ba-ba-bump

Definitely in a strange state.

I tried bumping it around the corner, but all attempts to follow links to go to page 4 take you to page 1 (posts 1-100).

However, if you edit the browser address line to go to /p300/ it takes you to the right place (though there's nothing interesting there yet).

Now the link to page 4 works, but my previous post is not yet visible (but 301 'ba-ba-bump' is).

Posting the last one (303, #2623031) made the previous one visible also (302, #2623027).

All of this stuff might be useful should someone choose to debug these minor problems.

X-posted from the Hall of Merit Needs You:

91. Joe Dimino Posted: December 01, 2007 at 01:36 AM (#2630505)
Dan, this is great news, thanks so much. Glad to see the great work everyone here has done will get some recognition. That's really exciting!

Looking forward, I really think we should focus on ranking Hall of Merit not HoF. That's where we'll really be able to help the most
if anyone actually pays attention to our work. Especially the Vets Committee, maybe we can help them actually elect someone.

Has there been discussion of how the ranking project could be/would be/should be structured? I guess there was talk of kind of starting with frontlog, mid-log, backlog and going from there?

Good work, Dan.

<a href=http://www.nytimes.com/2007/12/02/sports/baseball/02score.html?_r=1&oref=slogin> NY Times Article </a>

http://www.nytimes.com/2007/12/02/sports/baseball/02score.html?_r=1&oref=slogin

Has there been discussion of how the ranking project could be/would be/should be structured? I guess there was talk of kind of starting with frontlog, mid-log, backlog and going from there?

In post #240 in this thread I presented a proposal that would accomplish the task of prioritizing the HOM-not-HOF players. In #261 I gave this rationale for going beyond simply ranking them (they must be placed in an integrated list of all HoMers):

Since we agree this would be a worthy aim to devote our time to, it's all the more reason to go in the direction of something like "The Ten Level Spectrum of HoMers". Why? Because merely producing a ranked list of non-HOFers lacks an important aspect: the context of who are their peers already in the Hall.

If we simply say "Deacon White is the best non-HOFer", what does that mean? That he's as good as Willie Mays? or George Kelly? It's a mute determination.

But, if we can say, "Deacon White is the best non-HOFer. We rate him a low 6, on the same level as Willie McCovey, Bill Dickey and Harmon Killebrew", this speaks loudly about the quality of player we judge him to be.

Again, think about that proposal of mine. It's pretty simple to do, and I think it offers significant benefits.

Oops. The proposal for putting HoMers into ten levels is in post #242, not #240.

Thought I'd bump this back to the top. Has anyone had more discussion about a book? I have a fear that this exercise will all be lost unless it is put down on actual paper at some point.

Are andrew siegel and ronw going to get the MMP Project off the ground?

From the Group 4 election results thread:

"I thought it was likely that we would now have everybody pick a top 15 out of the 57, as opposed to the usual yearly format of top 15 out of 100 or so guys with votes - plus you could pick a 'wild card' or two beyond that if you wanted of course.

No need to prevent someone from taking a 28th- or 38th-ranked guy in his top 15, I don't believe.

It's just that the previous results can make the process more manageable for those ongoing voters who couldn't handle the size of a group or its antiquity.
By all means look at all 57 closely, if possible.
If not, at least be sure to fully consider the top 15 or so, and as many others as you can.

In some ways, this should be easier than a typical 'year' ballot, granting that it's tough when so many good ones are on at once. But it's a smaller universe of 57 overall."

.................

That thread also lists the "pct of possible votes" for all 57 candidates in the 4 different elections, which is a crude but potentially helpful measurement of general sentiment on merit.

I'm happy to take the lead on another thread of the project, with three caveats:

(1) We should do the composite balllot first.
(2) We need to decide if we are doing positional rankings or MMP awards.
(3) We need to have commitments from enough people (30?) to make the project worhtwhile.

(1) We should do the composite ballot first.

I still think it's largely an exercise in omphaloskepsis, very limited in usefulness for hall of fame arguements.

(2) We need to decide if we are doing positional rankings or MMP awards.

Again, I see limited utility in a positional ranking exercise.

(3) We need to have commitments from enough people (30?) to make the project worhtwhile.

Well, the first HoM election had 29 voters. I think 25 committed voters is enough to start.

No way.
Omphaloskepsis is "HoM Election History, Data and Analysis" by me and others, or the HOM Encyclopedia by Patrick W., or the metrical analysis of consensus and plurality by OCF.

By the way I googled "omphaloknepsis" and found only my own previous use of that non-word here at the Hall of Merit, and proper noun that I didn't investigate. And it's awesome that Google knows so much, and has organized its thoughts so well, to report such things virtually in an instant.

--
I agree that a score of participants is enough, or even fewer if it can attract some new blood from visitors to BBTF who did not participate in the Hall of Merit. That is a matter of pitching it in a way that is not too clubby or cliquish or [vocabulary fails me].

