I voted for Creighton once and got roundly bashed for my trouble. Ed Williamson is still on my ballot, which is fairly unique. Browning and C. Jones, which is not. Williamson is clearly one of the top 2 3Bs of the 19C, I preferred him to Ezra, others did not, but worst case he is #2. We have, of course, a bunch of part-time 3Bs from thence, but Ed would be the second 19C 3B specialist if he were elected. which of course he won't be.

fox is right, of course. Doc C did a 1941 MLE for Pesky but not for the purpose of additional MLE credit but only to fine-tune the WWII credit.

Dan R.,

The leading 1880s/1870s positional player candidates are, I think,

Pete Browning, Charley Jones, Fred Dunlap, Ed Williamson, Tom York, and Denny Lyons

Browning and Jones have serious shots at election

Dunlap and Williamson still get a few votes

York and Lyons were probably the last two other pre-1990 hitters to get votes.


The leading 1880s/1870s pitcher candidates are, I think

Mickey Welch, Tony Mullane, Tommy Bond, Jim McCormick, Jim Whitney

Welch has a core set of supporters

Mullane gets a vote or two

Bond got a vote or two until fairly recently

McCormick has enjoyed significant support

Whitney had some strong supporters back in the day.

If you are doing a full workup of pre-1893 baseball in your system, I for one will be very interested in seeing the result, both in terms of what your SD analysis finds re the AA, and your overall assessment.

Hahah, I forgot about Jones and Browning, I vote for them myself. I'll look into Dunlap, Williamson, York, and Lyons.

The problem with doing pre-1893 in my system is run estimation. BaseRuns still gives me good results for the 1890's (though still missing on teams like Boston) but going back into the 1880's it gets into *a lot* of trouble...I've tried to modify it, adding in big reached on error estimates and whatnot; it works for some league-seasons but not others. BP's EQR doesn't seem to do much better, at least as far as I can see (which should cast doubt on its WARP scores for the early game, since they obviously rely on EQR's accuracy). If anyone has a run estimator that reliably estimates team run scoring even within 5% back through 1871, I'd *love* to get my hands on it, and could then do my WARP back in time.

RE: Pesky and MLEs. I forget which voter did it, but they were using a 1941 MLE to create an additional data point to better estimate WWII credit for Pesky.

Right, I ran an MLE for his year in AAA in 1941. It raised his War credit scenario a bit, but it didn't bring him up to electable standards (in my opinion).

Ah, OK, but no one's saying Pesky should actually be credited for it at the major league level. If I credit Pesky for 43-44-45 with the average of his 42 and 46 performance, he'd probably be #1 or #2 on my ballot for next year. (Again, now that I have full AL WARP, I'm re-doing my ballot from scratch). Is there any consensus in the group on how much credit to give him?

Ah, OK, but no one's saying Pesky should actually be credited for it at the major league level. If I credit Pesky for 43-44-45 with the average of his 42 and 46 performance, he'd probably be #1 or #2 on my ballot for next year. (Again, now that I have full AL WARP, I'm re-doing my ballot from scratch). Is there any consensus in the group on how much credit to give him?

You're correct, Dan. My war policy had been to average the four surrounding seasons, which would have been 41, 42, 46, 47. But Pesky was in AAA in 41, so I had been giving him a zero for that year, which had lowered his war credit somewhat. So I did the MLE so that I could offer more realistic war credit by averaging those four years instead of three and zero.

Can you post your translated stats for him in 1941?

Weird little item with neither context nor much outside-my-head-interest. I've been tinkering with my system these past couple days, wondering why certain players break it and others don't. Anyway, it should lead to some changes on the next ballot, including the removal of a couple teddy bears. In general it's mostly the same guys in the same rankings, just moving around a couple-few slots. As I looked position by position, I noticed something interesting. At 1B I have the most possible agreement with the group I can. I currently figure the in/out line as generally the 17th-best player at each position. All of my top 17 1Bs have made it in. And Sisler just barely misses to boot. Granted everywhere else it's 2-5 guys worth of disagreement, but at least for once, in one little corner of the world, Dr. C can stand for Dr. Consensus.

Dan, this one's from the 1995 results thread. Feel free to tweak as needed! I used a .9 converter (on runs), but if you felt .87 or .88 was more appropriate, then you'd probably be right. I erred upward figuring that the high MiL leagues may not have been 100% farm systems at that time.

+++++++

here's Pesky's full AAA line.

146 G (of 154)
600 ab
93 r
195 h
48 rbi
25 2b
5 3b
1 hr
45 bb
37 k
16 sb
.325 avg

He played all 146 games at SS.

this is from marshall wright's book on the AA.

