   216. Sean Gilman Posted: June 27, 2006 at 08:54 PM (#2077921)
Yet nearly everyone who posts on HOM threads says with almost complete certainty, this guy should be ranked fourth, or seventh, or eleventh, or wherever

It may appear that way, but that's just a rhetorical illusion. The fact is that nearly every HOM voter has seen his or her (?) rankings and evaluations change over time, often quite significantly. How could that happen, if we're all so certain?
   217. Dizzypaco Posted: June 27, 2006 at 09:02 PM (#2077931)
It may appear that way, but that's just a rhetorical illusion. The fact is that nearly every HOM voter has seen his or her (?) rankings and evaluations change over time, often quite significantly. How could that happen, if we're all so certain?

Well I'm not saying people don't change their minds, but they do seem quite comfortable with their rankings at any one time, and I don't see how you can get to that comfort level.

One other thing and then I'll shut up.

I think a part of the "problem" is that, I think, some of the white guys were only compared to their white contemporaries from the major leagues by some members of the electorate, ignoring the NeLers in the analysis. If you place domination at your position and overall place in the ML as a criteria in your system, the absence of NeLers will distort things.

You have two choices. Either

a) You assume that statistics such as WARP, win shares, OPS+, etc. have roughly the same meaning over time - that is, an OPS+ of 130 means roughly the same no matter when it was accomplished.
b) You don't assume that they have the same meaning over time, due to changes in quality of play (due to such things as a lack of minorities in baseball).

If you go with A), the implication is that you will have roughly the same number of MLB players from each era, which also means you will overrepresent eras in which you have to add NgL players.
If you go with B), well, that a whole other can of worms.

I think most people are going with A), which causes the 1930's problem.
   218. Sean Gilman Posted: June 27, 2006 at 09:19 PM (#2077944)
Well I'm not saying people don't change their minds, but they do seem quite comfortable with their rankings at any one time, and I don't see how you can get to that comfort level.

Like I said, it's rhetorical. Our ballots are based on more or less objective information (stats of varying degrees of abstraction and complexity, translations for Minor Leaguers, Negro Leaguers and even, for some, between multiple Major Leagues) what we do with that information is almost entirely subjective. There's no particular reason to favor peak over career, or prime over peak, or Win Shares over WARP over OPS+ and ERA+. All approaches have their pluses, minuses and biases. You just have to figure out what seems best to you.

We all implicitly realize this, otherwise each of us would just make his own HOM on his or her own time and publish it somewhere for everyone else to admire. The fact that we choose to combine our ballots with those of 50 other people shows that we aren't totally certain about our own evaluations.

This illusion of certainty isn't confined to the HOM, but is an artifact of argument in general: it's hard to convince someone to vote for the candidate you advocate if your advocacy is littered with demurrals and equivocations. The nature of the debate requires us to act more certain than we can ever truly be.
   219. Dr. Chaleeko Posted: June 27, 2006 at 09:19 PM (#2077946)
The borderline players are a real problem. Yet nearly everyone who posts on HOM threads says with almost complete certainty, this guy should be ranked fourth, or seventh, or eleventh, or wherever.
Yes, this seems to be a problem as we go deeper into the backlog. We continually come against the uncertainty of certainty. I mean, you gotta post a ballot with 15 ranked guys on it, right? But I'm not sure that any of us really thinks that more than one or two (if that number) of the backloggers are definitely deserving. That's why they're in the backlog. This point, MAY buttress the smaller-hallers' argument, though I'm not certain that it does. Your standard for what gap should exist between the ins and outs will probably never be big enough or well-defined enough to allow any confidence in who is on which side of the borderline. But hey, at least in baseball we have stats to offer a hard-edged look along with the softer-edged subjective look at performance. Compare to, say, music where you see these long lists of 100-best rock records. Wow, talk about a no-accountability ranking model!

I have a feeling that even if Bill James had access to the MLE's, he still wouldn't include NgL players on the same lists as the others, because of these problems.

To be fair his top-100 player ranking in the BJNHBA did include several NgL players, MLEs or not.

I'm guessing Bill does have access to these MLEs either second hand or first (it's a free site after all), but I agree that he probably would chose not to use them as any definitive part of a player ranking, at the very least not without replicating the methodology himself. The player profiles of the NgL from his own NBJHBA are in many respects a distillation or patchwork of the same sources we've used. That's not a criticism of him. He had to work through forty-sixty years of blackball history in less than 30 pages. And he had to finish the rest of the book. Holway, Reily, Clark/Lester, these guys are devoting a lifetime of research to the leagues. That James sifted through this small pool of resources and made a good bunch of player lists shows that he's got a good idea of how huge and out of control the process of writing about the NgLs could have become if he didn't rein it in.

But I got the feeling that he was less than comfortable in making too many definitive statements about the NgLs probably for the very reasons Diz cites (not the uncertainty of the MLEs, but of the information undergirding them). For instance, I think he wrote that the only things he could say for sure were that Paige is one of the best (small number) pitchers ever, that Gibson was surely among the top (small number) catchers ever, and maybe one of the top (small number) players ever, and that Oscar Charleston was one of the best (small number) CFs ever. All of which are very defendable statements that virtually no one would or could attack as overstatements or as NgL pom-pomming.
   220. karlmagnus Posted: June 27, 2006 at 09:35 PM (#2077956)
(221) All three/four of which statements I would agree with, with a partial caveat on Paige, in spite of my Negro League skepticism. We do agree on some things, and the NgL section of the project has been an education.
   221. sunnyday2 Posted: June 27, 2006 at 09:36 PM (#2077958)
A HoM vote consists of a ballot with players ranking 1 through 15. This may create the appearance of "absolute certainty" as to who is #5 and who is #11. In the real world we're smart enough to know that absolute certainty is a chimera. As Chris Cobb often says, the MLEs are the best evidence we have. That doesn't make them perfect or anything near. They're the best we've got and the only alternative is to ignore the NeLers, and we're not gonna do that.