It's a real word (there are alternate spellings):

omphaloskepsis

noun
literally, the contemplation of one's navel, which is an idiom usually meaning complacent self-absorption

In the present discussion, I use it to describe efforts that are in the direction of fascination for long-time HoMers, while having no aim (and thus, little likelihood) to affect anything outside that micro universe.

I'm not greatly interested in the MMP project (though I will try to participate if it goes forward), but I agree that, if it starts in the present and goes backward in time, it might well encourage new participation in the HoM.

I see positional rankings as a valuable project for the HoM's presentation of its honorees, but it will require voters to figure out how to assess early players in relation to later players, and to assess players who starred in Negro and Latin leagues in relation to players who had the opportunity to play in the majors. That is a significant bar to entry for those who haven't already worked through those fundamental and difficult issues.

Are the two projects necessarily mutually exclusive?

Are the two projects necessarily mutually exclusive?

No; I think we're open to any option.

If the plan is to do positional rankings, a good way to approach it is by era, similar to the four-part ranking project just completed. You could use those eras. Or you could use an equal-era scheme. Maybe something like this:

Five 26-year eras, players' careers centered in the era

1867-1892
1893-1918
1919-1944
1945-1970
1971-1996

Speaking of eras, what of the Bill James (win shares) handling of 2B and 3B in three eras? Or catchers in two eras, pitchers in two eras, iirc.

For catchers, he adjusts the formula. There may be some evidence for assessing the adjustment in the pattern of catchers by time and letter grade (which he presents in one all-time list). I haven't looked at it.

For 2B and 3B, he presents the leaders in three lists for three eras: "early", "depression era", and "post-war". Twenty depression-era 3Bmen qualify for listing (Win Shares p144). Their letter grades:
A+ 0
Ao 0
A- Werber
B+ 3
B: 5
B- 3
C+ 4
C: 1
C- 1
D+ 1
D: 1

He gives 5 early A+ (Cross, Collins, Leach, Devlin, Tannehill) and one late A+ (Clete Boyer)

Sixteen depression-era 2Bmen qualify for listing
A+ Lonnie Frey, Burgess Whitehead
A: Bobby Doerr, Joe Gordon, Skip Melillo
A- 0
B+ 2
B: 2
B- 1
C+ 2
C: 3
C- 1
none below C-

He gives six early A+ (Bishop, Critz, Frisch, Maguire, McPhee, Danny Richardson [A to Hardy]) and five late A+ (Hubbard, Lemke, Maz, Reed, Robinson).

I haven't looked closely enough to support anything but a vague warning to users of the letter grades.

--
The general issue about ratings or rankings by fielding position is whether one should try to reflect changes in the game in a more complex way than equal time intervals.

I would participate in either the MMP (Most Meritorious Player, right?) or the positional rankings.

In terms of doing an overall ranking among the four groups of HOM/non-HOF, I think it would be wise to eliminate Jackson and Rose from the list. They are actually banned so they cannot be considered.

The general issue about ratings or rankings by fielding position is whether one should try to reflect changes in the game in a more complex way than equal time intervals.

KISS principle

which is, to me, secondary to the KILL principle: Keep It Logical, Lunkhead

I expect that any attempt to capture a change such as the place of secondbase on a defensive spectrum is futile.

On the other hand, the "simple" five 26-year periods seem to be cooked to match the 1892/1893 change in pitching distance
Five 26-year eras, players' careers centered in the era

1867-1892
1893-1918
1919-1944
1945-1970
1971-1996

It can also be simple to follow established usage, or popular usage.
I would call a joint at 1892/1893 established usage;
1919/1920 and 1942/1943 or 1945/1946 merely popular usage;
1918/1919 and 1944/1945, novel usage and therefore not simple.

Does anyone know whether the Hall of Fames's current use of 1943 is intended to represent a distinction between pre-wartime baseball and wartime baseball?
Or is it some particular number of years, like 23 = 5 + 15 + 3 for the old veterans committee?

It's a compromise scheme between the "established usage" scheme and an equal-era scheme. They actually match pretty closely.

If one is truly being fair to all eras, there should be some discernible relationship among the number of HoMers from each era.

Approximate number of HOMers in each of the proposed intervals:
1867-1892 30
1893-1918 50
1919-1944 60
1945-1970 50
1971-1996 40

The exact numbers will depend on the disposition of some of the boundary cases.

The high number for Era3 reflects the oft-noted "over-election" from the 1930's.
The low number for Era5 reflects the long careers of good candidates still ineligible.
The low number for Era1 reflects what?

Well, I think it would be useful for people to know if we think Grich was as good as Rose, or Santo or Dahlen as good as Shoeless Joe, etc.

I think it would be useful for people to know if we think Grich was as good as Rose, or Santo or Dahlen as good as Shoeless Joe

And even more useful for people to know if we think Grich was as good as Jim Rice and others already in the HOF.

Does anyone know whether the Hall of Fames's current use of 1943 is intended to represent a distinction between pre-wartime baseball and wartime baseball?
Or is it some particular number of years, like 23 = 5 + 15 + 3 for the old veterans committee?

