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

Thursday, July 08, 2004

1929 Results - Wallace, Thompson finally elected

Two long waits have finally ended, as Bobby Wallace and Sam Thompson were elected to the Hall of Merit.

Wallace was elected in his 10th try. For Thompson the wait was the longest in the history of the Hall of Merit. Big Sam came on the ballot in 1902, and was elected in his 28th year of eligibility. Thompson died in 1922, and will be inducted posthumously. Four players that Thomspon beat in 1902 have since been inducted (Pud Galvin-1910, Harry Stovey-1916, Charlie Bennett-1921 and Cal McVey-1914).

Thompson’s election was a fairly major upset. He was outpointed 601-526 by Sheckard in 1928. Thompson found his way onto two more ballots this year, while Sheckard fell off of one. Wheras both were named first or second on 7 ballots in 1928, Thompson managed 15 such votes this year, to Sheckard’s nine and beat him 629-616.

Sheckard remains the favorite for 1930, with Bob Caruthers and Dickey Pearce close behind.

RK   LY  Player             PTS  Bal     1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
 1    3  Bobby Wallace      703   40     8  6 11  2     3     3  1     1     2  1  2
 2    5  Sam Thompson       629   38     7  8  2  1  2  2     4  1  3  1  2  2  3
 3    4  Jimmy Sheckard     616   35     5  4  4  9  7        2  1     2  1
 4    6  Bob Caruthers      582   37     5  5  2  4  5     2  2  2  1  3  1  1  3  1
 5    7  Dickey Pearce      551   36     7  3  4  2  2  3     1     3  1  3  1  4  2
 6    8  Lip Pike           488   31     4  4  2  2  1  1  5  2  2  2  4  1  1
 7    9  Jake Beckley       448   31        4  2  5  1  6  4     1  1  2           5
 8   10  George Van Haltren 446   34        2  2  3  3  3  2  1  6  4  1  2  2  1  2
 9   12  Clark Griffith     381   33     1     1  1  2  2  5  4  1  1  3  2  2  3  5
10   15  Rube Foster        367   25     3  1  2     2  1  4  4  2     4  1  1
11   11  Jimmy Ryan         361   28           3  2  3  3  2  2  3  3  2  1  2     2
12   13T Roger Bresnahan    332   25     1  2     1  2  2  3  2  4  2        4  1  1
13   16  Mickey Welch       315   20     3  3  1     5     1  1  1        1  1  2  1
14   19  Cupid Childs       296   26        1     1  1  3  1  3  1  3  1  4  3  3  1
15   17  Hughie Jennings    295   22     1  1  1  2     2  2  3  2  3  2  1     1  1
16   13T Rube Waddell       293   22           2  4  2  3  2  2  2           2     3
17   18  Hugh Duffy         259   24           1  2  2     3  1        2  3  5  3  2
18   21  Pete Browning      246   18     1  1  1  3  2  1  1  1     1     3        3
19   22  Larry Doyle        218   18     1  2  1           1     1  2  3  1  2  3  1
20   20  Bill Monroe        216   19        1  1        1  1  1  3  1  1  6  1  2
21   23  Tommy Leach        205   20                 1  2  1     2  3  3  1  2  3  2
22   25  Charley Jones      186   15           1  1  1  1  2  1  3  2        1  1  1
23  n/e  Spotswood Poles    161   16           1  1              3  2  2  1     3  3
24   24  Ed Cicotte         147   12           1  1  1  2     2        1  1  1  1  1
25   26  Gavy Cravath       125   11                 2     1  2     2     1  2     1
26   30  Frank Chance       109    9.5            1     2  1           2  1  2     0.5
27   28  Ed Williamson       96    9                       1     1  3  2  1  1
28   27  Addie Joss          80    7                          3  1  2           1
29   29  Fielder Jones       68    7                       1     1  1     1  2     1
30   31  John McGraw         68    5     1              1  1              1        1
31   32  Ed Konetchy         53    4.5         1           1              2        0.5
32   33  Vic Willis          44    4                    1        1     1        1
33   39T Herman Long         41    5                                   2     1  1  1
34   34  Jim McCormick       38    4                                1     3
35   35  Tommy Bond          36    3                    1     1              1
36T  36  Lave Cross          34    3                    1        1              1
36T n/e  Bruce Petway        34    3                    1        1              1
38   39T Mike Tiernan        25    3                                1        1     1
39   37  Silver King         18    2                                1           1
40   38  Donnie Bush         18    1           1
41   42  Tony Mullane        17    2                                   1        1
42   41  Tom York            16    1                 1
43   43  Sol White           10    1                                   1
44T  44T Levi Meyerle         9    1                                      1
44T  44T Johnny Evers         9    1                                      1
46T  47T Harry Wright         8    1                                         1
46T  47T Jimmy Williams       8    1                                         1
48T  --  Ginger Beaumont      7    1                                            1
48T  47T Joe Tinker           7    1                                            1
48T  47T Ray Chapman          7    1                                            1
51T  --  Fred Dunlap          6    1                                               1
51T  --  Billy Nash           6    1                                               1
51T  44T Hippo Vaughn         6    1                                               1
Dropped Out: none.
Joey Numbaz (Scruff) Posted: July 08, 2004 at 03:35 AM | 83 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.

   1. Joey Numbaz (Scruff) Posted: July 08, 2004 at 03:52 AM (#722713)
Making this a hot topic . . .
   2. John (You Can Call Me Grandma) Murphy Posted: July 08, 2004 at 05:00 AM (#722802)
Looks like Pearce is in by '31, unless Harry Hooper has more support than I think he has. At any rate, Caruthers should best Pearce this year.
   3. DavidFoss Posted: July 08, 2004 at 05:01 AM (#722805)
... inexorable ...

:-)
   4. Dag Nabbit is part of the zombie horde Posted: July 08, 2004 at 05:14 AM (#722813)
53 players listed. Is that the record? 14 got first place votes. Can't wait to see how wide open the '32 ballot is.

Wallace & Thompson were the only ones on 80% of the ballots.
   5. John (You Can Call Me Grandma) Murphy Posted: July 08, 2004 at 05:28 AM (#722830)
... inexorable ...

:-)


The past few weeks I could actually use that word with a straight face about Pearce. :-)
   6. Kelly in SD Posted: July 08, 2004 at 07:16 AM (#722943)
To paraphrase ... "The journey to the Hall of Merit starts with the first step." - A relative of Big Sam at the induction ceremony.
   7. OCF Posted: July 08, 2004 at 07:42 AM (#722955)
Time for some more electoral navel-gazing: the latest consensus scores. I've been calculuting this for 9 years, since 1921. The 1929 results set new lows for: lowest average score (-0.5), lowest high score (+10), lowest low score (-21), and lowest best-possible score (+17). (The worst-possible score, from a ballot naming 15 players with no other support from anyone, would be about -50 except that the ballot would be disallowed.)

In the table below, the column "9 yr rel avg" is the average amount, per year voted since 1921, that the voter's score is above or below the average consensus score.The most extreme change in a score over the last few years is probably mine, from strong agreement to strong disagreement in five years.