I've got him at ~92 RC and 410 outs for a raw 6.06 RC/G.

The louisville colenols, his team, finished 2nd, 87-66. They scored 682 runs, 4.43 per game, below average for the league. They were second to last in HR in the leage. The league scored 4.62 r/g. Wright gives no ER or RA totals, but estimating based on innings and ERAs he gives, the team was probably giving up around 3.70 ER per game. Multiply by 1.10 to be safe, and you can estimate them giving up about 4.07 RA/9. OK, now add the est RA + RS, divided by twice the league's RS/9, then apply a heavy regressive weighting (25% on the season, 75% on the mean), and you get a .98 pf.

OK, so we can run through a very quick WS MLE of Pesky's 1941 season (since that's what I use to determine credit).

6.06 RC/G / .98 = 6.18 pkadj RC/G

6.18 / 4.62 = 1.34 ratio of player to league.

1.34 * 4.50 = 6.03 R/G in neutral league

6.03 * .90 = 5.43 R/G in Major League, neutral setting

5.43 / 4.5 = 1.21 Ratio of player to neutral MLB league.

1.21 * 4.23 = 5.12 Pesky's RC/G in 1941 AL setting in neutral park.

If he uses the same number of outs as he did in the AA, then 78 RC.

On an average AL team of 1941, he would rack of 15.6 BWS (per my own system of figuring it, not SFWS).

In the field, Pesky was very effective, says WS, 6.02/1000 innings, or ~111 games. Pesky played 146 games in 1941, crediting him at his career rate means 7.9 FWS for 1941.

So 15.6 + 7.9 = 23.5 MLE WS for 1941. Feel free to quibble, of course.

What difference does it make if I use three or four surrounding years?


1941 1942 1946 1947 = credit
-------------------------------
0 28 34 25 21.8
23.5 28 34 25 27.6


What's that do? First thing it takes him from 258 to 275 career WS. When I plug the results into my system, Pesky vaults from 7 pts to 15 pts. That's a lot but instead of putting him with Rizzuto, it pushes him into the territory of Maury Wills, Travis Jackson, Ed McKean. That's not implausible to me. His most positive comp would be Boudreau, though he's not quite up to boudreau's standard in my system, mostly because I don't give All-Star and MVP level credits for war credit seasons.

I figure 25 WS a year is plausible. You could argue for more. The guy could play, but like some others never quite hit the heights after the war that he did before. And like ARod, his move to 3B doesn't help. But once you plug 25 WS per year for the war years into the equation, on a 10 year prime basis the guy is a real candidate. If a 10 year prime is what you're looking for. Still, he is in my top 30 but no higher. Better than Cecil Travis, though. The years Cecil lost to the war are probably not even prime years, while Pesky's are peak years.

lso, are there any 1870's/1880's position player candidates that people are still supporting, or are Beckley and Van Haltren the earliest guys still showing up on ballots? Did anyone ever vote for Jim Creighton, for that matter, or did he simply die too soon to be a candidate?

Not enough to place on my ballott, but I have Fred Carroll entrenched in my consideration set. But I probably give him plenty more credit than others.

Nash is making the bottom of my ballot after years off it, but I don't consider him someone we desperately need in the HoM.

The Creighton candidacy was brought up before. He only pitched for four years (1859-62) and was dead at age 21. The best pitcher of his day, but just not enough of a career for even the peakiest of voters.

Prelim 1997 (elect 3)

Jim Creighton is #140. PHoMers are Dewey Evans, Keith Hernandez and Don Newcombe, though that is prelim. Perez, Leach, Trouppe, Cash, Quiz and Parker are next. Perez was actually in line for the 3rd spot but I moved Newk ahead just because I'm not knocked out over Doggie. Maybe Leach is better than Newk. They all have their flaws.

1. Edd Roush (1-2-2, PHoM 1976)—nice peak of 38*-33-30 (*short WWI season adjusted to 154), very well-rounded skills

2. Rollie Fingers (3-4-6, PHoM 1991)—there’s no uber-stat that says Fingers is ballot-worthy, but I go back to Chris Cobb’s old test—who do you want in the HoM? And on that simple basis, subjective as it is, I want the #3 reliever of all-time (as of 1995) in mine

3. Pete Browning (4-6-15, PHoM 1961)—do you believe in numbers?

4. Addie Joss (5-5-8, PHoM 1967)—not a workhorse, to be sure, but damn effective and not a “small sample” of IP

5. Dewey Evans (new, PHoM 1997)—who’da thought he’d end up better ‘n Fred and Jim? But there it is

6. Charley Jones (8-10-17, PHoM 1921)—even with hefty AA discounts

7. Orlando Cepeda (6-8-9, PHoM 1987)—pretty interchangeable with F. Howard, Cravath and (now) Tony Perez, but the best of the group

8. Ed Williamson (11-12-14, PHoM 1924)—more peak and more glove than than almost any other IF

9. Phil Rizzuto (9-7-10, PHoM 1995)—306 career WS with war credit. His tenth best season is 21 WS, compared to Nellie Fox at 19, Darrell Evans at 18, e.g.