In the Table of Contents, Veterans Committee rules, they are called "post-World War II players" and "pre-World War II players". Service in the Negro Leagues through 1945 counts toward ten years in the major leagues.

"And even more useful for people to know if we think Grich was as good as Jim Rice and others already in the HOF."

We already told them that with our votes.
I'm talking about the next step.

We already told them that with our votes.

We did? Hardly. Our votes told very litle about where we think he ranks among the hall of famers, or among the HoMers. We never compared him to Rice "with our votes" either. So, I'm talking about a relevant next step.

So am I.
Let's let the other voters decide what's relevant.
Deal?

(This is no longer relevant although it may suggest something relevant.)

For 2003, 2005, and 2007, but now revised out of business, the first stage of selection for Veterans Committee consideration was nomination of 200 players by a Historical Overview Committee. (Now there is a Historical Overview Committee for pre-WWII players, for managers/umpires, and for executives/pioneers, but it nominates much smaller lists. The largest lists that will be published, under the current system, correspond in size to the final ballot under the previous system.)

2005 and 2007 Revisions to Hall of Fame nomination by the Historical Overview Committee
- 200 players, revisions 2003-2005 and 2005-2007
- 60 contributors, revisions 2003-2007

@ post-WWII candidates under current rules [compare HOM Group 2]
* pre-WWII candidates under current rules, who remain under the first-stage jurisdiction of a Historical Overview Committee [compare HOM Group 3]



For 2007 the earliest 20 of the 200 nominees played in the 19th century. Of course none advanced to the second stage, but note how many HOM members did not make the first cut. (Herman Long was one of the 200 for 2003 and 2005, the only change in the 19c subset.)

Don't rush to enter the raw data yourself!
I plan to include it in a big table that I post at yousendit and to baseball-databank (which may get it into the next lahman database or on baseball-reference).

Let's let the other voters decide what's relevant.

Oh, I see. Well, I tried just observing for awhile and the result I see is this HoM thing waning in peoples' interest.

Tsk.

Frivolity time:

Imagine watching 3 full-time Hall of Merit hurlers on the same team in the same major league season.
HOM MLB Pitching trios (each must pitch at least 1 IP per team game, or 35 games, or 20 SV):

1949 Cleveland AL - Bob Lemon (22-10) and Bob Feller (15-12) and Early Wynn (11-7)
1950 Cleveland AL - Bob Lemon (23-11) and Early Wynn (18-8) and Bob Feller (16-11)
1951 Cleveland AL - Bob Feller (22-8) and Early Wynn (20-13) and Bob Lemon (17-14)
1952 Cleveland AL - Bob Lemon (22-11) and Early Wynn (23-12) and Bob Feller (9-13)
1953 Cleveland AL - Bob Lemon (23-15) and Early Wynn (17-12) and Bob Feller (10-7)

1966 Los Angeles NL - Sandy Koufax (27-9) and Don Drysdale (13-16) and Don Sutton (12-12)

That's it so far.

But consider that someday Glavine/Smoltz/Maddux will be listed NINE times here (1993-99 and 2001-02).

Nope, we've never seen anything like that (in the majors).

Anyone want to take a stab at a Negro League team with 3 HOM full-time hurlers in the same season?

Trivial: Gaylord Perry threw 13 full-time seasons with an HOM moundmate (8 w/Marichal, 2 w/Fingers, and 1 each with Blyleven, Jenkins, and Niekro).

Most frequent HOMer full-time pitching tandems:
Wynn/Lemon..........8
Koufax/Drysdale.....8
Marichal/Perry......8
Mathewson/McGinnity.7
Feller/Lemon........7
Waddell/Plank.......6
Lyons/Faber.........6
Alexander/Rixey.....5
Wynn/Feller.........5
Gibson/Carlton......5

Glavine and Smoltz will crush the field with 13 (and counting) someday.
Glavine and Maddux are 10.
Smoltz and Maddux also are 10.
And of course the trio's nine full-time seasons together are more than any DUO ever had before.

OK guys, I've been away for awhile. I'll catch up tomorrow.

If nothing's changed, the plan is to start ranking HoMer's by position now, right?

Howie, great list.

I'll read further back and catch up tomorrow.

1949 Cleveland AL - Bob Lemon (22-10) and Bob Feller (15-12) and Early Wynn (11-7)

Considering that Satchel Paige was also on this team, that's a lot of opportunities for Cleveland fans to watch HoMeritorious pitching.

Anyone want to take a stab at a Negro League team with 3 HOM full-time hurlers in the same season?

There aren't any, as far as I can tell. With only eight HoM NeL pitchers (Rube Foster, Jose Mendez, Joe Williams, Joe Rogan, Bill Foster, Martin Dihigo, Satchel Paige, Ray Brown) spread over a half century of baseball, there's hardly any two-pitcher overlap on the same teams.