Voter              1925 1926 1927 1928 1929   9 yr rel avg
Ken Fischer         13   12        12   10      7.1
Chris J             14   11   14   11    8      7.0
RMc                           17                7.0
David Foss          13    6    9    9    3      6.0
Howie Menckel        8    8   11   10    2      5.8
Devin McCullen      12    9   11   12    5      5.8
daryn               12    7   10    5    6      5.8
Sean M                    6                     5.3
Brad G              11   11   10   12    7      4.9
Joe Dimino          13    7    9    0    4      4.6
favre               11    7    8    9    2      4.6
mdb1mdb1            11    9    5    6    4      4.6
Tom H                6    6    6    7    5      4.3
Chris Cobb          10    7    6   12    6      4.2
TheGoodSamaritan     9    2    8    7   -3      4.2
DanG                 6    8    8   14    9      3.7
Don F               13    5   -3    7    2      3.7 
Sean Gilman         11    3    6    8    6      3.5
Philip              11    5    8    9    2      3.3
Andrew Siegel        8    1    6    8    2      3.0
Rusty Priske        11    8    6    6    8      3.0
Michael Bass         8    5    4    4    7      2.9
Al Peterson          5    3    8    6   -1      2.3
Kelly in SD                         6    1      2.1
Ron Wargo            5    6    7    7    0      2.0
stephen             11    3   -1                1.7
Rick A               9    2    1    3    2      1.5
Esteban Rivera       5    0    1    5   -2      1.3
Rob Wood            10    0    4    5    2      0.9
Adam Schafer         6    1    4    9    5      0.9
Michael D            8    3   -6    4   -2      0.7
PhillyBooster        6    1    4    3    3      0.6
OCF                 10    0    3    2  -11      0.2
Brad Harris              -5    5    7    1      0.0
Dolf Lucky           8   -1    2    4   -5      0.0
Carl Goetz           6    3    6               -0.6
Patrick W                 0    5    3   -1     -1.3
Yardape              7   -9                    -1.7
Max Parkinson       -2         4    3    1     -2.0
robc                -3    1   -5    5   -1     -2.1
Dan Rosenheck        2   -7    8    2   -4     -2.5
jhwinfrey                -5   -1    2   -1     -3.0
Zapatero             6   -7                    -3.8
Jeff M               1   -7    1    1   -8     -4.5
jimd                -7   -2   -4    2   -1     -5.0
karlmagnus          -4   -6    1   -2   -2     -6.0
dan b               -4   -6    0    0  -11     -6.2
Jim Sp               2  -11   -2   -5  -13     -6.3
EricC               -8   -8   -5   -6   -5     -9.8
sunnyday2           -8   -9   -3        -5    -10.2
Seaver 1969              -6   -5  -14         -10.7
yest                -9                  -8    -11.0
KJOK                -9  -16  -10   -9  -16    -13.2
John Murphy         -9  -18  -12  -13  -16    -13.2
Guapo                         -9  -12  -21    -16.2
      
Average            5.3  0.9  3.3  4.0 -0.5 
   8. Kelly in SD Posted: July 08, 2004 at 07:45 AM (#722961)
I can't see Hooper having enough support to damage Pearce's candidacy.
He does project to have the 3rd most WS in 1931 (assuming Sheckard gets in). But, in other measures of WS, Best 3 cons yr, Best 7 yrs, per 162g, Hooper will rank between 18th-25th. In OPS+, around 20th.
He has NO Stats AllStars and only 1 WS AllStar. He has no Black Ink. He received 5 points in the 8 MVP votes during his career. Defensively, WS has him as a C+ with no Gold Gloves.
Using BP #s: His 383 BRARepPos would be top 10, but per 162 would be about 20th. I don't know where his 55 FRAR would rank, nor his 96.8 WARP3.
These numbers do not argue well for a HoMer, IMHO.
   9. EricC Posted: July 08, 2004 at 11:26 AM (#723003)
I don't know where his 55 FRAR would rank, nor his 96.8 WARP3.

Hooper's 96.8 WARP3 will be the highest in 1931, so he has to be considered a serious candidate, but his Beckley-like lack of a peak will probably ensure that he doesn't get elected quickly. He will make my ballot.
   10. Jeff M Posted: July 08, 2004 at 12:11 PM (#723013)
How many ballots were cast?
   11. Michael Bass Posted: July 08, 2004 at 12:14 PM (#723016)
Very, very happy to see both Griffith and Foster moving up the tally, especially the latter. That said, there have been a lot of "big moves" in their range before that didn't stick (Rube Waddell), so we'll see if this jump is for real this year.

Looks to me like Caruthers/Pearce will be a great race this year. There may well be a case of "cold feet", much like hit Sheckard, with Bob this season. I myself looked at him again to make sure, but he's probably staying #2 on my ballot.
   12. andrew siegel Posted: July 08, 2004 at 01:25 PM (#723037)
If no one changes their preferences among the candidates (as if), Sheckard moves into the bonus points on 5 ballots, Caruthers on 3, and Pearce on 2. Also, again if no one changes his mind, Sheckard seems poised to join 6-8 ballots while Caruthers and Pearce only look to add 1-3 ballots. Looks like Sheckard opens with about a 75 point lead on Caruthers and around 115 or 120 on Pearce. Caruthers's lead on Pearce is ever so slightly bigger than it appears from last year's stats (40-45 points).
   13. John (You Can Call Me Grandma) Murphy Posted: July 08, 2004 at 02:15 PM (#723078)
Defensively, WS has him as a C+ with no Gold Gloves.

That's actually an excellent mark for a rightfielder using WS.

Hooper's 96.8 WARP3 will be the highest in 1931, so he has to be considered a serious candidate, but his Beckley-like lack of a peak will probably ensure that he doesn't get elected quickly.

I'm hoping for at least a few elections. :-)

How many ballots were cast?

48
   14. Guapo Posted: July 08, 2004 at 02:38 PM (#723103)
The 1929 results set new lows for. . . lowest low score (-21)

What do I win? Please, let it be an autographed Ed Konetchy baseball card!
   15. Dag Nabbit is part of the zombie horde Posted: July 08, 2004 at 02:48 PM (#723112)
Hooper's 96.8 WARP3 will be the highest in 1931, so he has to be considered a serious candidate, but his Beckley-like lack of a peak will probably ensure that he doesn't get elected quickly.

He has less of a peak than Beckley. Hooper only had three seasons where his OPS+ was as high as Beckley's career mark.
   16. Howie Menckel Posted: July 08, 2004 at 02:55 PM (#723121)
Back by zero popular demand. Players still on ballot in CAPS (Thompson and Wallace receive their final pts).

Through 1929, 5000+ vote points..

Thompson 12349
Bennett 11503
CARUTHERS 10011
H. Stovey 9576
PIKE 8431
Start 8378.5
McGinnity 8232
McVey 7985.5
Grant 7969.5
DUFFY 7339.5

PEARCE 6783
Galvin 6585
RYAN 6490
BROWNING 6361.5
VAN HALTREN 6272.5
Sutton 6070
McPhee 5921
JENNINGS 5901
Wallace 5808
SHECKARD 5623

BECKLEY 5121
   17. Patrick W Posted: July 08, 2004 at 03:30 PM (#723170)
In 32 years, 8 players have had a higher score in an election than an eventual
inductee and not yet been inducted themselves.

Player       Count (Years)
E.Williamson  2    '98-'99
Caruthers    10    '99-'02, '16-'19, '21, '26
Browning      3    '99-'01
Duffy         9    '07-'15
Ryan          4    '09-'12
Sheckard     10    '19-'28
Pearce        1    '26
Pike          1    '26
----------------------------------------------
Total        40

It seems likely that Caruthers and Sheckard will literally cut this list in half next year. Pearce & Pike can thank the Joe Jackson boycott for their appearance on this list. Most of the others can thank either Thompson, Grant, McVey, Galvin and/or Bennett.

Sheckard (3X), Duffy and Caruthers are the only players to recieve a score above 50% in a single year and not been inducted.
   18. ronw Posted: July 08, 2004 at 03:51 PM (#723193)
I looked at the constitution this weekend (and the Declaration of Independence, boy was Joe Dimino mad at the British) Two primary reasons for this project are to elect those forgotten by the Hall of Fame, and to keep out those who never should have been elected. So, how are we doing before we are inundated by FOFF?

17 HOMers who are not in the Hall of Fame:

Ross Barnes
Charlie Bennett
Bill Dahlen
Jack Glasscock
George Gore
Frank Grant
Pete Hill
Paul Hines
Joe Jackson
Grant Johnson
Sherry Magee
Cal McVey
Hardy Richardson
Joe Start
Harry Stovey
Ezra Sutton
Deacon White

Next year, it looks like Jimmy Sheckard and Bob Caruthers will be added to this list.

8 Hall of Fame players (or administrators with a substantial case for merit on their playing record) we have excluded from the HOM thus far, but have a decent chance of making it some day, along with 1929 points:

Jake Beckley (448) - It looks like he'll get in way off in the future.
Clark Griffith (381) - Also may make it. A good pitcher, made HOF in part because of longetivity of administration.
Rube Foster (367) - Again, made HOF in part because of longetivity of administration.
Roger Bresnahan (332) - Getting some support, but it may not be enough.
Mickey Welch (315) - 300 win clubber is gaining.
Hugh Jennings (295) - May never get to the HOM. Managerial accomplishments taken into account for HOF selection.
Rube Waddell (293) - Is dropping, but probably wouldn't care.
Hugh Duffy (259) - Is really dropping. He was #5 in one ballot.

Harry Hooper will soon be added to this group.