10. Frank Howard (12-14-23, PHoM 1987)—PHoM opening forced me to re-evaluate a lot of guys, and his numbers are just too good

11. Larry Doyle (10-11-5, PHoM 1975)—same OPS+ as Ed Roush

12. Reggie Smith (7-9-13, PHoM 1988)—cannot quite see how he’s not better than Jim Wynn

13. Dick Redding (16-16-11, PHoM 1971)—next best arm, great peak

14. Nellie Fox (13-13-12, PHoM 1971)—I’ve decided I like Rizzuto a bit better, though they are two very very comparable players. I think Phil’s peak was a bit more valuable

15. Elston Howard (15-15-7, PHoM 1994)—never really thought of him as a HoM or HoF or PHoM type of player, but I now see him as one of those few players whose opportunities were least commensurate to his ability

Close—i.e. right around in/out line, as I think we will elect another 10 or so backloggers before we’re done

16. Gavvy Cravath (14-21-21, PHoM 1995)
(16a. Keith Hernandez [17-new, PHoM 1997])
17. Don Newcombe (19-20-22, PHoM 1997)
18. Tony Perez (18-19-19)
19. Tommy Leach (20-21-16)
(19a. Quincy Trouppe [20a-22-50])
20. Norm Cash (21-21-43)

21. Dan Quisenberry (22-new)
22. Dave Parker (new)
23. Al Rosen (23-24-18)
24. Hack Wilson (24-25-28)
25. Burleigh Grimes (25-26-38)

HoVVG

26. Roger Bresnahan (26-27-31-)
(26a. Jim Wynn [27-28-52])
27. Sal Bando (28-29-29)
28. Johnny Pesky (29-30-80)
29. Eddie Cicotte (30-31-24)
30. Ken Singleton (31-32-71)
(30a. Jim Bunning [30a-31a-30b])

31. Bruce Sutter (32-33-44)
(31a. Joe Sewell [32a-33a-30a])
32. Alejandro Oms (33-34-40)
33. Bob Elliott (35-35-57)
34. Fred Lynn (34-new)
(34a. Ken Boyer [34a-33b-22a])
(34b. Don Sutton [35a-35a-20])
35. Luke Easter (40-40-25)
36. Lefty Gomez (36-36-32)
37. Luis Tiant (37-37-48)
38. Thurman Munson (38-38-37)
39. Chuck Klein (39-39-30)
40. Vern Stephens (41-41-33)

41. Fred Dunlap (49-49-59)
42. Bill Monroe (42-42-46)
43. Bobby Bonds (43-43-45)
(43a. Jimmy Sheckard [43a-43a-45a])
44. Bus Clarkson (44-44-82)
45. Cesar Cedeno (45-45-49)
46. Wally Berger (46-46-47)
47. Tommy Bridges (47-47-58)
48. Bucky Walters (48-48-27)
49. Graig Nettles (50-50-53)
(49a. Cool Papa Bell [49a-49a-49a])

50. Dick Lundy (55-55-26)
51. Pie Traynor (51-51-39)
52. Bob Johnson (52-52-60)
53. Luis Tiant (53-53-new)
(53a. Wes Ferrell [53a-53a-51a])
54. Hilton Smith (54-54-51)
55. Dizzy Dean (56-56-33)
56. Dave Bancroft (57-57-35)
57. Tommy Bond (58-58-34, PHoM 1929)
58. Hugh Duffy (61-61-62)
59. Jake Beckley (59-59-54)
60. Vic Willis (60-60-63)

HoVG

61. Bobby Estalella (67-67-42)
62. Gene Tenace (62-62-75)
63. Jim Rice (63-63-new)
64. Ben Taylor (64-64-73)
(64a. Biz Mackey [64a-64a-40a])
65. John McGraw (65-65-64)
66. Frank Chance (66-66-69)
67. Tony Oliva (68-68-76)
68. Vada Pinson (72-72-78)
(68a. Joe Kelley [68a-68a-64a])
(68b. Pete Hill [68b-68b-81a])
69. Lou Brock (69-69-65)
70. Rocky Colavito (70-70-56)