Rube Foster and Joe Williams pitched together on the Chicago American Giants for a while in 1912 or 1914, when Williams joined the team for its western barnstorming tour, but Williams was not with the team for the full year.
Jose Mendez and Joe Rogan pitched together for the Monarchs from 1920-25. This is by far the longest pairing, but Mendez was only a full-time starter in 1923.
Bill Foster and Joe Williams pitched together on the Homestead Grays in 1931.
Ray Brown and Joe Williams pitched together on the Homestead Grays in 1932. (Paige was with the Crawfords 1931-32, so there were three HoM NeL pitchers starting in Pittsburgh, but not all for the same team.)
Satchel Paige and Joe Rogan were teammates with the Monarchs in the later 1930s, but Rogan was no longer pitching, at least versus high-level competition, and Paige was not with the Monarchs full time.

There could have been other pairings in the Cuban Winter League, but I don't have the data to check that out.

Yeah, if I'd made the minimun 30 G instead of 35, Paige would make it a 1949 quartet.

The quartet started 89 games and hurled 738.3 IP - a majority of both the starts and the innings that year.

The Indians only used 10 pitchers, all of whom pitched at least 25 games.
Mike Garcia was 14-5 with a 2.36 ERA in 176 IP, and 38-yr-old Al Benton was 9-6 with 10 SV in 136 IP, with a 2.12 ERA.

Most of the innings were hurled by great pitchers, or guys who pitched great that year - or both.
And they came in 3rd place.

119 team ERA+ but a 97 team OPS+

Bob Lemon hit .269 with 7 HR and led the club in OPS+ at 134 !

Shouldn't we be ranking something?

Shouldn't we be ranking something?

Okay. I'll start first, Sean.

You're so short, ants can walk over you.

:-)

ZING!

:-D

There could have been other pairings in the Cuban Winter League, but I don't have the data to check that out.

I don't believe that there were any pairings of two or more HoM pitchers on the same team in Cuba, other than trivial cases.

The trivial cases: In 1915-16, Cyclone Joe Williams was pitching for San Francisco and Rube Foster, the team's manager and by then largely retired as active player, pitched one game, going 0-1. There's an quirky story behind that team. An all-Cuban version of the team returned to the Cuban League after a 12-year hiatus, and they clearly weren't able to compete, going 1-21. At that point, the Chicago American Giants were invited to go to Cuba to finish out the remainder of the season for them. They still had a losing record, but at least they were respectable, with Williams going 3-3, Frank Wickware 2-4, and Dismukes and Foster each at 0-1.

In 1925-26, the 38-year old Méndez pitched 6 games (31 innings) with a 1-1 record for Habana and the 19-year old Dihigo pitched in one game without a decision.

The final trivial case is that Torriente (who had a famously strong arm) was sometimes used as a backup pitcher. He and Méndez were teammates for many years, both with Almendares in Cuba and with the Cuban Stars in the United States, so there were several seasons when he pitched a game or two.

Seasons pitched in Cuba by HoM pitchers: Dihigo 19, Méndez 14, R Brown 5, R Foster 3, J Williams 3, B Foster 1, Rogan 1, Paige 1. And also early in their careers, Wilhelm 2, Bunning 2, and Wynn 1. (Wynn's first Cuban win was pitched against Dihigo in 1939-40.)

Also, Paige and Dihigo pitched in the same leagues in the Dominican Repubic in 1937 and in Mexico in 1938, but for opposing teams.

Actually, the Williams who pitched for the 1915/16 San Francisco/American Giants in the Cuban League was Andrew "Stringbeans" Williams, not Joe Williams, who was pitching & managing in the so-called Florida Hotel League at the time. Figueredo just gets it wrong in this case.

Another interesting thing about that Cuban League season (& something that isn't mentioned anywhere in English-language sources) is that it was divided into three parts, each with its own championship. The American Giants were brought in for the second series, then the Cuban players took over for San Francisco again for the third part.

Gary,

Thanks again for your expertise. This project owes an immense debt to you.

Shouldn't we be ranking something?

Yes, we should. I believe the next step is to start ranking all HoMers positionally, right?

Any other thoughts?

The general issue about ratings or rankings by fielding position is whether one should try to reflect changes in the game in a more complex way than equal time intervals.

I agree this could be an issue.

I can see going both ways on this, but I think in the end, it's best to rank by position with the understanding that fielding meant a little more at 3B in 1893 than 1993, for example.

Pitcher wise, I do think we should have 3 categories . . .

1) Pre-1893 starting pitchers
2) Post-1893 starting pitchers
3) Relief pitchers

However, we only have 4 relief pitchers (Wilhelm, Fingers, Gossage and Eckersley). I think that Eckersley is a reliever - he was not going into the Hall of Fame as a starter, despite his accomplishments there.

I could easily see combining categories 2/3, the problem is that then we get into the argument over how valuable relief pitching is relative to starting, and I'm not sure I want to head down that road with this part of the project.

Thoughts?

I can see going both ways on this, but I think in the end, it's best to rank by position with the understanding that fielding meant a little more at 3B in 1893 than 1993.