James does include NeL players in his top 100 in the NBJHA.
   222. Dizzypaco Posted: June 27, 2006 at 09:42 PM (#2077961)
Yes, this seems to be a problem as we go deeper into the backlog. We continually come against the uncertainty of certainty. I mean, you gotta post a ballot with 15 ranked guys on it, right? But I'm not sure that any of us really thinks that more than one or two (if that number) of the backloggers are definitely deserving. That's why they're in the backlog. This point, MAY buttress the smaller-hallers' argument, though I'm not certain that it does. Your standard for what gap should exist between the ins and outs will probably never be big enough or well-defined enough to allow any confidence in who is on which side of the borderline. But hey, at least in baseball we have stats to offer a hard-edged look along with the softer-edged subjective look at performance. Compare to, say, music where you see these long lists of 100-best rock records. Wow, talk about a no-accountability ranking model!

The point is well taken, butI think there's an important difference between evaluating borderline MLB players and borderline NgL players. If I'm comparing Ralph Kiner with Frank Howard, for example, I'll carefully examine their statistics, the evaluations of their defense, I'll compare how good each was in their prime, over the course of their career, and make some judgment over who was better - one that I am reasonably comfortable with. Others may disagree, but at least we are all speaking the same language. If, on the other hand, there is a player from the 1920's, who played in the NgL, who was a similar type of player as Kiner and Howard, I probably wouldn't be able to make a very educated judgment on where he fits in with Kiner and Howard. This problem is probably even worse for people whose value is more closely tied to defense.

I think my thoughts on the issue are pretty close to Bill James, as described by Dr. Chaleeko.
   223. DavidFoss Posted: June 27, 2006 at 09:50 PM (#2077970)
Yeah, putting MLB and NeL players in the same electoral pool is unique to this process. I don't think anyone else has tried to do that. It certainly was challenging, but I agree it was much more educational this way than putting them into separate pools. It sounds corny, but it makes it feel more like a single HOM instead of a MLB-HOM and a NeL-HOM.
   224. Jim Sp Posted: June 27, 2006 at 09:52 PM (#2077977)
I agree with you, Jim, in this sense: I think there are more great players than there were 120 years ago. However, that doesn't mean that, on average, the greatest back then were not comparable to the greatest of today. IOW, if your system indicates that nobody from the 19th century could compete with the elite of today, then that's a very questionable system to me.

Yes, we are in agreement. Anyone who can't come up with at least (say) 20 19th century players that they would elect isn't giving them a fair shake. For example anyone who doesn't think the following ten are no-brainers doesn't deserve a vote.

Kid Nichols
Dan Brouthers
Roger Connor
Buck Ewing
Cap Anson
George Davis
Bill Dahlen
John Clarkson
Ed Delahanty
Jesse Burkett

Oops, I just disqualified the entire HoF Veterans Committee.

:)
   225. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 10:53 PM (#2078028)
Oops, I just disqualified the entire HoF Veterans Committee.

:)


Heh.
   226. Dizzypaco Posted: June 27, 2006 at 11:03 PM (#2078035)
On the other issue everyone is discussing...

When comparing players across eras, everyone is making assumptions, yet almost no one is acknowledging those assumptions. When choosing players for the HOM, do you think:

a. The number of players from the Major Leagues, regardless of race should be relatively consistent (e.g. 25 from the 1890's. 1910's, 1930's, 1950's etc. Any Negro League players would be extra.
b. The number of players from the major Leagues and Negro Leagues combined should be relatively consistent. Keep in mind that under this system, the number of Whites would be much, much higher in the first few decades, when there were no Blacks playing anywhere, versus any later decade
c. The number of White players inducted from each decade should be relatively consistent
d. The number of players per existing team (either Major or Negro League) should be relatively consistent
e. There should be no attempt to maintain consistency over time.

Which of these five you choose will have a huge effect on who you think should be elected. It appears, from the results, that many voters are picking (a). This is causing the increase in elected players from the 1930's, and the perception that the 19th century is underrepresented (which I disagree with).

John, which do you agree with?

b. The
   227. Sean Gilman Posted: June 27, 2006 at 11:22 PM (#2078066)
The number of electees per year was determined by number D.

I like to think that the voters make their ballots based on the relative merits of the various candidates in themselves, not on how they fit into some theoretical quota or timeline. I guess that would be option E: not making any assumptions, treating a pennant as a pennant, being fair to all players and eras and so on.
   228. John (You Can Call Me Grandma) Murphy Posted: June 27, 2006 at 11:30 PM (#2078090)
John, which do you agree with?

E looks good. B is close, but since I believe there are more great players as we move toward our time, it's not really a good assessment of my views.

When I construct my ballot, I never set up a quota of so many players from each decade, BTW.
   229. DavidFoss Posted: June 27, 2006 at 11:31 PM (#2078095)
the perception that the 19th century is underrepresented (which I disagree with).

I'm not sure anyone said that. Starting in 1898 gave us a nice head start on Cooperstown and the "candidate gap" in the late 20s gave us an oppurtunity to focus directly on them for quite a few years.

I think the concern is that we (or at least me) simply want 19th century players to continue being considered. Their candidacies have not expired.
   230. OCF Posted: June 28, 2006 at 12:12 AM (#2078254)
I think I'm somewhere between d and e. Baseball in the 19th century - at least, professional baseball at the highest level - was largely a regional game, played by whites from the northeast and Great Lakes regions. It wasn't until after 1900 that that changed with the significant recruitment of white southerners (Cobb, Jackson), Great Plains residents (Johnson, Alexander), Californians (Cravath) and large numbers of African-Americans (most of whom were southern.) I'm more bothered by the shortage of 1940's players and more worried about whether we're going to elect enough integration-era National Leaguers than I am with 19th century players. Yeah, I do still vote for Van Haltren and occasionally Duffy, but they're not Hamilton or Delehanty, Their omission (or the omission of Childs, or McGraw, or Herman Long, or Charley Jones) would not be a major crime.