9 players who will not make the HOM, but are in the Hall of Fame:

Frank Chance (109) - Most supported of the Cubs trio, made the HOF in part on managerial exploits, in part on playing career, and in part on a poem.
Addie Joss (80) - A tragedy, but the same will soon be said about Ross Youngs.
John McGraw (68) - Had a bit of managerial experience taken into consideration in the HOF vote.
Vic Willis (44) - Dennis Martinez probably won't make it either.
Johnny Evers (9) - These are the saddest of possible words . . .
Joe Tinker (7) - Trio of Bear Cubs and fleeter than birds . . .
Chief Bender (0) - Has never even merited a single HOM vote, and is the only Hall of Famer with that distinction.
Tommy McCarthy (0) - Wow. I still shake my head that he slipped into the Hall of Fame. Someone voted for this possible all-time worst HOF selection in a past HOM election.
Hall of Famer Happy Jack Chesbro (0) - Happy Jack's title as worst HOM-eligible Hall of Fame pitcher will soon end, because . . .

Rube Marquard is joining this list in 1931.

Miller Huggins and Wilbert Robinson had some playing career (they each made Bill James' all time top 100 position lists), but made the HOF primarily as managers. Bucky Harris will soon join them.
   19. DavidFoss Posted: July 08, 2004 at 04:01 PM (#723206)
Hall of Famer Happy Jack Chesbro (0) - Happy Jack's title as worst HOM-eligible Hall of Fame pitcher will soon end, because . . .


... he'll be elected next year in a landslide!!!

WOOHOO!
   20. DanG Posted: July 08, 2004 at 04:56 PM (#723291)
Calling all supporters of John Dwight Chesbro...elect the living!

Enjoying his retirement in exotic Massachusetts, he isn't likely to last more than two more elections.
   21. Al Peterson Posted: July 08, 2004 at 07:50 PM (#723609)
Thanks Ron for running the comparison of HOM vs HOF. It goes to show you that pitching and defense are more heavily weighted by Cooperstown than by this group. How many pitchers among the HOM/non-HOF group? Yeah, that would be zero until Caruthers is elected and even he is a hybrid.
   22. OCF Posted: July 08, 2004 at 10:12 PM (#723790)
So many different players got 1st or 2nd place votes in the 1929 election that we can try to assemble a team from just them. In what follows, the two numbers in the parentheses are 1st and 2nd place votes.

C: Bresnahan (1, 2); could also claim Pearce.
1B: Beckley (0, 4)
2B: Doyle (1, 2); also Childs (0, 1) and Monroe (0, 1); could claim Pike as well.
3B: McGraw (1, 0); could also claim Browning (1, 1).
SS: Wallace (8, 6); also Pearce (7, 3), Jennings (1, 1)
LF: Sheckard (5, 4)
CF: Pike (4, 4); also Van Haltren (0, 2)
RF: Thompson (7, 8)
P: Caruthers (5, 5), Griffith (1, 0), Foster (3, 1), Welch (3, 3)

A full team.
   23. EricC Posted: July 08, 2004 at 10:45 PM (#723831)
He has less of a peak than Beckley. Hooper only had three seasons where his OPS+ was as high as Beckley's career mark.

True. It should also be kept in mind that Zack Wheat was almost an exact contemporary of Harry Hooper. I'll have Wheat ahead of Hooper when he becomes eligible in 1933.
   24. DavidFoss Posted: July 08, 2004 at 11:06 PM (#723859)
Hooper is quite similar to HOF-er Max Carey. Carey's has some mean SB and SB% numbers, but neither one will get my vote.
   25. John (You Can Call Me Grandma) Murphy Posted: July 08, 2004 at 11:16 PM (#723895)
Wheat, Carey than Hooper for me. The latter two don't look like they'll ever be on my ballot, while the former is not remotely an inner circle candidate.
   26. Kelly in SD Posted: July 09, 2004 at 12:22 AM (#724261)
Someone with an understanding of WARP1 to WARP3 conversions, please help.

While trying to compare Wheat and Hooper, I saw something I didn't understand on BPro's player pages. Hooper's WARP1 was 107.8 and WARP3 was 96.8. Wheat's WARP1 was 123.9 and WARP3 was 85. I have not read the yearly BPro books so I didn't know what the difference was so I looked at the Glossary. It said that WARP3 adjusts WARP1 for league quality and schedule length.
Since the players played in 154 game schedules, schedule lenght should not be much of a factor though a team could have more games played b/c of more ties.
If I understand it right, Hooper's WARP is reduced by 10% because of league quality and Wheat's by 32%. If that is right, then if Wheat had the same WARP1 has Hooper, his WARP3 would be about 73? Was the National League only 75% the quality of the American League from 1909-1926?
Am I misunderstanding what the adjustment from WARP1 to WARP3 means?
I would like to understand WARP because many voters use it.
If this should be posted in another area, please let me know.
   27. Jeff M Posted: July 09, 2004 at 12:34 AM (#724313)
Someone with an understanding of WARP1 to WARP3 conversions, please help.

and

Was the National League only 75% the quality of the American League from 1909-1926?

I don't think you are misunderstanding the conversion. Many people, including me, don't use WARP3, but instead used WARP1 with a season-length adjustment where appropriate. I think us not understanding specifically how the numbers are adjusted is grounds enough not to use WARP3.

My resolve to use only WARP1 is reinforced when I see an example like this. The adjustment you mention appears to treat the NL as poorly vis-a-vis the AL during that period as the AA is treated vis-a-vis the NL from 1882-1891.
   28. Jeff M Posted: July 09, 2004 at 12:45 AM (#724361)
Wheat...is not remotely an inner circle candidate.

Guess it depends on the size of the circle. Wheat looks a lot like Clemente in value to me, unless you timeline them. In WS, they're the spittin' image of each other and in WARP1 they're darn close.

WHEAT, ZACK
Games: 2,410
WS Peaks (3-5-7-9): 95-149-199-243
WS Career: 380
WS per 162: 25.54

WARP1 Peaks: 31-49-64-78
WARP1 Career: 124
WARP1 per 162: 8.34

CLEMENTE, ROBERTO
Games: 2,433
WS Peaks (3-5-7-9) 94-149-200-247
WS Career: 377
WS per 162: 25.10

WARP1 Peaks: 32-51-67-82
WARP1 Career: 122
WARP1 per 162: 8.14
   29. EricC Posted: July 09, 2004 at 01:24 AM (#724519)
Someone with an understanding of WARP1 to WARP3 conversions, please help.

While trying to compare Wheat and Hooper, I saw something I didn't understand on BPro's player pages. Hooper's WARP1 was 107.8 and WARP3 was 96.8. Wheat's WARP1 was 123.9 and WARP3 was 85.


WARP2: Hooper 93.5 Wheat 82.3

I believe that the conversion from WARP1 to WARP2 is mainly a replacement level offset. Let's say that Hooper and Wheat both played about 18 seasons of about 550 plate appearances per season. Then Davenport is implicitly saying that the replacement level for the AL was typically (93.5 - 107.8)/18 = -0.8 wins per season lower than some reference point, and the replacement level for the NL was about (82.3 - 123.9)/18 = -2.3 wins per season lower than the same reference point. Thus, the WARP difference between the NL and the AL replacement level averaged 1.5 wins per season over this era.

I've also calculated league factors, and agree that the NL was weaker than the AL over this era. However, I estimate the difference to be more like 2 Win Shares per season for a player with 550 plate appearances. Using 3 Win Shares = 1 win, my analysis would suggest a difference between the league replacement levels of about 0.7 wins per season over this era, or about half of Davenport's numbers.

I think that real distortions occur if WARP1 is used without any correction for league strength (Hornsby will be overrated, for example, not that that matters in his case), but the corrections in going from WARP1 to WARP2 may be too extreme.

Wheat...is not remotely an inner circle candidate.

But will he be a first-ballot HoMer?
   30. Jeff M Posted: July 09, 2004 at 02:14 AM (#724752)
Eric:

Can you explain how you calculate your league factors? I'm interested in the methodology because I think it is a very thorny problem.

It always seemed like an impossible task to me, because it seems like you can't just use the league stats and standard deviations. If the hitting numbers are better in one league, that might mean that the average hitter is better in that league and the average pitcher is the same, in which case the league is stronger. Or it might be that the average hitter is the same but the average pitcher is worse, in which case the league is actually weaker, even though the hitting numbers are higher.

Similarly, if you use runs, it's hard to tell if more runs equals more quality (better hitting against the same pitching) or less quality (poorer pitching against the same hitting).