71. Mickey Welch (71-71-70)
72. Marvin Williams (74-74-68-44)
73. Wilbur Cooper (73-73-67)
74. Jim McCormick (75-75-83)
75. John Clapp (84-84-86)
(75a. Billy Pierce (75a-75a-70b)
76. Dave Concepcion (76-76-41)
(76a. Early Wynn [76a-76a-70a)
77. George Van Haltren (77-77-74)
78. Rusty Staub (78-78-71)
79. Jimmy Ryan (79-79-79)
80. Jim Fregosi (80-80-84)

81. Tommy John (81-81-new)
82. George Burns (82-82-81)
83. Tony Mullane (83-83-72)
84. Dolf Luque (92-92-HM)
85. Ron Cey (85-85-61)
86. Ernie Lombardi (86-86-66)
87. Herman Long (87-87-NR)
88. Urban Shocker (88-88-85)
89. Bert Campaneris (89-89-90)
90. Boog Powell (90-90-NR)

91. Red Schoendienst (91-91-92)
92. Mike Tiernan (99-99-HM)
93. Luis Aparicio (93-93-55)
94. Tony Lazzeri (94-94-88)
95. Artie Wilson (95-95-94)
96. Bobby Veach (96-96-89)
97. Wally Schang (97-97-97)
98. Maury Wills (98-98-77)
99. Bobby Murcer (110-110-HM)
100. Buddy Bell (100-100-new)

Hall of Pretty Good

101. Kiki Cuyler (101-101-79)
102. Carl Mays (102-102-92)
103. Jim Kaat (103-103-98)
104. Ellis Kinder (104-104-HM)
105. Steve Garvey (105-105-HM)
106. Bobby Avila (121-121-NR)
107. Rabbit Maranville (107-107-86)
108. Joe Tinker (108-108-HM)
109. Johnny Evers (109-109-HM)

110. Andy Cooper (117-117-HM)
111. Gil Hodges (111-111-HM)
112. Spot Poles (112-112-HM)
113. Al Oliver (113-113-HM)
114. Cecil Travis (114-114-NR)
115. Billy Nash (115-115-NR)
116. Mickey Vernon (116-116-HM)
117. Lon Warneke (126-126-HM)
118. Bill Byrd (118-118-HM)
119. Sol White (119-119-HM)
120. Pancho Coimbre (120-120-HM)

121. Amos Otis (127-127-HM)
122. Catfish Hunter (122-122-HM)
123. Silvio Garcia (123-123-HM)
124. Hippo Vaughan (124-124-HM)
125. Jake Fournier (125-125-HM)
126. Dave Orr (140-140-HM)
127. Roy White (128-128-NR)
128. Roger Maris (141-141-HM)
129. Lave Cross (129-129-HM)
130. Jose Cruz (130-130-HM)

131. George Scales (131-131-HM)
132. Sam Rice (132-132-HM)
133. Denny Lyons (133-133-HM)
134. Vida Blue (134-134-HM)
(134a. Pud Galvin [134a-134a-NR])
135. Silver King (135-135-HM)
136. Fielder Jones (136-136-HM)
137. Davey Lopes (137-137-HM)
138. Jim Whitney (138-138-HM)
139. Virgil Trucks (139-139-HM)

140. Jim Creighton (145-145-NR)
141. Leroy Matlock (148-150-NR)
142. Sparky Lyle (142-142-HM)
143. Ray Dandridge (143-143-HM)
144. Darrell Porter (144-144-new)
145. Mickey Lolich (146-146-HM)
146. Ron Guidry (147-148-new)
147. Frank White (149-new)
148. Ken Griffey, Sr. (new)
149. Dick McAuliffe (150-147-NR)
150. Toby Harrah (NR-149-new)

1997 prelim

Really getting down to the teddy bears here...may be getting a little heavy on the really old timers at this point. Perhaps I'm being too conservative with the borderline newcomers? Will examine. Right now, Dewey Evans is at #23, Parker at #28, and none of the other newbies are even close.

1. Dizzy Dean
2. Ed Williamson
3. Vic Willis
4. Elston Howard
5. Al Rosen
6. Pete Browning
7. Gavvy Cravath
8. Luis Tiant
9. Charley Jones
10. Nellie Fox
11. Eddie Cicotte
12. Edd Roush
13. Hugh Duffy
14. Larry Doyle
15. John McGraw

Is it known what Creighton died of?

According to his SABR bio page at http://bioproj.sabr.org/bioproj.cfm?a=v&v=l&pid=16900&bid=770, Creighton died of a ruptured inguinal hernia.

Ouch.