By the structure of our process, a number of 19th century players were initially largely rejected, but with the backlog still relatively manageable, survived in the backlog and until campaigns built around them gained momentum: Pearce, Pike, Start, Caruthers, Jennings, Griffith - and perhaps eventually Beckley. Some of these candidates fared better than those who were nearly elected but backslid: Van Haltren, Sewell. It's been a lot harder to survive the introductory year lately, because the backlog is so large. Is Frank Howard's candidacy already dead? Should it be?
   231. Brent Posted: June 28, 2006 at 12:30 AM (#2078331)
Dizzypaco (#209) wrote:

I just don't know how the level of play of the Negro Leagues compared with the Leagues, given that the teams never played against each other in a meaningful game. I'm sure a lot of work has been done to establish MLE's for Negro League players, but my confidence in those numbers is not particularly high.

As this project has gone along, one of the most exciting developments was finding new ways to check and validate the MLEs. The first versions were largely based on players who made the transition from the Negro Leagues to MLB in the late 1940s and early 1950s. It was a very small sample and there were concerns about how well the results could be extrapolated back to earlier decades. When we came to Alejandro Oms, however, we noted that he was playing in Cuba alongside major leaguers like Dolf Luque, Mike Gonzalez, Bobby Estalella, and also alongside American Negro League greats like Oscar Charleston, Jud Wilson, Mule Suttles, etc. Chris Cobb and Gary A were able to use this information to develop translations between 1920s MLB and the Cuban League. It was great to see that this second source of information confirmed decisions we had already made regarding players like Wilson.

For an even earlier period (1908-13), Gary A was able to compile box scores of games between Cuban League teams and major league teams that were visiting Cuba, giving us statistics like OPS+ against major league pitching and defense. (See Gary A's blog. Again, the results confirmed our earlier decisions to elect Grant Johnson and Pete Hill.

I think some of the research that has taken place on these threads has been exciting, even cutting edge. And I am far more confident of the Negro League MLEs than I was in the early days. Even with all this work, I think our evaluations require a mix of judgment, historical knowledge, and good sense in addition to statistical analysis. That's more true for the Negro Leaguers than for other players, but every era has its complexities. (Dare I mention that our latest crop of candidates includes a couple who were able to extend their careers by DHing?)
   232. John (You Can Call Me Grandma) Murphy Posted: June 28, 2006 at 12:51 AM (#2078432)
(Dare I mention that our latest crop of candidates includes a couple who were able to extend their careers by DHing?)

Not only should it be mentioned, but it should be stressed, Brent.
   233. Dizzypaco Posted: June 28, 2006 at 02:06 AM (#2078804)
For an even earlier period (1908-13), Gary A was able to compile box scores of games between Cuban League teams and major league teams that were visiting Cuba, giving us statistics like OPS+ against major league pitching and defense. (See Gary A's blog. Again, the results confirmed our earlier decisions to elect Grant Johnson and Pete Hill.

This only confirms my fears. MLE were derived from what were generally exhibition games, which is a very shaky way method. I know there's nothing better, but I would never say that the results of a few exhibition games confirms a decision to elect a given player. Even the methods used for the Cuban players are a little questionable.
   234. Howie Menckel Posted: June 28, 2006 at 02:24 AM (#2078886)
It worked out that "pennant is a pennant is a pennant" won out, however unintentionally.
My listings have shown that the number of ALers or NLers has been remarkably consistent over time (generally slightly rising, mostly AL having a couple more per year) - the spike in the late 1920s and 1930s is due to the rising quality of the Negro Leagues and the rising number of stars.
It turned out that, in effect, the group didn't pit the borderline Negro Leaguer against the borderline MLer - it elected both.

I agree that overall there were more stars in that era, aided by 3-league issues and being between the wars. So the spike is somewhat justified by any measure, and one could argue that while the Negro Leagues are at their best, there should be more HOMers total.

As for Negro League MLEs, I guess I'd be popular among our guests here. I've been skeptical but respectful of those numbers, and rated guys like Cool Papa Bell more highly than Willard Brown - contrary to the MLEs.
It's a tough process, but I like the diversity of opinions and the results. Nobody and no group is perfect, but it's working pretty well.
   235. Brent Posted: June 28, 2006 at 03:14 AM (#2079013)
This only confirms my fears. MLE were derived from what were generally exhibition games, which is a very shaky way method. I know there's nothing better, but I would never say that the results of a few exhibition games confirms a decision to elect a given player.

Be careful about bringing your 21st century prejudices to discussions of baseball history. These exhibition games were not played by a bunch of prospects that the MLB teams were taking to Havana for tryouts. (To begin with, the teams didn't have farm systems from which to bring prospects. Also, they weren't going to spend the money to take a bunch of prospects to Havana. Check out Gary A's list of opposing pitchers in games Jose Mendez pitched against major league teams--they included Mathewson, Plank, Bender--some of the best pitchers of the era.) These series were generally held in October or November; they were not part of spring training. At that time it was common for major league teams to play exhibition games or series against teams from other leagues, and from the evidence I've seen these series were hard fought. The major leaguers considered it quite embarrassing to lose to teams from "lesser" leagues.

Basically, the issue comes down to whether Negro Leaguers should be included in the HoM, or whether we're going to exclude them just as they were excluded from "organized" baseball during their careers. Personally, I would never have considered participating in an HoM that excluded them. Since Negro Leaguers are included, then we have a few options on how to compare their records to major leaguers. In principle, it's really no different than comparing records of major leaguers who played in the Federal League with the AL and NL. You can compare those who played in both leagues (Robinson, Campanella, Doby, Irvin, and others for the Negro Leagues; Luque, Gonzalez, Marsans, Estalella, Minoso and others for the Cuban League). Or you can look at their records in head-to-head competitions--there were hundreds of games played between major league teams and Negro League or integrated Latin American teams. Fortunately, the two approaches have largely given us the same answers.