Does replacement level analysis solve that? It doesn't seem like it would, since it is just using the same keague hitting and pitching numbers. Also, in mature leagues like the NL and AL, why would there be different replacement levels in the two leagues, from a conceptual point of view?
   31. sunnyday2 Posted: July 09, 2004 at 02:14 AM (#724755)
Prediction time:

1930--Sheckard, Caruthers
1931--Pearce (John is buying)
1932--Santop, Pike or (in case Pike also catches a case of Sheckard-itis) Beckley or Van Haltren
1933--Johnson, not Wheat? OK, then the second among Pike or Beckley or Van Haltren.

This, BTW, is deeper than I ever thought we would get into the backlog. I really thought Wheat would be a shoo-in until reading the above comments. I also personally plan to have Mendez in the #1 or 2 spot in '32 but I don't expect y'all to agree.

1934--Cobb and Speaker
1935--Collins and Lloyd
1936--Alex and Smokey Joe
1937--Torriente and Heilmann
1938--your guess is as good as mine--the third among Pike or Beckley or Van Haltren, or Mendez or Ben Taylor or Wheat or Max Carey--OK, no more cop out, I'll say that Pike and Beckley and Van Haltren are all in by now, and Ben Taylor
1939--Wheat and Mendez
1940--coming up on the outside is Rube Foster, and Max Carey
   32. Jeff M Posted: July 09, 2004 at 02:16 AM (#724764)
keague league
   33. sunnyday2 Posted: July 09, 2004 at 02:29 AM (#724834)
BTW, I have no clear idea of where Wheat will be on my ballot other than to say that at a glance he looks to be squarely in the Stovey-Kelley-Sheckard-Magee-Hill class, all of whom have been borderline on my ballot, none being in my PHoM yet but all still on my consideration list for PHoM. If Wheat is "in" that class, but at the top, he could go in my PHoM in a hurry (1933) but if he is only somewhere in the middle of the pack, then it could be a long slow climb.

As to the Clemente comp, it is probably politically incorrect to say that this just indicates how vastly overrated Roberto has become. His case is a lot like Joe Jackson's. He has been overrated by history because it makes a better tragedy that way. Not that Clemente, like Jackson, won't go into my PHoM (or the HoM for that matter). He will.

But remember the Clemente family getting all huffy because he didn't make the all-century team? I mean, c'mon.
   34. Howie Menckel Posted: July 09, 2004 at 02:37 AM (#724870)
Ridiculously unscientific comment: I watched Clemente play, and he's the most elegant player I'v ever seen.
no SABRmetric pts for that, just is what it is..
   35. sunnyday2 Posted: July 09, 2004 at 02:49 AM (#724925)
Howie, I would agree with you. Clemente illustrates for me the point about ability and tools and grace and athleticism being different than value.

I have also opined somewhere that Clemente is a great comp for Cris Torriente, but that is based on a 10% discount for Torriente off of KJOK's MLEs. Using Chris Cobb's 5% discount, Torriente looks better. With no discount, taking KJOK's MLEs at face value, and I am sure K believes he has already applied the appropriate difficulty or league adjustment and he may be right and Chris and I wrong. If that's the case, then Torriente was a lot better.

Also, using the Wheat/Clemente comp, that puts Torriente in Wheat's (and by extension, Stovey, Kelley, Sheckard, Magee and Hill's) class. I have a hunch Torriente was better than any of them, though I intend to continue to apply the 10% discount and then ask if the player is better, rather than starting at face value and wondering if he is worse.

PS. Shortly, I will be deciding among Stovey, Kelley, Sheckard, Magee, Hill and Wheat for my PHoM. Any advice?
   36. sunnyday2 Posted: July 09, 2004 at 02:59 AM (#724941)
Finally and then I will log off and go to bed. Thanks to Ron for the HoF comp and to Al Peterson for pointing out the obvious (which of course had not been obvious enough for me).

Gotta recheck those pitchers.

Mickey Welch, for one, has moved from out of consideration all the way up to 15-20 on my ballot and I think he will indeed be in my consideration set for a long time to come.

Rube Foster is moving up, too, and he will be in my PHoM soon. I also still like Tommy Bond and Jim McCormick.

And I like Waddell better than Griffith. But in addition to all of that, I will need to look more closely at the 250-299 WS pitchers, a whole bunch of whom are coming eligible soon or recently. I've pretty much written off the whole bunch, but that is almost surely a mistake. Not that I relish the thought of more analysis....
   37. Kelly in SD Posted: July 09, 2004 at 03:50 AM (#725010)
Another question re: WARP1 to WARP2 adjustments:
Does BPro adjust the quality level each year so the WARP2 reduction factor will vary with each year? Or do they go with a moving average of 3 or 5 seasons? If their adjustments are based on just one season's worth of data, that seems to me to be too small of a base from which to draw conclusions.
Also, how do they take into account a concentration of outliers in one league like Cobb, Speaker, Johnson, Collins, and JJackson or a concentration of sucky players on one team (like the 1915-22 As or various Boston teams)?
There glossary on their website doesn't describe how it works.
   38. Jeff M Posted: July 09, 2004 at 03:50 AM (#725011)
Howie, I would agree with you. Clemente illustrates for me the point about ability and tools and grace and athleticism being different than value.

Yes, and reiterates my explanation of the conceptual difference between the Bill James Sim Score (which is more likely to match style) and my WS Sim Score (which matches value, not style).

I'm not old enough to have seen Clemente play, but I've always heard how graceful he was, and my comment about Wheat/Clemente wasn't meant to put Wheat in that category or suggest the same level of athleticism. My comment just compared their values at various stages under WS and WARP.

Of course, the Clemente career numbers would have gone up a bit if he hadn't died, but his peak numbers probably wouldn't have changed, since he would have been 38. Wheat was still very good at 37, but faded at age 38 and 39 (as most players do).
   39. Kelly in SD Posted: July 09, 2004 at 06:34 AM (#725077)
To clarify my questions in Post 37.

How does one judge the quality of one league? Does an accumulation of poor players drag down the other players because their competition is lessened. For example, the 1909 NL had 3 teams lose at least 98 games. Or does a group of outliers raise the quality of a league?
Second, how does one compare the quality of leagues? Do you compare the performances of players who changed leagues in an offseason after taking into account park, league, and age factors? Do you compare every player to an "average" player and see if there are more players who are below-average to a certain degree? Or is one league's "average" or ".500" player compared to the other league's "average" or ".500" player, or to the respective league's "replacement" player?
But, if you measure value of players as opposed to performance of players, each league should have the same total value because each league should play the same number of games and have the same number of wins and losses (allowing for the random rain/dark out tie not completed). Based on value, if players in different leagues had the same value, but different performances would that serve as the comparison to establish league quality?
Okay, its 11:30 and I need to go bed. Sorry if any of this is too rambling to make sense.
   40. Joey Numbaz (Scruff) Posted: July 09, 2004 at 09:48 AM (#725108)
Great job on the HoM/HoF comparison!

RE: WARP1/3 and Wheat/Hooper

Looking at the components . . .

WHEAT WARP1/2 (NL 1909-27)
BRAR 747/603
BRARP 578/462
FRAR 385/142
FRAA 24/-97

HOOPER WARP1/2 (AL 1909-25)
BRAR 611/550
BRARP 431/383
FRAR 364/294
FRAA 114/55

One major difference is that Wheat played LF for his entire career, Hooper RF. In the all-time adjustments, it's quite possible that RF takes on a higher significance, relative to LF than it had from 1909-27. This would explain why Hooper loses 59 runs and Wheat 121 (relative to average).

It is very likely that the AL was significantly better than the NL during this era. Most of the mega-stars, Cobb, Heilmann, Speaker, Johnson, Ruth played in the AL. The White Sox despite a worse record were heavy favorites in the 1919 World Series, things like that make it seem pretty obvious.

As to how you'd 'figure out' which league was better, I'd look at players that switch leagues, adjust for age and see how they perform (better/worse). I'd be almost positive that's how Prospectus does it, and I think their conclusions generally make a lot of sense intuitively, for what that's worth. If their system didn't show the AL with a clear edge for this era, I'd be surprised.

So the Wheat/Hooper thing looks like it is partly league quality, partly the all-time fielding adjustments.
   41. Joey Numbaz (Scruff) Posted: July 09, 2004 at 09:50 AM (#725109)
"As to how you'd 'figure out' which league was better, I'd look at players that switch leagues, adjust for age and see how they perform (better/worse)."