I don't see any argument against that, Joe, or else the third base rankings would be worthless, IMO.

However, I don't think the position of pitcher should be split into separate categories. Albert Spalding should be able to duke it out with Sandy Koufax and Hoyt Wilhelm, IMO.

On pitching groups: both a pre-1893 pitcher group and a relief pitcher group are too small to be worth ranking. There's only Spalding, Galvin, Radbourn, Keefe, Clarkson, and Caruthers in the pre-1893 group, unless one tosses Rusie in here. Is there much value in ranking these six guys?

If we must split the pitchers for procedural reasons, I'd say a period split that gets the 63 into approximately equal groups of around 20 would make for a manageable ballot, say 1871-1920, 1921-1960, 1961-present. That more-or-less breaks out to pre-lively ball, lively-ball pre-integration, and the integrated game.

We could also work out a good balloting system for selecting out the top 20 pitchers, then the next 20 pitchers, then the bottom 23 in three successive rounds of voting. I think that would give us the right kind of results, but I don't know enough voting theory to propose a method that I am confident would be fair.

I can see pitchers and maybe OFs, too, in 3 preliminary groups and then combined voting.

But I also pictured combined "HOM not HOF" voting, and that has yet to happen.

Just seems easier to segment a large group first, then vote on the best of the best. You weed out the ones who prove irrelevant later - obvious to some voters, but make it more manageable for many others...

Pitching was much different before they moved the mound back. I don't think Clarkson et al should be ranked with the others. I would put Rusie in the modern group.

There's a clear deliniation in the three groups, IMO - old distance, new distance, relievers.

James ranked relievers separately in the HBA - he had 30 righties, 10 lefties and 5 relievers. I don't really see the righty/lefty distinction, but I think he was correct with the starter/reliever distinction.

I don't see any distinction that took place in 1961 - other than to make easy groups. But the groups aren't easy. Where does Whitey Ford fall then, for example?

Old distance, new distance, reliever seems reasonable.

I could maybe see splitting out 1893-1920 as 'deadball era'. Also, if that makes it easier to rank.

Howie - I assumed OF would just be LF/CF/RF, right?

Well, I'd like to see real numbers.

I don't see too many guys, especially LFs/RFs, as being so distinct from each other on pct of games at certain OF spots.

But maybe I'm wrong.

Pitching was much different before they moved the mound back. I don't think Clarkson et al should be ranked with the others.

Pitching was certainly different, but I don't see the best hurlers as necessarily inferior to later generations placed in their proper context. Therefore, the pre-1893 shouldn't be segregated, IMO.

fyi

I am finshing up penning an article for SABR entitled "wins and losses for batters". Not all that stat-geeky; much more 'just for fun' look at hitters and how you can make up a very simple W-L mark for them to compare them with pitchers. I come up with conclusions such as Lefty Grove = Ted Williams and Ron Santo = Bert Blyleven.

Anyone who'd like a copy, email me at han60man@aol.com and I'll send you a draft.

We could hash out what to do with the pitchers while we are starting on position players. Could we set an order for the positional votes, and get started, while we continue to discuss the pitching vote? What position do we start with, and when?

I'll reiterate my view that the best arrangement would be one that would lead to a single rank order of all pitchers, just like what we will have for position players.

What if we voted for pitchers in three rounds? In the first round, voters pick their top 30 out of the 63 HoMers. The top 20 finishers in that election make up the first 20, and they are removed from further consideration. Then there is a second round. Voters pick their top 30 out of the remaining 43 pitchers. The top 20 finishers in that election make up the second 20, and they are removed from further consideration. Then there is a third round, in which voters put the remaining 23 in rank order, and the results of that election fill out the roster.

There might be some noise around the joints between the different ballots (i.e. do we really know, using this system, that we rate #20 higher than #21?), but my intuitive sense is that it would be a reasonably fair approach. Others with better grasp of voting theory can correct me on this point.

I don't see any distinction that took place in 1961 - other than to make easy groups. But the groups aren't easy. Where does Whitey Ford fall then, for example?

If we must divide by eras, I still think 1961 is a better divide within the modern game than any other that's around, but, if we want to go that route, we should talk through the options.

As to borderline cases. Whitey Ford and possibly Hoyt Wilhelm are the only two difficult cases with this divide, which, as era markers go, is not a lot of borderline cases. As to 1961 as marking a distinctive change, there's expansion, and there's the fact that integration was more or less complete after this point: no more quotas, and the first black HoMer pitcher to have a full career in the majors has just gotten established.

What other late markers could be used? Well, the "big strike zone" era begins just a couple of years after 1961. If one uses that, Ford and Wilhelm become obvious "pre" cases and Drysdale and Koufax become the questionable borderline cases. This seems like a practical alternative marker.

We may come to look on the establishment of the five-man rotation in the 1980s, or the offensive explosion of 1994 as major era markers, but they are too late to help us now.