None of us claim that the MLEs are perfect, and like Howie Menckel I mix in a heavy dose of evidence from expert opinion. For me, that's what makes this project interesting. I mean, how boring is it to plug numbers into a spreadsheet and have it spit out a rating. The thing that keeps it interesting is to ask difficult questions, to bring in new types of evidence, and to continually question your assumptions.
   236. sunnyday2 Posted: June 28, 2006 at 03:43 AM (#2079037)
>When comparing players across eras, everyone is making assumptions, yet almost no one is acknowledging those assumptions.

We've been at this for 3+ years. All assumptions have been hashed ad infinitum. The fact that they're not all rehashed in the 1979 ballot discussion to the contrary notwithstanding.

Navel gazing is good. Diz, you may be new to HoM navel-gazing, but we are not.
   237. sunnyday2 Posted: June 28, 2006 at 03:47 AM (#2079041)
PS. I dont' think the choices in #228 are exhaustive.

F. We should elect the best players available regardless of era, but using consistent methods and criteria.

This is not the same as "no effort to maintain consistency." F would include consideration of the size of the player pool (the population from which the players are drawn) and of the opportunity pool (the number of teams) among many other things. But consistency would not be the goal, it would be a guidepost.
   238. sunnyday2 Posted: June 28, 2006 at 03:49 AM (#2079044)
>This only confirms my fears.

Omigod. We have screwed up. All is lost. Effort wasted. Despair!

Is there a point to this other than: Despair!
   239. Mark Shirk (jsch) Posted: June 28, 2006 at 08:13 AM (#2079153)
Sunny is getting cranky again
   240. Chris Cobb Posted: June 28, 2006 at 02:10 PM (#2079218)
This only confirms my fears. MLE were derived from what were generally exhibition games, which is a very shaky way method. I know there's nothing better, but I would never say that the results of a few exhibition games confirms a decision to elect a given player.

No, no, no, no.

MLEs were derived from:

1) the batting records of players who played multiple seasons in both the Negro Leagues and the major leagues in the 1940s and 1950s. These records were adjusted for aging patterns and for the contexts of league offense and park, insofar as that was possible. This provided the basic conversion rates that were used for play in the organized Negro Leagues, 1920-1948.

2) the batting records of players who played in the major leagues and the Cuban Winter League. These were used to find conversion factors for the CWL during the 1920s, when there was most cross-over between the two, and also during the early 1910s, when several other players also crossed over. These factors set a competition level for the Cuban leagues that, for the 1920s, turned out to be quite close to the conversion rate that was suggested by comparing the batting records of black players who played in both the Negro League and the CWL, using the standard conversion rates established by method 1.

3) the records of games in Cuba between major-league teams and Cuban/American black teams. These conversion rates can be compared to the conversion rates derived from the 1910s by method 2, and again the findings of both conversion methods are reasonably consistent.

So, yes, some "exhibition games" were used in deriving conversion rates. But "exhibition games" are not the sole evidence for conversion rates for any period of CWL or NeL play during any period: there is always also a view of rates derived from ordinary, "official" competition.

The "results of a few exhibition games" never "confirm the decision to elect a given player." In the cases of Grant Johnson and Pete Hill, their records in several seasons of play in Cuba, amounting to several hundred at bats, provide a body of evidence. Conversion rates derived by both method 2 and method 3 indicate that both players, both past their peaks, were performing at a level that was significantly above average for the major leagues at that time. This evidence, which is entirely consistent with (1) the more fragmentary statistical record of their play in North America and (2) their contemporary reputation as great ballplayers, may be said to confirm our decision to elect them, which was based on a more fragmentary statistical record and their contemporary reputation. "Confirm" is not too strong a word, because the case for both players (or at the very least for Grant Johnson) is past the point where a fully informed person could raise any other objection than "much of his career is without statistical documentation" to his election. There is no counter-evidence that urges caution. Was Grant Johnson better or worse than Sam Crawford or Bobby Wallace? Impossible to be sure: make your best estimate in ranking them together on a ballot. Was Grant Johnson better than Joe Tinker and Johnny Evers? Almost without a doubt. Should he be a Homer? Undoubtedly.

Obviously, the record is not as complete, nor as easily interpreted, as the record for major-league players. The fundamental question for the project is: "Is it preferable to accept this higher degree of uncertainty and rank the players excluded from the major leagues because of racism together with major-league players, or is it preferable to aim for a lesser degree of statistical uncertainty by excluding the excluded players from the study?" For the purposes of the project, which is to honor the best players, it is clearly preferable to include all the players who might have been the best players within the pool of players under consideration. If the purpose of the project were to validate our analytical methods (as James's in the NBJHBA was in part, since his listing of players amounted to a defense of his win shares system), we might have preferred the other approach. But the HoM is just about the players. It's been our task to try to find the best methods to evaluate them that we can, given the data we have to work with.

Certainty is impossible, and that's a big reason why we vote, as Sean Gilman said a while ago. Have are we in the process of creating a perfect Hall of Merit? Most likely not. Better data and better analytical tools will be developed, and those may show some of our choices to have been mistaken, or some may feel we have assessed merit in the wrong way, or made some mistakes given the data we had available.

Joe Dimino's vision for the HoM was "Something better," not "something perfect." And I think it's hard to argue that we are not doing better; there's a good argument to be made that we are doing this project of identifying merit across the whole of professional baseball history better than it has been done before. That's not a small claim, but it's the only claim that really matters.
   241. Dr. Chaleeko Posted: June 28, 2006 at 02:27 PM (#2079231)
Sunny is getting cranky again

I love it when he's cranky...as long as he's cranky at someone else!

I agree that overall there were more stars in that era, aided by 3-league issues and being between the wars. So the spike is somewhat justified by any measure, and one could argue that while the Negro Leagues are at their best, there should be more HOMers total.

A few possible reasons why the 1930s are swollen with homers:
1) It was easier to dominate the 1930s---not likely

2) The second big war caused careers that would have centered on the forties to be centered on the thirties---possible in some cases

3) The Negro Leagues created a third/fourth league worth of players, whose careers centered on the 1930s---very much contributing

4) The contraction of the NgLs after integration, but before expansion causes baseball talent pool of the 1940s and 1950s to be smaller by comparison, thus fewer HOMers---pretty logical

5) The rise of alternative sports in the late 1940s and 1950s siphons off high-level talent from those eras, making them seem less well represented---I've got a feeling this one contributes too, but not as much as it will in the 1960s.