I thought it was obvious, but just in case, I also meant that you adjust for park and offensive level of the league too . . .
   42. Joey Numbaz (Scruff) Posted: July 09, 2004 at 09:52 AM (#725110)
BTW - I think Wheat was a helluva player - I imagine I'dve voted him #1 for the past several elections - I'm pretty sure I'd take his career over Baker's. Hooper is going to be an interesting one, I haven't done the research on him yet, but eyeballing it I'd probably have him behind Beckley to start.
   43. EricC Posted: July 09, 2004 at 11:05 AM (#725114)
As Joe pointed out, WARP1 to WARP2/3 conversions may also include different positional adjustments for different positions, something that I neglected in comment #29.

Can you explain how you calculate your league factors? I'm interested in the methodology because I think it is a very thorny problem.

....how does one compare the quality of leagues?

I agree that it's a thorny problem. The principle on which I've attempted to analyze league factors is that, on average, players' performance vs. age will fit a pattern, and that collective deviations from these career arcs indicate changes in league quality. For such analyses, it may be best to use a metric such as win shares per game (my choice) or equivalent average, that has a fixed year to year average. Interyear comparisions of performance of each individual player then provide a large set of equations to solve. The best fit to these equations yields the league factors.

If the hitting numbers are better in one league, that might mean that the average hitter is better in that league and the average pitcher is the same, in which case the league is stronger. Or it might be that the average hitter is the same but the average pitcher is worse, in which case the league is actually weaker, even though the hitting numbers are higher.

Interesting points. The difference between the two cases is difficult to detect when using metrics that are tied to a fixed average. If, say, all of the pitchers in a league
were kicked out after one season and replaced by worse pitchers the next, then, because, the average win shares per game per player would stay the same, there would be no indication of a change in league quality in my methodology. Comparison of performance of players who switch leagues helps to calibrate the league factors, but, unfortunately, not that many players (and particularly few good players in their prime) switched between the NL and the AL in the post-1903 period.
   44. Howie Menckel Posted: July 09, 2004 at 11:42 AM (#725117)
Clemente's OPS+ in 1972 (138 in 378 ABs) was his 10th in a row over 130, which is a remarkable achievement.
130 also is his career OPS+ mark; he was mediocre for a number of years before taking off.

I think he could have Aaron-ed it for several more years.
   45. Jeff M Posted: July 09, 2004 at 01:17 PM (#725164)
About a year ago, I compared the statistics of everyone who was a semi-regular in both the AA and the NL to get a feel for the league quality differences. To qualify, they had to have at least one 140+ at-bat season in both the AA and NL.

Initially, I looked at every season those hitters played and there were 541 AA player seasons and 970 NL player seasons. There were several problems with that. First, the disparity between the AA sample size and the NL sample size. Second, the average age of the NL hitter in the set was 29 and the average age of the AA hitter in the set was 27, which might skew the data. Third, Joe pointed out I shouldn't go beyond 1892 because of the pitching rule change in 1893.

So I only used seasons between 1879-1892. Here are the results (with standard deviations in parens):

American Association

Avg. Age..............27.1 (3.8)
Player Seasons......541
OBA.....................0.324 (0.057)
SLG.....................0.360 (0.079)
OPS.....................0.684 (0.129)
OPS+...................1.065

National League

Avg. Age..............27.1 (3.4)
Player Seasons......529
OBA.....................0.308 (0.059)
SLG.....................0.348 (0.086)
OPS.....................0.656 (0.133)
OPS+...................1.014

Here are the good things about the study: equal sample size; players in both sets have the same age; standard deviations are in line with each other; no concern about pitching rule changes; no pollution with data of players who didn't play in both the leagues.

Here are the bad things about the study: maybe OBP, SLG and OPS aren't the right measures, but I wasn't sure what else to use; more importantly, I don't know what the results mean. All the NL hitting numbers are lower by about 5%, but we KNOW that the AA was quite a bit weaker over its life.

That's why it's such a thorny problem. In this case, I interpreted the numbers to mean that hitters who played in both the AA and the NL were average hitters when they played in the NL and above average hitters when they played in the AA, indicating (I think) that the AA pitching was weaker and/or that the other hitters in the AA weren't established a lower average.

Then there's the problem of quantifying that into an AA discount. And even if you could quantify it, how would you quantify it on a year to year basis, since accumulating enough player seasons to make the study valid means you have to look at players over a period of time (i.e, hardly anyone played in both the NL and the AA during the same season)?

I suppose someone could run the numbers for the NL and AL players over a specified period and come up with a similar chart. However, it's a lot of work and I still don't know if it is conclusive because good hitting might be good hitting or might be bad pitching.
   46. Chris Cobb Posted: July 09, 2004 at 01:39 PM (#725182)
sunnday2 wrote:

I have also opined somewhere that Clemente is a great comp for Cris Torriente, but that is based on a 10% discount for Torriente off of KJOK's MLEs. Using Chris Cobb's 5% discount, Torriente looks better. With no discount, taking KJOK's MLEs at face value, and I am sure K believes he has already applied the appropriate difficulty or league adjustment and he may be right and Chris and I wrong. If that's the case, then Torriente was a lot better.

I tend to apply a 5% discount to i9s ML projections, not to KJOK's MLEs. I haven't made use of them in my rankings yet because I'm not sure how to get from the offensive context in which he places the players to their contemporary one. I know KJOK has explained how this might be done somewhere, but I think that is buried in the old threads to which we no longer have full access.

I might end up applying the same discount to them that sunnyday2 does, but at this point I don't know.
   47. karlmagnus Posted: July 09, 2004 at 01:46 PM (#725189)
Looking at Jeff M.'s numbers. Given Caruthers, Foutz et. al. it may be that AA managers operated on a strategy of hiring pitchers who could knock the stuffing out of the ball, whereas NL managers didn't. We know now (or think we do) that hiring pitchers purely for their pitching ability, and ignoring their hitting, is optimal, but people didn't know that in the 1880s, and in the AA managers may have gone for a pitcher who could hit, on the gorunds that any defects in his pitching, he'd make up with the bat.

With the Caruthers/Foutz team in St Louis, the strategy was very successful, so it wouldn't be surprising if others tried it. A smaller study looking at the batting stats of AA and NL pitchers, 1882-93, would tell us whether this was the reason for the difference.
   48. Jeff M Posted: July 09, 2004 at 01:53 PM (#725196)
Also, I think Joe may have suggested one time that you can look at fielding numbers to help determine quality of play. That certainly makes some sense and is clearly evident if you compare seasons that are years apart.

I tried it with the AA/NL comparison in my struggle to come up with appropriate discounts.

I didn't want to use fielding percentage (for a variety of reasons), so I looked at what percentage of runs were earned. The theory was that the higher percentage earned, the fewer unearned runs and the better the defense. The better the defense, the better the league.

Over the ten years of the AA, 59.4% of the runs were earned (on average) and there was a .0042 standard deviation. Over the same period in the NL, 58.6% of the runs were earned (on average) and there was a .0054 standard deviation. That study didn't help me much, since I know the AA is weaker.

So I tried fielding percentage. Over the ten years of the AA, the average fielding pct. was .9067 and the average fielding pct. in the NL was .9110. That shows weakness in the AA, but is it statistically significant? I don't know.

I then calculated the average AL fielding pct from 1916-1925 and the same for NL. AL had a fielding pct. of .9662 and NL had a fielding pct. of .9661.

*For completion, I also determined that the percentage of runs earned in the AL from 1916-1925 averaged 80.2% per season (with a standard deviation of .039) and the NL had a percentage of 79.3% (with a standard deviation of .025).

I give up for now.
   49. PhillyBooster Posted: July 09, 2004 at 02:51 PM (#725288)
Another thing to consider is that the AA might have been made up of all "Pitcher's Parks," so even though the hitters hit a lot better against AA pitchers, they hit better in pitcher's parks so it was a wash.

There is, I believe, no way to tell the difference by looking at just the hitters. The best I can tell, the only way to truly gauge the difference is to look at pitchers also.

If pitchers performed 5% worse in the NL, then it is likely due to the hitters being 5% better. If the pitchers are 20% worse, then maybe the hitters are 5% better and the parks increase scoring 15%.
   50. Jeff M Posted: July 09, 2004 at 03:39 PM (#725373)
....I'd look at players that switch leagues, adjust for age and see how they perform (better/worse). I'd be almost positive that's how Prospectus does it, and I think their conclusions generally make a lot of sense intuitively, for what that's worth. If their system didn't show the AL with a clear edge for this era, I'd be surprised.