Who's this Joseph Leggett, who according to that article was the best player of Creighton's day?

pre-1893:
Beside the viable Browning and Jones, something notable about the pre-1893 (60'6") survivors is that Bond, Welch, Dunlap and Williamson have all appeared at or near the top of someone's ballot. Tom York enjoyed a strong supporter 80-100 years ago who has since tempered his advocacy of anyone/everyone. Not only Fred Carroll but Duke Farrell and John Clapp garnered a vote here or there as findings during the early stages of the neverending search for meritorious catchers.

Welch may be the most interesting case, not to say strongest. Electorally, I guess that he has some latent supporters who have given him up as a lost cause. What's interesting is that he worked mainly as a teammate of Tim Keefe in Troy and New York. Keefe was elected rather quickly with a near-consensus that he was clearly better than Welch, call the margin "X". Subsequent research by (?)Chris James on (?)quality of opposing teams or offenses persuaded many readers to re-envision that margin as something like X/10 or X/3. Several people made remarks such as "I'm not sure Keefe was better than Welch" --of course, they didn't all become Welch voters. Jim Galvin and Early Wynn and Don Sutton suffered some punishing criticism but Welch is the one 300-g-w in the foyer.

A problem (analytical problem, at least) with the 1880s National League is the sometimes gross imbalance. Chicago fielded an all-star team for most of the decade, a la Boston in the 1870s. Successive rivals Providence, Detroit and New York were all-star teams or nearly so for a season or four. The league never produced .850 at the top and .150 at the bottom, so the average quality of opposition (measured strictly within league by WL%) never differed by +/-.100 from top to bottom. At worst or "best" Baltimore, Boston, and Cleveland in the 1890s achieved similar margins over Louisville, St Louis, and Washington only thanks to expansion, top to bottom in a 12-team rather than 8-team league.

As far as I know, Joseph Leggett is one of the amateur stars still "lost": we have no birth or death data. Because he never played in the majors from 1871, there is no organized effort to track him down, as there is for Charley Jones, the greatest player with time and place of death or burial unknown as of Fall 2006.

In the SABR biography (Baseball's First Stars, 1996), Bob Tiemann calls Leggett 'Quickly acknowledged as "the master of his position". He was the incumbent catcher and captain when the Excelsiors acquired Creighton, so he secured the famous fastballs at least as their on-field receiver if not also as a negotiator. (Creighton was probably the first professional baseball player in some sense.) "The Excelsiors were especially hard hit by the call to military service. Leggett apparently did not join the army but he played sparingly after 1860." Tiemann shows him playing in 0-5-5-2-1-14 matches, 1861-66, finis. In 1856 (last season of the Knickerbocker era) he played for the Wayne club in Brooklyn and he joined Excelsior late in 1857 scoring eight runs in his one match, a 41-23 win. The list of known 1856-and-1866 players is short and Leggett is on it.

In contrast we know that Creighton died in his father's home, at a known street address in Brooklyn, still standing iirc. It will be a nice 150th birthday present for "organized baseball" if someone finds Joseph Leggett alone and some family ID frosting on the cake.

By the way, I don't believe we know who was the best player, except that by 1862 Creighton was obviously the best player. Leggett and Creighton's Excelsior club played zero matches in 1861 which is not to say zero games. Leggett narrowly lost the NABBP presidential election at the annual meeting that December --Tiemann. Personally, I suspect that both ballplayers and writers (Henry Chadwick) gave the game less attention in 1861, then resumed life as usual, to some extent, when they realized that the war would not be short. NYC was not a war zone, of course. Nor was it a center of abolitionist thought or sentiment, or a safe place for escaped slaves. It was or became a stronghold of anti-war politics and activism. . . . blah blah this doesn't help learn what Leggett or his family was doing. For what it's worth, I guess he could buy his way out of military service if it came to that.

He was the incumbent catcher and captain when the Excelsiors acquired Creighton

He'll need a very big catcher bonus at this point to get onto anyone's ballot. ; )

Time for my 1997 prelim. Barring a change of mind, three bats make my PHOM: Veach, Stargell and Cepeda. Evans makes the tail end of my ballot (I have never consciously given a hedge vote.) Dave Parker remains in my consideration set, but way deep (12 among RFs - somewhere around 125 if I ordered them that far). None of the other newcomers makes my consideration set, although I suppose Templeton could feasibly be tacked on to my shortstop queue around 20 deep.