6) The rise of the farm system in the 1930s initially made it easier for some teams to identify and cultivate superstars, but in the 1940s and 1950s the level of play became more uniform after the (near) universal adoption of the farm system, making it more difficult to stand out---who knows
   242. John (You Can Call Me Grandma) Murphy Posted: June 28, 2006 at 03:17 PM (#2079279)
Joe Dimino's vision for the HoM was "Something better," not "something perfect." And I think it's hard to argue that we are not doing better; there's a good argument to be made that we are doing this project of identifying merit across the whole of professional baseball history better than it has been done before. That's not a small claim, but it's the only claim that really matters.

Very nicely put, Chris.
   243. Mark Shirk (jsch) Posted: June 28, 2006 at 03:22 PM (#2079281)
Of course there is also

7) The 1930's just happened to have a few more great players.

and

8) all of the above (except I don't really accept 1 when compared to surrounding decades)
   244. Dizzypaco Posted: June 28, 2006 at 04:31 PM (#2079334)
1) the batting records of players who played multiple seasons in both the Negro Leagues and the major leagues in the 1940s and 1950s. These records were adjusted for aging patterns and for the contexts of league offense and park, insofar as that was possible. This provided the basic conversion rates that were used for play in the organized Negro Leagues, 1920-1948.

If this is true, then I would have a high degree of confidence of the conversion rates in the 1940's, and a low degree of confidence in the conversion rates in the 1920's, in that there weren't any players who played in the Negro Leagues in the 1920's and the major leagues after 1947. If you are using aging patterns, you are making huge assumptions which aren't necessarily justified.

2) the batting records of players who played in the major leagues and the Cuban Winter League. These were used to find conversion factors for the CWL during the 1920s, when there was most cross-over between the two, and also during the early 1910s, when several other players also crossed over. These factors set a competition level for the Cuban leagues that, for the 1920s, turned out to be quite close to the conversion rate that was suggested by comparing the batting records of black players who played in both the Negro League and the CWL, using the standard conversion rates established by method 1.

The Cuban winter leagues, or what might have been also described as exhibition games for major league players. Or can you somehow prove that the major leaguers playing winter ball gave exactly the same effort as they did during the regular season? Once again, massive, possibly unjustified assumptions are being made.

3) the records of games in Cuba between major-league teams and Cuban/American black teams. These conversion rates can be compared to the conversion rates derived from the 1910s by method 2, and again the findings of both conversion methods are reasonably consistent.

More exhibition games. I would never make conversion rates from the results of barnstorming tours, or, for example, the current winter leagues.

I'm not saying that conversion rates have to be perfect. I'm saying that we need a reasonable amount of confidence in the numbers for them to be useful in these sorts of things. Conversion rates for the Negro Leagues in the late 30's and 40's passes this test, in my opinion, even if they are not perfect. The others do not. We have no idea if these numbers are any good or not.

If the purpose of the project were to validate our analytical methods (as James's in the NBJHBA was in part, since his listing of players amounted to a defense of his win shares system), we might have preferred the other approach. But the HoM is just about the players. It's been our task to try to find the best methods to evaluate them that we can, given the data we have to work with.

Do you really believe this garbage about Bill James? What in the world could you point to as evidence? Bill James devised a system call win shares. He then went about rating players based largely based on his new system. What in the world is wrong with that? How in the world is that different from what HOM voters have done? Did you even read the book?

Generally, what I think you have done, is this: Devise a system that can't be tested in any reasonable way, and assume it works. If you think I'm wrong, point to the games played between these players and Major Leaguers, in which the results actually mattered to the Major Leaguers.
   245. John (You Can Call Me Grandma) Murphy Posted: June 28, 2006 at 04:35 PM (#2079340)
Dizzypaco, which NeLers do you feel don't belong in the HoM?
   246. Chris Fluit Posted: June 28, 2006 at 05:12 PM (#2079382)
No, dizzy. What we've done is devise a system using the best available means and criteria. Then we've proceeded to use that system to the best of our abilities (including healthy skepticism from certain voters) rather than perpetuate the errors of history against players of color. The system may well be flawed and imperfect but it is immeasurably superior to the alternative of exclusion.
   247. Dr. Chaleeko Posted: June 28, 2006 at 05:14 PM (#2079386)
there weren't any players who played in the Negro Leagues in the 1920's and the major leagues after 1947.

i shouldn't do this, but Satchel Paige did...of course he's the exception to every rule.
   248. Dizzypaco Posted: June 28, 2006 at 05:26 PM (#2079409)
Dizzypaco, which NeLers do you feel don't belong in the HoM?

That's my point. I don't have any idea. I accept the MLEs for 30's and 40's Negro Leaguers - I think that the methods are pretty sound for the conversion, although I haven't studied the results as much as others. Its the Cuban players, the older Negro Leaguers that I don't think people know enough about. I'm not saying that the guys elected in don't belong - just that we don't really know. And its why I think its a mistake to try to put these players on the same type of ballot as people who we know so much more about.

I understand that its the conversion rates are the best method possible - I'm not disputing this fact. But the fact that it is the best available method does not necessarily mean that we should have confidence in the results. It was a worthwhile exercise, but I think that the lack of confidence in the results should have been more explicitly acknowledged, and taken into account.

What I believe should have been done is this: Identify what people feel is the proper number of Cuban and Negro Leaguers that should be elected to the HOM. Then attempt to determine who was the best players in each league - who were the best Cuban players? Who were the best NgL playing in the 1920's? And so on. Have an election. Perhaps rely on those people who have really studied the history of each of these leagues. Use statistics to answer these questions. But don't try to compare a Cuban player playing in the 1920's with a major leaguer playing in the 1960's - we have no real basis for comparison, IMO.
   249. Dr. Chaleeko Posted: June 28, 2006 at 05:48 PM (#2079444)
That's my point. I don't have any idea.