My goal, of course, is not to tear down BP, because they know a lot more about this stuff than I do.

However, I have concerns about us discounting NL players (i) if we can't seem to empirically show the NL was weaker and (ii) even if we could, we'd have a hard time quantifying it on a year-by-year basis. BP indicates that the NL was weaker, but doesn't explain why or how the specific year-by-year strength numbers were arrived at.

I determined which hitters from 1916-1925 hit in both the AL and the NL during that period. Out of 909 NL hitters who played in the NL in those 10 years, 186 also hit in the AL during that period. Of those 186, only 36 had at least one 140+ PA in a season in both the AL and the NL <u>over their careers</u>. I realize that if I took the list of NL players during that time period and examined if they played in the AL at any time the list would get a little longer. But even if I could get the number to 100, it would be such a small sampling of crossover semi-regulars over a long time period that I don't think it would mean anything (and certainly wouldn't permit a year-by-year quantification of a discount).

The discounts we apply (if any) to NL players will definitely have an effect on where they rank on the ballots. So I believe we ought to be cautious.
   51. PhillyBooster Posted: July 09, 2004 at 03:51 PM (#725403)
I concur with Jeff that, although it is likely that the NL was weaker in these years, it is a weakness that is lost in the noise, much the way the 1888 and 1889 AA was likely somewhat weaker than the NL, but it is impossible to say so with any statistical accuracy.

In my mind, until the 1950s, a pennant is a pennant in either league. As it is, the NL has more HoMers for each year, so far, from 1902 on. That, I believe, is not because we are downgrading NLers, but rather an example of how the NL was weaker (fewer stars).

After the 1950s, I see actual rationale for the NL being the superior league (they integrated more quickly), and -- in theory, at least -- the two leagues in the 1950s and 1960s should have similar numbers of white HoMers with the NL jumping ahead with extra black HoMers.
   52. Jeff M Posted: July 09, 2004 at 04:12 PM (#725453)
There is, I believe, no way to tell the difference by looking at just the hitters. The best I can tell, the only way to truly gauge the difference is to look at pitchers also.

You are probably right Philly. Of course, pitchers suffer from a similar sample size problem.

I determined which ptichers from 1916-1925 pitched in both the AL and the NL during that period. Out of 371 NL pitchers who played in the NL in those 10 years, only 85 also pitched in the AL during that period. Of those 85, only 33 had at least one 40+ IP in a season in both the AL and the NL over their careers. That number would be cut in approximately half if I had used 100+ IP as the standard.
   53. sunnyday2 Posted: July 09, 2004 at 06:21 PM (#725784)
I think Kelly asked a really important question, and I don't mean "How do you measure league quality." Not that that isn't an important question, actually, but at least that question is being discussed.

But he also asked, so, assuming you CAN measure league quality, "What do you do with that?" And that question hasn't been responded to.

What he said is, if we're measuring and honoring value here, isn't value = value? Didn't the 8 team AL of 1915 playing a 154 game schedule offer the same total value as the NL in the same year? If we discount for league quality or difficulty, we're saying there's less value there, which is not true. I mean, look at WS. There are (154 X 8 X 3)/2 of them available in each league.

Good question.

(An aside: I think it's also worth noting that there are two kinds of league difficulty adjustments. One is for different leagues at the same time, the other is for different times whether same league or different [that would be the infamous "timeline"]. On several dimensions, the timeline and the league difficulty [at the same time] adjustment are the same thing, the same question, and raise the same issues. In other respects, they're different, and thus we have two different names for them. This post addresses both the timeline and the league difficulty adjustment.)

Anyway, the question is why discount? If we discount, are we NOT then measuring value?

I think we are NOT then measuring value. Rather, the formula becomes something more like:

Value X Difficulty Adjustment = Merit

The Difficulty Adjustment approximately = the ability of the average player or the total cumulative ability or the replacement value or whatever of all the players in the league. But the formula does not = ability, because the value is indeed a constant over the short term, though it expands with the number of teams and the number of games played.

But the difficulty adjustment does water down value in order to recognize that over time the quality of play (defined how?) has gotten better. You might believe it's gotten better because the pool is bigger or because training and coaching and nutrition, etc. etc. etc., are better, but that really doesn't matter. Play has gotten better, I think we all agree, but some people are more determined and some less to get that into consideration.

But all of that aside, the big question for me is how to apply "average value" or a difficulty adjustment to *individual* players. When we apply a difficulty adjustment, we assume that every player in year X is equally different than every player in year Y, and that seems to fly in the face of common sense.

IOW, Dan Brouthers earns 35 WS in 1885 (I'm making the number up) and Jack Clark earns 35 WS in 1985. The NL in 1895 is only .60 as difficult as in 1995 so therefore Clark's 35 WS represent 40 percent more...more what? More value? More ability? More merit? But how do we know that the 40 percent average difference applies at the extremes, among the outliers? I don't see how we can know that.

So, to summarize and then I'll shut up, I go toward pure value because 1) it's in the spirit of this project, 2) it's unfair to think that Dan Brouthers, lacking the "necessities" of today, could possibly have done any mroe than he did, and 3) we really don't have any idea how to apply a timeline or a difficulty adjustment against the extreme cases, which are the only cases getting any votes here.

Does that make sense?
   54. Kelly in SD Posted: July 10, 2004 at 02:49 AM (#726886)
sunnyday2,
The point about each league having the same amount of value available each year is a significant point that got lost in my post. Thank you for bringing it out.
Also, thank you to Jeff M for bringing up how small are the data points available for using 2-League players to compare the quality of each league. Personally, that is not enough for me to trust that there was a large quality difference between each league.
I'll generally follow a "pure" value method and just have to make sure I do enough research.
   55. Howie Menckel Posted: July 11, 2004 at 06:12 PM (#728794)
Negro Leaguer career HOM point totals
GRANT 7969.5
JOHNSON 3118
Monroe 1764
Foster  1372
HILL 706
S White 297
Poles 161
Fowler 157
G Stovey 43
Petway 34
did I miss anybody?

DORMANT: No points in last three years
Griffin 1513.5
Whitney 852
O'Neill 791
Clements 265
Fowler 157
Hutchinson 141
Foutz 136
McGuire 119
Lyons 114
Mathews 96
Orr 86 Farrell 82 Thomas 67 Phillippe 57 Seymour 53 Zimmer 51 G. Stovey 43 Joyce 36 Latham 36 H. Davis 31 Dalrymple 31 Creighton 29 Stivetts 24 E. Smith 21 Hoy 20 McKean 17 Chesbro 14 Cuppy 10 Larkin 10 Hahn 8 Tannehill 7 Gleason 7 Buffinton 6 Leever 6 Powell 6 Hartsel 6 Selbach 6 Weyhing 6 McCarthy 6 Burns 6 Clapp 6 Nichol 6 Sunday 4 C. Welch 3 Werden 1 Cummings 1
   56. Sean Gilman Posted: July 12, 2004 at 01:14 AM (#729753)
This year's (very late) HOM Game features the New York Giants and the Pittsburgh Pirates.
Notable Giants include Carl Hubbell, Mel Ott, Edd Roush, Travis Jackson, Bill Terry and Fred Lindstrom.
The Pittsburghs feature the Waner Brouthers, Burleigh Grimes, and Pie Traynor.
   57. jimd Posted: July 12, 2004 at 09:31 PM (#730864)
This is what I remember of BP's description of the league comparison procedure as described in one of their annuals. The backbone of the process involves developing the evolution of each league from year to year; this involves
comparing everybody to themselves in other years (like the Cramer/Kremer (sp?) study but with BP's own metrics) and thus uses the largest sample possible. The internal evolution of the league provides far more data than the cross-league samples, though all of the cross-league samples from more than a century of trades and free-agent signings provide the data for the calibration of the two leagues relative to each other. There is nowhere near enough data to do a proper league vs. league comparison during much of the early history, so their numbers will derive much more from the relative evolutions, not from the cross-league moves during those periods (though each cross-league move is weighted just like each stay-in-the-league season-pair).

Next, the "trendline" from 1947-to-now is subtracted from this data, so its the overall deviations from this trendline that are considered to be the year-to-year quality deviations (leagues get weaker after expansion, stronger after contraction or when the talent pool gets deeper with no expansion, etc). There is no "timeline" per se, anything which might resemble such a thing over a period of time represents (in BP's opinion) quality evolution at a rate different than the modern post-WWII quality evolution. The most significant one we'll encounter is a rapid increase in quality in the NL 1925-40 (as the farm systems kick in?) and less so in the AL.
   58. jimd Posted: July 12, 2004 at 09:51 PM (#730882)
Why the AL is stronger than the NL during the 1910's.