WWII is a major marker, but it may have had less effect on the way the game was played than either integration or expansion. Because of its effect on pitching careers, there are fewer borderline cases (Feller and Newhouser?). But it's very close to 1920.

Integration would be another possible marker, and one for which I don't think there would be any borderline cases, but it's nearly as problematic as WWII in its closeness to 1920. Also, since there are no black HoM pitchers from the 1950s (likely a consequence of the contraction of opportunities for black players, especially pitchers, that was an ironic result of the breaking of the color barrier in the majors), 1947 isn't as meaningful a marker for HoM pitcher groups as it is for position players.

Looking my own ruminations over, I'd say that 1963 would be pretty much equivalent to 1961, and that 1947 has its good points as a divide.

But overall, I'd rather see us get to a rank order that is not broken out by periods.

Pitching was certainly different, but I don't see the best hurlers as necessarily inferior to later generations placed in their proper context. Therefore, the pre-1893 shouldn't be segregated, IMO.

Segregating does not in any way imply 'inferior'. It just implies vast difference, which was the case.

I think if we were going to divide by more than just 1871-1893/1893-present/relief Chris, I'd go with these groups (which are not necessarily even, but reflect major changes in the game).

Group 1 - 1871-1892
Group 2 - 1893-1920
Group 3 - 1921-1945
Group 4 - 1946-present
Group 5 - relief aces

1946 present could be split into 1946-1980 and 1981-present, to reflect the emergence of the bullpen and reduction of starter workload, but we don't have a ton of guys in post-1980 yet anyway, so that seems kind of pointless.

As for the positions, I assume we just go:

C, 1B, 2B, 3B, SS, LF, CF, RF

Afterwards, when we have our lists, if we want to put them all in order - we could do it in a "Survivor" type format . . . where we either go bottom to top, or top to bottom . . .

Top to bottom

We rank the top guy in each group. This gives us 5 pitchers, 8 position players on each ballot.

The top guy is the champ. Next week we add #2 from his position to the group and vote again. Wash, rinse and repeat throughout the project. Weekly ballots (maybe even semi-weekly? It's only slotting one new guy each time).

Bottom to top

We do the same thing only start with the lowest ranked guy at each position. The lowest rated player each week drops and the new player from his list is added for the next election. This method builds suspense and is probably preferable.

This would keep us occupied for a few years, and be relatively easy to administer.

I'm not sure how we'd go about adding new electees though.

One thing these ideas would require is a group that will for the most part commit to voting all the time.

Think we can get this hashed out by the weekend and start discussing the catchers next week?

I'll also send an email to the Hall of Merit group once we get this figured out, in case people have stopped checking the site with the lack of activity the last few weeks.

356. Chris Cobb Posted: April 15, 2008 at 08:44 AM (#2744472)
We could hash out what to do with the pitchers while we are starting on position players. Could we set an order for the positional votes, and get started, while we continue to discuss the pitching vote? What position do we start with, and when?

My advice is start with C, 1B, 2B, SS, 3B, presumably one at a time. Two things will happen at the same time. Interested parties will continue to express themselves here, re pitchers and outfielders --their values, what they know of voting theory, who knows what else. Over at C and other "easy" ones, everyone who participates will gain better understanding of some challenges involved for pitchers and outfielders, subclassified or not.

Excellent advice Paul. I will post a thread Sunday to start discussion of the catchers next week, unless anyone objects . . .

I don't see any distinction that took place in 1961 - other than to make easy groups. But the groups aren't easy. Where does Whitey Ford fall then, for example?

Right around 1960 was a crucial change in how pitchers were used. Before then, pitcher usage was marked by fluidity. Just about any starter could be used as a reliever and vice versa. The modern firm split hadn't yet sunk in. Fun fact: in the 1960 AL, every single pitcher with 20 starts had at least 1 relief apperance. Every damn one. Now, only about 25% do. But what happened in 1960 in teh AL happened literarlly dozens of times in the previous decades. It has never happened since then.

And starters weren't used in a regimented, orderly fashion that we now think of when we use the word rotation. Look at those Cleveland Indians staffs from teh early 1950s. Even when no one was injured (which was usually the case) their days rest was all over the place. Instead of going ABCDABCDABCD ad infinitum, they went ABCDCADCBEDBAF.

The best example I can think of the shift that occurred then came in 1960/1961. Ralph Houk replaced Casey Stengel as Yank manager. Whitey Ford, for the first time in his career, worked exclusively as a starter. He had 39 starts, far more than ever before. Instead of being reshuffled around as Casey had always done with him, Houk put him on a set days rest.

By 1963, the Dodgers and Giants both had two men start half the team's games -- in the previous 40 years that had only happened once in all of baseball (the Spahn/Sain Braves). Maybe I'm missing one or at most two other examples. But two teams did it in one year. They could do it because the 4-man rotation had replaced the every-man fluidity of before. Those two Dodgers in 1963 had no relief appearances. The pair of Giants had only 1.