If we always got stuck on the uncertainty of the enterprise, there would be no quelling the fear of being wrong. Then we'd always be waiting for some better information to come along, and we'd never be able to advance past the 1910 ballot.

Some wise folks suggest that we should embrace what we fear as a way to cope with it. If that's so, the HOM has a bear hug with uncertainty.
   250. sunnyday2 Posted: June 28, 2006 at 06:04 PM (#2079472)
Jeez, now everybody's cranky.

I just objected to the glass is half empty, Monday morning critique.

I didn't do the MLEs so whatever faults they have are Chris' and Doc's...no, seriously, Chris and Doc's MLEs represent one of the best efforts I have ever seen to try to understand NeL records in some kind of apples to apples fashion. Sure, it was a case of fools rush in where Bill James fears to tread. So what? It needed to be done. Someday somebody might do it better...or they might not. They (Chris and Doc and those who helped them along, you know who you are) should be congratulated...and the HoM project too.... I mean, reasonable criticism is fine, we all do it (just some of us are more cranky than others). But what I saw did not strike me as constructive in the least.

And besides it was midnight and I was just home from a ballgame. So, yeah, I was cranky. (Thank god the locals won!)

Sorry, my cranky comments were not constructive either, of course. To summarize:

>It was a worthwhile exercise, but I think that the lack of confidence in the results should have been more explicitly acknowledged, and taken into account.

1) Well, now we learn that this has been a worthwhile exercise. That is gratifying. It certainly seemed that the opposite was being stated previously, and

2) I think the lack of confidence or the potential variance or whatever you want to call it has always been absolutely crystal clear and explicit. I know because I was here. Otherwise we would have elected several more NeLers, I mean, look at the MLEs. The 50+ voters have all had their takes on those MLEs and many of those NeLers got elected with 25-30-35 votes, just like the ML backloggers do.

This is not a stupid or superficial bunch of HoMies here.
   251. Jim Sp Posted: June 28, 2006 at 06:14 PM (#2079499)
Dizzy,
What I believe should have been done is this

If I might paraphrase...you don't buy the premise of our elections (trying to compare all of these players on a single ballot) but you find the related conversation interesting. That makes sense, you're welcome to participate in the discussion without casting a ballot. We're 80 "years" into this project so you'll understand that we're really not that open to doing a different project or changing the rules on this one. On the whole I'd say that our results have been pretty reasonable and you'd probably enjoy voting if you can learn to live with the premise. FWIW we are not including Japanese players in these elections, that could have been considered a similar issue, but the group picked a scope for the project and we're running with that.

Its the Cuban players, the older Negro Leaguers that I don't think people know enough about.

There are some people in this group (not me) that know a lot about these groups of players. The rest of us are obviously highly influenced by their input.
   252. Dizzypaco Posted: June 28, 2006 at 06:19 PM (#2079509)
If we always got stuck on the uncertainty of the enterprise, there would be no quelling the fear of being wrong. Then we'd always be waiting for some better information to come along, and we'd never be able to advance past the 1910 ballot.

Its all about the degree of uncertainty. We don't know for sure how good a certain player is defensively, but based on various sources, we have a fairly good idea. We don't know for sure how a player playing in the 1890's compares with someone playing in the 1950's, but with some reasonable assumptions, and a decision to treat people equally across eras, we can make educated choices.

The same for MLE's. If we have reasonable confidence in the numbers, as I feel is justified with MLE's for NgL players from the 40's, we can made educated choices.

If there is no reasonable confidence in the numbers, if they are closer to a stab in the dark, then to embrace these numbers is not always the wisest choice. And that is the fear with some of these players.
   253. karlmagnus Posted: June 28, 2006 at 07:13 PM (#2079618)
We also elected Joe Start, Dickey Pearce and Lip Pike, much/most of whose value came before 1871, playing what was essentially an entirely different game. At least the Negro Leaguers were playing approximatly the sasme game as their ML contemporaries, even in Cuba.
   254. Chris Fluit Posted: June 28, 2006 at 07:31 PM (#2079670)
Dizzy, I suspect that lack of confidence is a major reason why Jose Mendez and Dick Redding went into the backlog while several Negro League pitchers of the '30s were elected in their first year of eligibility. That lack of certainty has had an effect on the ballot. But sooner or later, we have to make some sort of a decision on those players. For most voters, we're operating within the parameters of reasonable certainty. We obviously can't operate under absolute certainty- we can't go back in time and change the system that prevented these players from getting a fair shake in the first place, and we likewise can't go back in time and document the actual performances to your satisfaction. We can only look at the best available evidence and draw our conclusions from that. We may not be 100% certain that Mendez or Redding were better than Billy Pierce or Bucky Walters (to name two pitchers behind them on the ballot). But we also can't be 100% certain that Mendez or Redding were worse than Don Drysdale or Jim Bunning (to name two pitchers who have been elected recently). We could be underestimating their ability as easily as overestimating it. All we can do is to try and be as fair as possible. And that means giving due consideration to these players, even without absolute certainty. Some of us are convinced and we accord them high spots on our ballots. Others of us are unconvinced and leave them off the ballot entirely. It is the combination of opinions leads to a fair result. And that result is certainly more fair than excluding all black players who played before the 1930s.
   255. John (You Can Call Me Grandma) Murphy Posted: June 28, 2006 at 07:36 PM (#2079682)
Voting in a HOF election is not for the weak of heart. :-)
   256. jimd Posted: June 28, 2006 at 08:16 PM (#2079779)
If we always got stuck on the uncertainty of the enterprise, there would be no quelling the fear of being wrong. Then we'd always be waiting for some better information to come along, and we'd never be able to advance past the 1910 ballot.

We'd have never started.

Before the NeL debates there were all the debates about the values of 19th Century pitching, catching, fielding, league quality, pre-National Association, rule changes, etc.