Create 16 teams all of the same strength, and have them play a balanced schedule of 150 games; they each go 75-75 with random variation, total 1200 wins. Now, let four of them get in addition one of the four best players of this period (Johnson, Cobb, Speaker, Collins) replacing a replacement player at his position. Each of these players accumulates about 12 WARP-1 per year, year after year. The 12 teams without them now go 72-78 while those four teams go 84-66, still total 1200 wins. Take the four good teams and 4 of the other teams, call them the AL (average 78 wins). The other 8 teams, call them the NL (average 72 wins).

The AL is now 8% better than the NL, and there is no difference between them other than 4 super-stars replacing 4 replacement players. The leagues are small enough that this star imbalance makes a measurable difference.
   59. sunnyday2 Posted: July 12, 2004 at 10:14 PM (#730895)
Thanks, jimd, for the information. The method (and theory) seems reasonable enough. Except I go back to Kelly from SD's question. So what?

It seems to me that when we "discount" a player's record--and for purposes of example, let's call him Larry Doyle--when we discount his record (his WS or his WARP or whatever), we are saying that he is 6 percent (staying with the jimd's example) worse than a player with the same record, the same value, whatever, in the other league. Why?

Because he didn't have to hit against Walter Johnson?

Or call that player, er, ah, can't think of a good example of a deadball NL pitcher, a borderline HoMer. Scrolling, scrolling...OK, Hippo Vaughan. He's 6 percent worse than, say, Eddie Cicotte (I know they're not "identical"), because he didn't have to face Cobb, Speaker and Collins?

Now I also realize that BP would argue that we're not just talking about Johnson, Cobb, Speaker and Collins. By in theory, if we were just talking about these four, the difference is 6 percent. So we're talking theory here.

If less than 6 percent of Larry Doyle's AL comp are against Walter Johnson, then how do you get to a 6 percent discount? If less than 6 percent of Vaughan's PAs are "not" against Cobb, Speaker and Collins but against their bizarros, then why a 6 percent discount?

And even if Larry Doyle's bizarro does bat against Walter Johnson 6 percent of the time, how do you know Larry Doyle would really do 6 percent worse in the AL? I mean, the best players are outliers anyway. Maybe they would "rise to the occasion." I know I shouldn't say that, since I don't particularly believe in "pitching in a pinch." But there it is.

Can anybody extend jimd's discussion or provide a theoretical discussion of why an individual player should be assumed to progress or regress at the same rate as the league average? Thanks.
   60. OCF Posted: July 12, 2004 at 10:17 PM (#730897)
let four of them get in addition ... (Johnson, Cobb, Speaker, Collins)

There's a fifth player at that level, and he's in the National League: Pete Alexander. Now the AL also had Crawford, Baker, Jackson et al. How do they compare to the likes of Cravath, Doyle, Wheat? At that point it becomes hopelessly circular, because these are precisely the players we need to rank.

Of course, the 1915 Red Sox bought this young rookie southpaw from the Orioles...
   61. Kelly in SD Posted: July 12, 2004 at 10:52 PM (#730922)
Posts 58-60 bring me back to my question regarding comparing leagues and outliers. jimd described in post 58 how 4 outliers could drastically change how two leagues are compared. Is this right? Is this a valid method if a few players could warp the comparison between leagues?
Wouldn't it make more sense to compare the replacement level? But, a bottom-feeding team or 2 could pull the replacement level down below what it "should" be because they, either through choice or economics, have a good number of players who are not "replacement" level. I believe this occurred often in baseball, at least through the 1940s. There are various Philadelphia, Boston, and St.Louis teams in both leagues that would qualify.

Even comparing all players to an "average" performer runs into the same problem when there are several outliers who pull the average up or a very poor team which distorts the level of an "average" player. If there are several good outliers, it will appear as though the league "average" is a better player than historically normal. If there is a team of worse-than-replacement players or that has many worse-than-replacement players, the league could show a lower "average" player, but the league would have more players who were above-average because of that.

So, depending on you measure the quality of a league, there are a variety of ways that the actual players and teams could distort the results. If you measure by the "average" player, several outliers could distort the level of an "average" player. If you measure the quality of a league by measuring how far above or below the "average" individual players are, this can be distorted by a team of very poor players that pulls the "average" down so that players who should show as very-slightly positive to very-slightly negative all show positive.
These poor teams could pull down the replacement level if that is your measuring stick of choice.

Finally, have people considered how a dominant team could distort the quality of a league as well. For instance, does a dominant team in a league (like Boston 1912-1918 AL or Phil 1910-1914 AL or the Chi, NY, Pit group in the NL 1901-1913 or the late 20s Yankees or As) distort a league's average or replacement levels.
   62. Kelly in SD Posted: July 12, 2004 at 10:53 PM (#730923)
Sorry, I didn't finish my thought because I just go tthe mail and found out I got in to Boston Univ and UDub for LLM in Taxation. Now, my wife and I just need to decide where to go.
Any advice?
   63. DavidFoss Posted: July 12, 2004 at 10:56 PM (#730926)
Of course, the 1915 Red Sox bought this young rookie southpaw from the Orioles...

Yeah... well in 1916, the Cardinals had a young 20-year old 3B-SS with a 150 OPS+ who looks like he'll only get better once finds a place in the infield to play.

There were some all-time greats in the NL as well. The AL greats do seem to trump the NL greats when matched up head to head though.

The superstars are going to be easy to induct. Two similar borderline candidates -- one in each league -- will be interesting.

Hooper vs Carey is the only one that pops into my head at the moment. Is there a better one? Wheat vs Hooper is a better positional match, but I don't think the league difference can make up for the qualitative difference in that case.
   64. Dag Nabbit is part of the zombie horde Posted: July 12, 2004 at 11:40 PM (#730947)
Is this thread not showing up in the main menu of the HoM homepage for anyone else? I get it on the sidebar, but not with the rest of the election results.
   65. jimd Posted: July 13, 2004 at 12:03 AM (#730961)
Sunnyday2, if you don't want to discount, that's your prerogative.

That said, there's a significant difference between batting against Cobb, et al and a replacement level hitter. If Hippo pitched in the AL he could expect to face them 50-60 PA during the season, about 5% of his batters-faced. If those players were exactly twice as productive offensively as the average player, then this is worth a 5% discount on Vaughn's pitching numbers. (Cobb is more like 4-5 times as productive in BRAR as a league average player.)

Similarly for Doyle. Johnson was saving 110-120 runs per season compared to a replacement pitcher according to BP; that's about 2 runs for each full-time player he was pitching to (2runsx7teamsx8players=112). Doyle was about 40 batting runs above replacment; knock off 2 runs for facing Walter and that's a 5% discount.

(I'm not using fractions here, so I'm surprised that these numbers are anywhere close.)

There's a fifth player at that level, and he's in the National League: Pete Alexander. Now the AL also had Crawford, Baker, Jackson et al. How do they compare to the likes of Cravath, Doyle, Wheat?

OCF, if you consider Alexander to cancel out Johnson then that cuts the AL margin from 8% to 6%. If you want to start counting 2nd-tier stars then you have to name more NL'ers than AL'ers to cut into the margin. I don't see any significant surplus of NL 2nd-tier stars; it's more like a further tilt toward the AL, but that may be just reputation.

Wouldn't it make more sense to compare the replacement level?

Kelly in SD, I'm not saying what I did is how BP does it. I know they do calculations involving replacment level and all, as mentioned above by EricC. What I'm doing here is a back-of-the-envelope sanity check which says that if all else were equal between the leagues and you substituted 4 superstars for 4 replacement level regulars you get a very substantial league imbalance. And the AL had those 4 superstars.

These leagues are small, half the size of today's leagues; it would take an imbalance of twice as many players to have the same magnitude of effect today.
   66. Adam Schafer Posted: July 13, 2004 at 05:27 AM (#731313)
Was very pleased to see my #2 and #3 guys on my ballot go in last "year".

1. Mickey Welch (2) - The arguments against him just don't stack up against the arguments on his behalf. He wasn't quite as good as Keefe, but really wasn't much worse at all. I like to think of it as something like Glavine was to Maddux. Not quite as good, but would've been the #1 starter on most any other team. They pitched in the same park in the same era for too long for their extremely similiar stats to be coincidental. Welch pitched much too long for his career to be considered all luck.