When it comes to pitcher usage patterns, the early 1960s mark one of the most profound changes in the history of MLB.

There, that's one less page from my book for you all to read.

1.
Since institution of the 154-games schedule in 1904 there have been only a few seasons where any team played fewer than 151 games.

Minimum team games played (150 or fewer, 1904 to date)
150 - 1926
149 - 1906, 1907, 1933, 1934, 1935, 1938
148 - 1942
147 - 1945

143 - 1995
138 - 1919
123 - 1918
112 - 1994
103 - 1981
The last five are famous exceptional seasons (in 1972, a sixth exception, the minimum was 153 games). The seven seasons 1926-1945 (bold) represent only 11 teams.

Between 1919 and expansion very few pairs of teammates started even 75 games: seven pairs 1920-23 and six beginning 1948. So there is no match with a team that played 150 or fewer games, nor did the the 1942 White Sox or 1945 Indians get 74 starts from two pitchers. So I will focus on 76 starts.

2.
Between 1919 and expansion only eight pairs of teammates started 76 games.
1920 * CHA Faber 39, Williams 38
1920 * CHN Alexander 40, Vaughn 38
1920 * CLE Coveleski 38, Bagby 38
1921 o CHA Faber 39, Kerr 37
1950 * BSN Spahn 39, Bickford 39 = 78 of 156, 50%
1950 o BSN Spahn 39, Sain 37
1950 o BSN Bickford 39, Sain 37
1956 * PIT Friend 42, Kline 39 = 81 of 157, 51.6%
* at least half of team games
o fewer than half of team games

The great anomaly is Pittsburgh in 1956, with Bob Friend and Ron Kline starting 81 of 157 games. Otherwise only the three pairs in 1920 and Spahn and Bickford in 1950 started half of team games.

3.
During the same period 1920-1961, only 12 pitchers started 40 games for one team. George Uhle with 44 in 1923 tops the list followed by Bob Feller in 1946 and Bob Friend in 1956 with 42.

38 starts in 154 games and 40 starts in 162 games represent the same share of team games. There were 62 pitcher-team-seasons with at least 38 starts during those 42 years. Beginning in 1962 there were 74 pitcher-team-seasons with at least 40 starts, 72 in the 1960s and 1970s and the last by Charlie Hough in 1987. During 1962-1979, the first 18 years under 162 games, 40 starts were about three times so frequent as 38 starts during 1920-1961, about four times if the comparison is to 1924-1961. The increase in number of teams doesn't account for much of that.

4.
The number of doubleheader games played should be included in the analysis. Comparing the 1930s and 1970s, the number of game days increased in proportion greater than 154 to 162.

(Yes, the Spahn/Sain Braves but Vern Bickford gets a raw deal.)

Doc N
By 1963, the Dodgers and Giants both had two men start half the team's games -- in the previous 40 years that had only happened once in all of baseball (the Spahn/Sain Braves). Maybe I'm missing one or at most two other examples. But two teams did it in one year. They could do it because the 4-man rotation had replaced the every-man fluidity of before. Those two Dodgers in 1963 had no relief appearances. The pair of Giants had only 1.

When it comes to pitcher usage patterns, the early 1960s mark one of the most profound changes in the history of MLB.

There, that's one less page from my book for you all to read.

5.
Accounting for doubleheaders, this might show up clearly as an early 1960s change. Or it might show up as an early 1960s change in the National League, presumably by emulation of the powerhouse Giants and Dodgers.
Without accounting for doubleheaders, the record is thin and it shows the two leagues remarkably out of phase.

Teammate pitchers with 80 starts (1961 to date, 26 pairs)
National League (14 pairs including 9 Dodgers)
Chi 69(3)
L.A 62* 63 65(3) 66 69(3)
Phi 66
S.F 63

American League (18 pairs including 8 White Sox)
Bal 70
Chi 71 72(3) 73 74 75(2)
Cle 73
Det 71 72 73 74
Min 65
Oak 70 74
Tex 74

That is 14 in the National League of the 1960s, 17 in the American League of the 1970s, plus the 1965 Minnesota Twins. Almost complete separation in time.

6.
By the way, seven of eight pairs for the 1971-1975 Chicago White Sox include former reliever Wilbur Wood. Next year they converted both relievers Rich Gossage and Terry Forster to starting.

Joe wrote:

I will post a thread Sunday to start discussion of the catchers next week, unless anyone objects . . .

I have no objection, but I would like to have the thread!

OK, I'll set it up now (or later today at worst).

Very interesting Chris - I was aware of the fluidity to rotation transition (this is very noticable if you play an old (1920s or 1930s let's say) replay with Diamond Mind or something, because of all the double-headers (and extra off days). I had no idea of when the transition occurred.

Thanks!

I also sent an email to the group to encourage participation. If you didn't get this message, you aren't registered for the Yahoo group.

OK, Joe. I'll bring this over here.

I wrote:

There is, to me, obvious benefit in weighing in our collective opinion on "their guys", telling exactly why they are not deserving of being honored.