Voting in a HOF election is not for the weak of heart. :-)

And it's a piece-of-cake compared to a HOM election ;-)
   257. John (You Can Call Me Grandma) Murphy Posted: June 28, 2006 at 08:29 PM (#2079814)
And it's a piece-of-cake compared to a HOM election ;-)

Jim, this week I think I have made more typos and boneheaded mistakes than I have had in the whole year. :-(
   258. KJOK Posted: June 28, 2006 at 09:02 PM (#2079907)
there weren't any players who played in the Negro Leagues in the 1920's and the major leagues after 1947.

Bobby Estelella and Quincy Trouppe were close, plus you did have 1920's Negro League players like Dolph Luque who played in the majors also...
   259. Chris Cobb Posted: June 29, 2006 at 03:34 AM (#2080600)
Dizzypaco wrote:

The Cuban winter leagues, or what might have been also described as exhibition games for major league players. Or can you somehow prove that the major leaguers playing winter ball gave exactly the same effort as they did during the regular season? Once again, massive, possibly unjustified assumptions are being made.

This comment illustrates the main problem with all of your objections to the use of Cuban stats for MLEs. While admitting that you are ignorant about the Cuban players and their leagues, you nevertheless proceed to pass judgment on the value of the statistical evidence that documents their careers. Here's the situation: the major-league players playing "winter ball" in Cuba in the 1920s were NATIVE CUBANS. This was their home, and their home league. It's where they played before, during, and after they played in the U.S. major leagues, which a few of them were lucky enough to be white enough to get into. The idea that they put less effort into their play in their home country, or treated the games as exhibition games, is utter nonsense. It is quite probable that they cared _much more_ about the outcome of the CWL season than they did about the outcome of the major-league season.

More exhibition games. I would never make conversion rates from the results of barnstorming tours, or, for example, the current winter leagues.

Again, you show that you are simply applying assumptions based on conditions after Latin American ball became a de facto farm system for the North American major leagues to conditions prior to this change in the organization, which took place in the 1950s. You have no basis for legitimate criticisms of our use of this evidence because you have no understanding of the context in which the evidence was generated.

I wrote:

If the purpose of the project were to validate our analytical methods (as James's in the NBJHBA was in part, since his listing of players amounted to a defense of his win shares system), we might have preferred the other approach. But the HoM is just about the players. It's been our task to try to find the best methods to evaluate them that we can, given the data we have to work with.

Dizzypaco responded:

Do you really believe this garbage about Bill James? What in the world could you point to as evidence? Bill James devised a system call win shares. He then went about rating players based largely based on his new system. What in the world is wrong with that? How in the world is that different from what HOM voters have done? Did you even read the book?

I have read it cover to cover, and use it regularly as a reference during this project. I don't see, actually, why you think I was claiming that there was anything wrong with what James did: I meant no criticism whatsoever. James' top 100 rankings of players at each position is the single most comprehensive evaluation of baseball players ever undertaken. In its kind, it is more ambitious and more valuable than the Hall of Merit. That said, the NBJHBA is nevertheless, unavoidably (and probably quite designedly) a defense of the win shares system, in that James introduced the system to the baseball world in the book and then showed the results of applying it to the systematic ranking of players by position. If the reader thinks that his rankings are horse****, (which is what James thinks of some of the early _Total Baseball_ rankings), that would reflect poorly on his system. Broad acceptance for win shares depended to a significant extent on the plausibility of the rankings the system helped to generate, and James was surely well aware that he was giving his comprehensive metric the most high profile exposure possible by showcasing it prominently in the long-awaited _New Bill James Historical Baseball Abstract_. He also believed that it was a valuable metric and that his rankings were improved by his use of it, since he delayed the publication of the NBJHBA in order to use win shares in it, but I would venture the possibility that he thinks the win shares system is ultimately more important that the all-time rankings he used it to create, and that he therefore saw the rankings as serving the system as much as he saw the system serving the rankings.

As to additional evidence: you might notice, as others in this discussion have pointed out, that James includes Negro-League players in his top 100 comprehensive ranking, but he doesn't include them in his positional rankings. If he has enough confidence in his ability to judge those players to place them in the top 100, why not place them in the positional rankings also? You suggested, and I think the suggestion is somewhat plausible, that James didn't think there was enough evidence available to rank the borderline players, or at least that he personally was not sufficiently well-informed to do so. I would note, however, that he felt he knew enough to rank the top 10 players from the Negro Leagues at each position, and it would be the rare case when many more than those ten players would need to be placed within the top 100 at a position. I have no doubt that James could have made some well-educated guesses if he had decided that it was a good idea.

I would suggest that one major reason he decided that it was _not_ a good idea was because he wanted to use win shares as the primary basis for his rankings. He explicitly included a subjective element, but he wants the reader to be able to assess the lists purely on the basis of the win shares numbers. With no comparable numbers for the Negro-League players, his overall analysis would have had to be much more subjective in order to integrate them, and the value of the lists as a demonstration of the reliability of win shares as a comprehensive metric would have been diminished. Notice that James doesn't include any systematic win share numbers for his overall top 100, in which he includes the Negro-League players. He doesn't use that set of rankings as a demonstration of the value of win shares at all. It therefore appears to me that James was trying to find a reasonable compromise between the imperatives of giving the Negro-Leaguers their due and making as effective a demonstration of the value of the Win Shares system as possible. I think the compromise he made was a good one.

The Hall of Merit doesn't have a new comprehensive metric to demonstrate to the world through our project, so we don't face the same imperatives that James did. Therefore, we can do things a bit differently, and produce a perspective on the top baseballplayers of all time that has somewhat different strengths.

Generally, what I think you have done, is this: Devise a system that can't be tested in any reasonable way, and assume it works. If you think I'm wrong, point to the games played between these players and Major Leaguers, in which the results actually mattered to the Major Leaguers.

What is a reasonable test for win shares, or for WARP1? To a considerable degree, a major test both systems face is a non-systematic judgment of the "reasonableness" of their results, isn't it?