2. Jake Beckley (4) - Again, I'm a career lover

3. Rube Waddell (5) - The top 5 in strikeouts for 10 consecutive years. he's #10 in the all-time ERA leaders.

4. Roger Bresnahan (6) - It's no secret that I love catchers. I would've ranked Roger higher had he caught more and played the OF less during his peak years.

5. Lip Pike (7) - I can see him finally getting in one of these days.

6. Hughie Jennings (8) - Nothing new to add

7. George Van Haltren (9) - Still just a shade above Ryan

8. Jimmy Ryan (10) - See Van Haltren

9. Bobby Carruthers (11) - I'll give him a solid spot ahead of Griffith for now. I've been back and forth on whether to place him ahead or behind of Griffith, but have decided beyond a shawdow of a doubt to go with Bobby. I hope with Bobby's election soon, that Welch will garner more of the attention that he deserves.

10. Clark Griffith (12) - I love the guy, so I hate placing him this low, but this isn't about someone being our "favorite"

11. Eddie Cicotte (13) - Underrated in my opinion. May not be HOM material, but underrated nonetheless.

12. Dickey Pearce (14) - I'm not an easy person to convince, but I read a couple ballots that really struck me and made me reconsider placing him on my ballot. I don't want to overreact, so I'm keeping him low right now, but as I sort through my feelings on him a little more, I can see myself moving him up 4 or 5 spots.

13. Hugh Duffy (15) - Back onto my ballot. No new thoughts on him

14. Jimmy Sheckard (n/a) - Not a big fan of Jimmy. I'll give him props for being the 14th best choice on my ballot. THat does say a lot.

15. Rube Foster (n/a) - I've liked Rube, not enough to put him on my ballot, but I do see him as more deserving than anyone else. This last spot was a toss up between him and Cupid Childs. It is rather disappointing seeing these guys on my ballot though. That should be changing real soon though as we get some better quality players eligible.
   67. Adam Schafer Posted: July 13, 2004 at 05:32 AM (#731318)
sorry guys, wrong thread with that post...
   68. John (You Can Call Me Grandma) Murphy Posted: July 13, 2004 at 05:33 AM (#731319)
Adam, any reason why you posted your ballot here?

:-)
   69. Adam Schafer Posted: July 13, 2004 at 05:35 AM (#731320)
yeah, it's late, i was rushing to get it in before i fall asleep, and i had multiple windows open and posted in the wrong one...there's just not enough caffeine when i work 20 hour days!!
   70. John (You Can Call Me Grandma) Murphy Posted: July 13, 2004 at 05:41 AM (#731324)
there's just not enough caffeine when i work 20 hour days!!

I hope that's a wee exaggeration (for your own sake).
   71. Adam Schafer Posted: July 13, 2004 at 05:44 AM (#731326)
i wish it was an exaggeration. friday is my only "normal" day when i put in 14 hours. working 2 full time jobs makes the hours add up in a hurry.
   72. John (You Can Call Me Grandma) Murphy Posted: July 13, 2004 at 05:47 AM (#731330)
i wish it was an exaggeration. friday is my only "normal" day when i put in 14 hours. working 2 full time jobs makes the hours add up in a hurry.

I'm amazed that you have time for this. You must be a real baseball fanatic (like the rest of us aren't) :-)
   73. DavidFoss Posted: July 18, 2004 at 04:41 PM (#740038)
Not sure if anyone has mentioned it yet, but this election makes the 1891-1895 Phillies the first All-HOM outfield. (Delahanty/Hamilton/Thompson)
   74. Michael Bass Posted: July 18, 2004 at 05:08 PM (#740057)
Good stuff!

What's the closest we've gotten to an all-HOM infield? What, if any, infield is likely to be the first?
   75. Max Parkinson Posted: July 18, 2004 at 05:33 PM (#740095)
Michael,

I have serious doubts that it happened all in the same game, but for the 1888 NY Giants, Jim O'Rourke played 4 games at 1st base, Roger Connor played 1 game at 2nd base, JM Ward was the everyday SS, and Buck Ewing played 21 games at third...
   76. DavidFoss Posted: July 18, 2004 at 05:52 PM (#740132)
Many 70's and 80's HOM-ers were quite multi-positional, so they may have been able to pull it off, but I couldn't find a team that had regulars at all four spots.

Several with 3:

1873 Bos - O'Rourke/Barnes/Wright/3B
1874 Bos - O'Rourke/Barnes/Wright/3B
1875 Bos - McVey/Barnes/Wright/3B
1876 Chi - McVey/Barnes/SS/Anson
1877 Chi - Spalding/Barnes/SS/Anson
1879 Cin - McVey/2B/Barnes/Kelly
1882 Buf - Brouthers/Richardson/SS/White
1883 Buf - Brouthers/Richardson/SS/White
1884 Buf - Brouthers/Richardson/SS/White
1885 Buf - Brouthers/Richardson/SS/White
1887 NYG - Connor/2B/Ward/Ewing
1893 NYG - Connor/Ward/SS/Davis

With 4-IF instead of 3-OF, this one is going to be much less likely.
   77. Howie Menckel Posted: July 18, 2004 at 06:10 PM (#740168)
Teams with at least four HOM hitters, any position, 10 G minimum

1873-75 Boston NA White Barnes O'Rourke Wright
1876 Chicago NL White Barnes Hines Anson McVey
1877 Chicago NL Barnes Hines Anson McVey
1879 Cincinnati NL White Barnes Kelly McVey
1879 Providence NL Hines O'Rourke Ward* Wright Start
1881-82 Providence NL Hines Ward* Wright Start
1881-84 Buffalo NL White O'Rourke Brouthers Richardson
1885-86 New York NL O'Rourke Ward Ewing Connor
1886-88 Detroit NL White Brouthers Richardson Bennett Thompson
1887-89 New York NL Gore O'Rourke Ward Ewing Connor
1890 Boston PL Kelly Brouthers Richardson Stovey
1890 New York PL Gore O'Rourke Ewing Connor
1891 Boston NL Kelly Stovey Kelley Bennett
1891 New York NL Gore O'Rourke Ewing Connor Glasscock
1892 New York NL Gore O'Rourke Ewing Richardson Keeler
1892 Philadelphia NL Connor Hamilton Delahanty Thompson
1893 New York NL Kelly Ward Connor Davis
1895 Louisville NL Brouthers Glasscock Clarke Collins
1898 Philadelphia NL Delahanty Flick Lajoie Thompson
   78. DavidFoss Posted: July 18, 2004 at 06:38 PM (#740242)
The HOM Plaque Room has falled way behind again. So much so that posts there do not show up in Hot Topics!

I recreated the Cap Standings from the existing plaques. I also extrapolated the Cap Standings to all the current inductees. Its a provisional extrapolation, though, as I'm not the one who decides who gets which cap. :-)

Should we volunteer to help Joe out by making plaques for the 1919-1929 inductees?
   79. John (You Can Call Me Grandma) Murphy Posted: July 18, 2004 at 09:35 PM (#740502)
Should we volunteer to help Joe out by making plaques for the 1919-1929 inductees?

No need to, David. He has all of them up to 1929 (and including this year :-) from me. Joe was planning on giving me the "keys" to the Plaque Room so as to help him out due to his busy schedule (I'm not sure if he has time for a bathroom break). Hopefully, we can get the Plaque Room up-to-date again.
   80. DavidFoss Posted: July 18, 2004 at 10:41 PM (#740577)
Great, thanks...

Yeah, I saw from reading Gleeman that Joe is in Cincy for the SABR meeting.
   81. John (You Can Call Me Grandma) Murphy Posted: July 18, 2004 at 11:02 PM (#740608)
Yeah, I saw from reading Gleeman that Joe is in Cincy for the SABR meeting.

I wish I was.
   82. Joey Numbaz (Scruff) Posted: July 19, 2004 at 03:45 AM (#740926)
I just sent Jim the note John, hopefully you get the access quickly . . . thanks for the help!

--Joe
   83. John (You Can Call Me Grandma) Murphy Posted: July 19, 2004 at 03:56 AM (#740932)
I just sent Jim the note John, hopefully you get the access quickly . . . thanks for the help!

Thanks a lot, Joe! As soon as Jim allows access, I'll have the Plaque Room up-to-date in fairly short order. Then you'll be able to make those bathroom pit stops again. :-)

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
greenback calls it soccer
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.1687 seconds
49 querie(s) executed