1)Hugh Duffy (PHOM)
2)Ben Taylor (PHOM)
3)Edd Roush (PHOM)
4)Nellie Fox (PHOM)
5)Tommy Bridges (PHOM)
6)Lefty Gomez (PHOM)
7)Charley Jones (PHOM)
8)Vern Stephens (PHOM)
9)Gavvy Cravath (PHOM)
10)Bob Johnson (PHOM)
11)Bobby Veach (PHOM)
((11a)Willie Stargell)) (PHOM)
12)Orlando Cepeda (PHOM)
((12a)Ken Boyer))
13)Al Oliver
14)Tony Oliva
15)Dwight Evans
16)Jim Rice
17)Wally Berger
18)Dizzy Dean
19)Bus Clarkson
((19a)Darrell Evans))
20)Dan Quisenberry
21)Bruce Sutter
22)Ernie Lombardi
((22a)Jimmy Wynn))
23)Alejandro Oms
24)Reggie Smith
25)Ron Guidry
26)Al Rosen
27)Mickey Welch
((27a)Jim Bunning))
((27b)Billy Pierce))
28)Sparky Lyle
29)Dick Redding (PHOM)
30)Ron Cey

Calling out for a little help as I review Marv Williams's MLEs!

I'm looking for particular information about three lower minor leagues of the early 1950s.

The leagues:

-Gulf Coast League (1953)---Williams played at Laredo

-Northwest League (1954)---Williams played Vancouver

-Sally Leauge (1955)---Williams played in Columbia

What I need
1) Schedule length (better yet individual team records or games totals!!!)
2) A sense of whether the league was a hitters, pitchers, normal league for scoring (r/g even better!!!)
3) This is the optional item---A sense of how hitter/pitcher friendly Williams's parks were.

If anyone with one of them old guides that list the MiL standings and stuff for each year can help, I'd be much obliged. Thanks!!!!

-Northwest League (1954)---Williams played Vancouver

Are you sure about this one? I show the Northwest League being dormat after 1947 until April 26, 1955, and Vancouver being in the Western International League in 1954?

You're right KJ, I meant to put that. The WIL either morphed into the NWL or the Vancouver team of the WIL entered the NWL. Lester/Clark have it listed as the NWL, which is where I was pulling the name from.

Thanks for pointing that out. So

-Western International League (1954)---Williams played for Vancouver

Among the "HOF, not HOM" who played mainly before 1960, eleven are not among the worst Hall of Famers according to any of the critics our generations have produced, as far as I know:
Welch, Duffy, Beckley, Joss, Roush, Bancroft, Grimes, Traynor, Cuyler, Dean, Gomez

Three stand out for flying under the radar here. In my opinion, Kiki Cuyler is simply easy to evaluate for our purpose: not bad, but everyone sees at a glance that the Meritorious line must be drawn above him. That leaves Dave Bancroft and Pie Traynor. Bancroft has the better case, and a good one I believe, but I am sure that Traynor too has been dismissed by many voters without getting as much attention as he deserves, measured in words.

Is Bancroft hurt by the drop of his OPS+ specifically below "100" at the end? (dropping from 105 in twelve full seasons to 98 in fifteen) Do modern metrics, such as they are without benefit of pbp data, reveal that he was not an outstanding fielder? Does anyone find "if Dave Bancroft, then Travis Jackson" compelling? (OPS+ 107 in twelve less than full seasons, must under 6000 pa to Bancroft's just under 7000).

I support Beckley, Roush, Bancroft as meeting HoM standards.

Duffy, Grimes, Traynor, Cuyler and Dean are just below my cutoff with Welch, Joss and Gomez another notch below. I'd have to say I think Tinker, Evers and Chance should be in this conversation also. Also John McGraw, Vic Willis, Hilton Smith, Roger Bresnahan - I wouldn't throw any of those guys out of the current HoF.

One of the interesting things about this project is how I've noticed that the HoF is shrinking over time. Each successive generation it seems to be more difficult to get in.

Among the "HOF, not HOM" who played mainly before 1960, eleven are not among the worst Hall of Famers according to any of the critics our generations have produced, as far as I know:
Welch, Duffy, Beckley, Joss, Roush, Bancroft, Grimes, Traynor, Cuyler, Dean, Gomez

Hey! That looks like my ballot. Actually, I only have four of those guys on my ballot, but four more are close and Cuyler is the only one out of my top 50.

Hilton Smith is one of those guys that gives me that un-funny funny feeling....

You're right KJ, I meant to put that. The WIL either morphed into the NWL or the Vancouver team of the WIL entered the NWL. Lester/Clark have it listed as the NWL, which is where I was pulling the name from.

Thanks for pointing that out. So

-Western International League (1954)---Williams played for Vancouver

To answer your question - unfortunately, none of these teams have any game or other data in my database, which makes me think they probably are NOT listed in the standard guides - you'll probably need to use newspaperarchive.com or something similar to find them.