But let me address the test you propose. In a strict sense, we can never know how hard players are actually trying to win. Pedro Martinez got lit up by the Red Sox tonight: did he _really_ want to win? We can't know. In a broader sense, though, it's obvious that the results mattered to the "Major Leaguers." As I pointed out above, in all of the CWL games, the major leaguers were native Cubans, playing in their home league. It is highly unlikely that they played half-heartedly in these games. Dogging it would never have been tolerated by their fans or by their teammates. Although they were good players, they were not (except in the case of Dolf Luque) the top stars in the league. As to the exhibition games between major-league teams and Cubans, in some cases the major-leaguers may have taken the games lightly. But if you were to read accounts of these series, you would see that most of the time they took them quite seriously. The onus is on you, I think, to learn something about this evidence before you dismiss it, not on me to demonstrate that I and others in the project have not proceeded as cavalierly in our study of the data as you imagine that we have.

If there is no reasonable confidence in the numbers, if they are closer to a stab in the dark, then to embrace these numbers is not always the wisest choice. And that is the fear with some of these players.

Acceptance of these numbers is not a requirement for any HoM voter, and we haven't elected any players purely on the basis of MLE evidence -- that is, we haven't elected any player who looks great by the numbers but who doesn't also have a reputation as a great player. What is a requirement for the HoM voter is that all eligible players be given full and fair consideration. And that is, in the design of the HoM, the wisest choice.
   260. Brent Posted: June 29, 2006 at 04:02 AM (#2080621)
Thanks, Chris.
   261. Mark Shirk (jsch) Posted: June 29, 2006 at 04:54 AM (#2080667)
Wasn't this also during the period that some players may have needed their CWL paychecks, if not to survive then to at least live comfortably? It wasn't like this was the era of multi million dollar baseballers.
   262. John (You Can Call Me Grandma) Murphy Posted: June 29, 2006 at 12:35 PM (#2080767)
Thanks, Chris.

Double thanks. You said everything that I would have if I was as eloquent.
   263. Howie Menckel Posted: June 29, 2006 at 01:12 PM (#2080788)
dizzypaco,
Given your bedside manner, I pray you aren't a doctor in real life.

You have a germ of a reasonable concern - which gets buried under an avalanche of undoubtedly inaccurate/highly unlikely beliefs.
Feel free to take note of the responses, and incorporate them into a new try.

Gentle suggestion: Clearly, you can stick with your overall premise if you like, and it's not absurd on its face. But please try to make a more sensible defense of your stance.
   264. Howie Menckel Posted: June 29, 2006 at 01:15 PM (#2080791)
The part about how Cubans might not have been taking games in CUBA all that seriously reminds me of the questions to Toni Kukoc on whether he'd be able to handle the pressure of an NBA playoff game.
He used to laugh heartily, knowing that playing for his country had entailed a whole lot MORE pressure - and he'd note that in some European countries, "hostile crowds" mean fans are throwing things like coins, batteries, etc. on the court during play.
You may recall that Kukoc performed quite well in NBA playoff games, thank you....
   265. Dr. Chaleeko Posted: June 29, 2006 at 02:34 PM (#2080834)
The Ty in Chris Cobb comes out!

I think Chris's rebuttal points out one of the most difficult mental roadblocks for analyzing the pre-War period. And much of this is conjuecture, but bear with me. Before the farm system (or perhaps, simply, before the war) MLB games "mattered" to people in about half the states in the union. There were other leagues that mattered much more to their local constituencies, because those attendees could see the players in the stands or maybe even knew them, and because those leagues were not player-development factories and chased real, meaningful pennants. I'm talking mostly about the NgL, PCL, MxL, and CWL, but there are many others.

Today we hardly understand this. Today MLB is, worldwide, the highest peak you can reach as a player. There are other mountains, notably Japan, but none is high enough to rival the majors, and/or have been made contractually subservient to MLB. My impression is that back then, the NgL, the PCL, the MxL. the CWL, and the coming Japanese leagues, all began to climb to a point where they were nearing the height of MLB and therefore either presented a threat or an opportunity to it. We know this is true because history tells us so.

1) The MxL raided MLB for players, and MLB responded quickly and decisively. Eradicating the threat. Within ten years the MxL was made into a classified league.

2) MLB relocated and expanded in response to the PCL's threat to go major, which ultimately caused the PCL to enter a farm agreement with the big leagues rather than risk further encroachment on its territory.

3) MLB saw opportunities for cheap talent in the NgLs and integrated.

4) The CWL and PRWL were all subsumed under the player-development mantra, making the annual Carribean Series less a nationalist baseball fetschrift.

What's yet to happen with Japan, I just can't say.

Anyway, the mindset that MLB was so much better or more important is, I think, a product of the post-War boom, the reigning in of minor and independent leagues in the 1930s-1950s, and the big money that came along in the 1970s. And it's now deeply ingrained. When the colonial officials of Mass Bay Colony forced the Penobscot Indians to abandon all of their traditional tools for hunting, gathering, and processing foods, the institutional knowledge of milling grains or dressing the catch vanished within three generations. It's difficult to relearn what you have unlearned. And this, I think, is one of those cases where anyone raised with the supremacy of MLB (like most of us!) has to unlearn.
   266. sunnyday2 Posted: June 29, 2006 at 03:03 PM (#2080862)
Doc, you address your comments to the pre-war period. I think they continue to hold for black players through the integration period (through 1960 or so). We do not really understand the meaning of their accomplishments in the NeLs or the MLs or especially the MiLs in what was a period of transition at every level. Here is what I wrote in another thread:

If there is an era that is underrepresented, I believe it is the integration era for black players--players who, like Minoso, split their careers between the NeL and the "white" M or MiLs. Lots of guys (Wilson, Clarkson, ???) moved into the "white" M or MiLs but didn't get a fair opportunity to play in the MLs. We have elected (I am going off the top of my head here) 10-12-15 black players from 1935-45 and a similar number from 1955-65, but only about 4-5 from 1945-55. (Somebody posted a list a couple months ago, anybody help me here?)
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
dirk
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.0950 seconds
49 querie(s) executed