Joe responded:

I'm not sure that there is . . . the goal of this project was never to knock the Cooperstown mistakes down a peg. It was more to see who was getting shortchanged by the Hall of Fame.

The Hall of Fame is never going to 'unelect' anyone (barring an OJ Simpson or Pete Rose caliber post-career scandal). So I don't really see much point in that.

As a group we've decided that Ferrell, Schalk and Lombardi (in this case) are below the 20 listed above. I think that's enough, right?

I think that’s again stopping short of being all we can be.

By not electing the “HOF 3” (more so, by having them far from contention for election) we have, in actuality, knocked them down a peg. In effect, we have unelected them by putting others in their place. Sure, that wasn’t ever our goal, but it’s certainly a real byproduct of the process.

So we’ve done that; I agree that we don’t want to go further down that road. What we should be doing is creating context for the HOF electors. What we should be doing is offering discussion and evaluation of their guys vis-à-vis our guys and other leading candidates.

By analyzing where their guys fall short, by studying how those elections came about, by offering a comprehensive analysis of not just the HoMers, but the HOF-not-HoM, plus a few other strong candidates, we would accomplish much more.

And I’m not suggesting anything all that difficult. There’s the 3 HOF catchers plus 4 others getting some semblance of support in our elections (Howard, Schang, Parrish, Munson). We’ve all presumably studied these guys, so offering reasons for their placements should not be too hard. Besides giving our sagacious analysis to the Hall electors, there is also the benefit of further scrutiny of the candidates for our future HoM elections (especially beneficial for new voters).

Or y’all can just be content to deal with “our guys”.

I could maybe see having separate elections positionally to rank everyone who got a vote in the last HoM election as well as Hall of Famers that aren't in the Hall of Merit. To set up a pecking order, add to the list, whatever you'd want to call it.

But I wouldn't want them all on one ballot. I don't want to have to explain why player X (just barely out of the Hall of Merit) is ranked ahead of player Y (elected to the Hall of Merit long ago).

Is that something like what you are thinking?

Ya know, that could work. An even simpler idea is, instead of separate elections, to add two subsections at the end of the HoM positional ballots. The first subsection would be for "My next ten". The second would be for any remaining HOFers, with their approximate ranking in "my system". A sample ballot could look like this:

1. Gibson--(with the required explanatory blah-blah)
2. Bench
3. Berra
4. Carter
5. Hartnett
6. Campanella
7. Cochrane
8. White
9. Ewing
10. Fisk
11. Santop
12. Dickey
13. McVey
14. Simmons
15. Freehan
16. Torre
17. Bennett
18. Bresnahan
19. Trouppe
20. Mackey
-------------------------
21. Howard
22. Schang
23. Parrish
24. Munson
25. Lombardi(HOF)
26. Tenace
27. Porter
28. Petway
29. Boone
30. McGuire
-------------------------
~40. Ferrell(HOF)
~45. Schalk(HOF)

Then do likewise for each of the other positions. You could make explanations for the two subsections optional, requiring explanations only for the HoMers. Again, simple.

except that in many cases some of my "next ten" outrank the HoM guys.

except that in many cases some of my "next ten" outrank the HoM guys

Then dispense with the numbering for the next ten. Right or wrong, Joe said he doesn't want to open that can of worms.

Then dispense with the numbering

I should have said, dispense with continuing the number sequeance and number them from 1 to 10.

Paul,

Something definately happened in the ealry 1960s. It was in that same period managers stopped leveraging their pitchers against particular opponents. It had been common from at least the 1880s to 1950s, but by 1965 it was dead and gone.

"dead and gone" overstates that situation somewhat. Even today, rotations are sometimes tweaked so that eg the Red Sox ace pitches the first game in Yankee Stadium instead of the last game in Tampa the previous night. But much less common than it once was.

Southworth managed the Boston Braves 1946-1950 and half of 1951. The following six "couplets" represent the pitchers who started more than 12 games for the Braves during those six seasons (all at least 15). Line one gives the numbers of starts, line two the numbers of reliefs.

It looks like sharp definition of the starting pitcher role was a two-season anomaly. Suddenly in 1949 and 1950 the big three worked only two (1949) and three (1950) games in relief.

1946
34 27 21 21 16
_3 _1 _4 15 _8

1947
35 35 30 20
_3 _5 _6 _2

1948
39 35 30 22
_3 _1 _7 11

1949
38 36 36 22
__ _1 _1 _8

1950
39 39 37
_2 _1 __

1951
36 33 22 20 19 15
_3 _4 _4 _5 14 _5

For the 1956 Pittsburgh Pirates, on the other hand, Bob Friend and Ron Kline worked 7 and 5 games in relief as well as 42 and 39 starts. For number three Vern Law it was 7 and 32.
The 1956

The 1948 team was the pennant winner. The 1949 and 1950 teams both finished fourth. So, I speculate, Billy Southworth did not become the genius of his day, forging a new norm in the new use of pitchers a la Tony LaRussa.