Excuse me, I didn't include the Negro League members (but I don't recall any of them named among the worst Hall of Famers, essentially, except in this forum).
Tinker, Evers, Chance, and Bresnahan have all been named, I think, honored for the poem or the shinguards.

--
Ranked by Bill James among the 12 best at fielding position or the 36 best pitchers:
Dizzy Dean, Dale Murphy, Sal Bando
After election of Jim Wynn, that's all until Don Mattingly becomes eligible.

Ranked by Bill James among the 15 best or 45 best pitchers: in these ranks 13-15 and 37-45 about half are not in the HOM including five who played before WWII, one under the radar here
Wally Berger, Edd Roush, Pie Traynor, Carl Mays, and Lon Warneke
The others: EHoward, Munson, Perez, Fox, Aparicio, Fregosi, Rosen, Nettles, Brock, Bonds, Parker

--
Dave Bancroft, WARP1 111
How many in the foyer does Clay Davenport rank higher?

--
TomH #23 called for rotten tomatoes.
Maybe I should have written this in second person, but let it be.

McGraw, Chance, Bresnahan, Monroe, Elliott, and Walters on the ballot (top 15 backlog) constitute a warm welcome for heavy hitters with playing time issues. One rotten tomato for Charley Jones down around number 50 with Easter and Fregosi. It isn't that too many outfielders have been elected, for Reggie Smith and Bobby Bonds are around number 15 and Bob Johnson is in the top ten. (The rest of TomH's not-ballot 15 is four pitchers and three longevity cases--Beckley Evans and GVH.)

Fox is in the top twenty, the fourth "throwing infielder" after McGraw, Monroe, Elliott. The next ten include Concepcion, Traynor, Nettles, and Leach; the next ten Cey, Rizzuto, Bando and Aparicio (about 40). Aparicio like Fox above and Maranville just below is a traditional extreme longevity/durability candidate. Cey and Bando are heavy hitters akin to Elliott. Bancroft fits well among the others by type: merely good batters, given fielding position (if Rizzuto qualifies, thanks to a generous view of WWII); merely full-length careers; excellent in the field by reputation (if Traynor qualifies). So how does Bancroft not rank at least high 30ish with Leach? One rotten tomato for leaving him out of this company and a second for leaving him out of the fifty-man backlog you have listed.

Maranville, Doyle, and Fregosi seem to be in the forties. Given the values revealed in the whole table, I will spare a fourth rotten tomato for placing Laughing Larry so low.

Eric - Just ignore that post - I found the data...I'm about to head out of town in just a second to have no internet access for the weekend, so I'll email you all the data I have when I return..

Thanks, Kevin!

Someone mentioned Johnny Evers and Joe Tinker favorably.
Reading in the 20-year-old BJHBA that Evers was a greater offensive force than Frisch, Lazzeri, Herman, Doerr, and Fox, I gleaned this from the entries for 2Bmen.

Offensive Winning Percentage
784 Hornsby
742 Collins
711 Morgan
703 Lajoie
700 Robinson
665 Carew
653 DOYLE
646 Gehringer
628 Grich
619 Gordon
602 EVERS
601 Herman
593 PRATT
591 Frisch
588 HUGGINS
582 Lazzeri
578 Doerr
There is no entry for BUDDY MYER, who is covered with Billy Herman as two of the best matched players ever. I suppose he is in that group.
531 Fox
523 Randolph
521 Schoendienst
439 Maz
415 Frank White

DanR previously asked about pre-1893 candidates. And he noted #103
The problem with doing pre-1893 in my system is run estimation. BaseRuns still gives me good results for the 1890's (though still missing on teams like Boston) but going back into the 1880's it gets into *a lot* of trouble... I've tried to modify it, adding in big reached on error estimates and whatnot; it works for some league-seasons but not others. BP's EQR doesn't seem to do much better, at least as far as I can see (which should cast doubt on its WARP scores for the early game, since they obviously rely on EQR's accuracy). If anyone has a run estimator that reliably estimates team run scoring even within 5% back through 1871, I'd *love* to get my hands on it, and could then do my WARP back in time.

DanR,
Beside the players of the 1870s/80s whom you have not rated at all, our understanding of several 1890s players might change significantly given a significantly different systematic treatment of pre-1893, simply because their major league careers began before 1893. Beckley, Duffy, GVH, and Ryan have a lot of ink here; believe it or not, they once had about equal numbers of votes. Long and Nash (left side of the Boston infield before Collins), who are now flying with Eric Chalek and John Murphy. For anyone whose prime years straddle 1893 you may be able to say with some confidence, based on 1893 and after, that he is a likely or unlikely subject of notable revision when you are able to do his entire career.

The player superficially most like Dave Concepcion is Lave Cross. He played a long time. Maybe he was "historically great" at 3B (only then is he Meritorious). Contemporary opinion suggests that that is plausible. Both WARP and WinShares seem to say yes. If everyone here believed it, several would be voting for him, because he was a league-average career batter, unlike Maranville and Mazeroski. (Every career voter who believes it should be voting for him. Maybe every one is!) I say no more than "superficially" like Concepcion because the short career big bats at 3B in his time --John McGraw and Bill Joyce, at least-- may yield a rating for Cross very different from Concepcion's.

Might someone come from nowhere --beyond Lave Cross-- and be a plausible HOM candidate during the next decade, by some combination of advances in data-gathering and digitization, analysis and persuasion? Maybe a longtime catcher, because voters are inclined to award some extra "catcher credit" and ratings of catcher defense are especially immature. Deacon McGuire is the obvious candidate, although I can imagine it for a strong batter who didn't catch as much, such as Jack Clements. The problem is, McGuire needs to be evaluated as an excellent catcher, Clements as a very good one, in order to get going. Either would be a major re-evaluation which must be considered unlikely.

Was Mike Griffin "historically great" in CF? Contemporary opinion leaves that open, I think, but WARP and WinShares seem to say excellent and very good respectively. (In my opinion, he Merits some support from prime voters if he was merely excellent but others disagree.) Anyone who has never viewed his record should do it. It looks best at Baseball Prospectus; he should be Clay Davenport's favorite forgotten man.

Paul, thanks very much for your feedback. I'm trying to estimate salaries for pre-1893 guys just by using BRAA + FRAA / 9 and 1893 replacement levels...Cross definitely falls short for me. Concepción played shortstop in an era where both replacement level for infielders and standard deviations were exceedingly low; Cross played 3B (when it was more like 2B, but still not close to SS) when StDev's were really high.

I'll give Griffin a look.

Yesterday I wrote:
Dave Bancroft, WARP1 111
How many in the foyer does Clay Davenport rank higher?

in the foyer - that is, eligible for the HOM but not elected.
A few. More than I expected. This group hasn't remotely elected everyone with WARP1 rating 100 (as someone said casually) and I found six rated above Bancroft. I think I checked every non-pitcher in 1996 Ballot Results and every "throwing infielder" in the Hall of Fame plus a few other candidates for WARP1 love(*).

WARP1
133.5 Maranville
116.8 BECKLEY
116.7 Leach
116.5 Cross
115.2 Tinker (including one big FL season)
113.0 PEREZ
111.6 Bancroft

*at least Bartell and Myer (Bancroft and Herman twins, Bill James once thought), Gardner, Cross, Latham, Ritchey, and Pfeffer (superslick fielders in some opinions).

Hi, DanR.
Regarding your system, my eye was not on the standard deviation ball but on some others. I presume that standard deviations were really high in the 1870s and 1880s, although your particular variety may yet be impossible to measure.
By the way, I don't recall that you have anything new on catchers in particular, even with pbp data. But I am really pontificating for everyone's benefit :-!

Worst player with 100 BP WARP1 has got to be Willie Davis.

I imagine stdevs in the 70s and 80s were insanely high. But to measure stdevs I need wins above average, to measure wins I need runs, and to measure runs I need a ####### run estimator that is remotely accurate for those leagues.

I do have catchers in my spreadsheet...I now have all position player seasons with over 50 PA since 1893 in both leagues. Defense-wise I just use FRAA and Fielding WS for C, with a rather tight stdev for their defense as per Chris Dial's 2006 numbers.

Dan,

Joe D and I were talking about the 19th C. run estimator issue. You just can't get the runs to add up due to the influence of all those errors. I ended up concurring with him that if base runs couldn't do it, then the last resort was to apportion the leftover runs to all batters in equal proportion in some manner. He mentioned that the Stats All-Time Handbook figures RC this way, so that could be a start for you.

58. Phil Hughes a Condom ('zop) Posted: March 27, 2007 at 04:05 PM (#2319085)
DL, the "smell test" is not a satisfactory measure of statistical significance.

Lets take the following top 100 ballot distribution:
12 . SS
10 . 2B
10 . C
_3 . 3B
13 . LF
14 . CF
12 . RF
12 . 1B
14 . P

That 3 really sticks out in this ballot, right? Doesn't seem to "pass the smell test". But the standard deviation of players-per-position for this ballot is nearly 3.5, so 3 is only 2 standard deviation below the mean

60. DL from MN Posted: March 27, 2007 at 04:33 PM (#2319107)
> the "smell test" is not a satisfactory measure of statistical significance

Nothing else in my spreadsheet meets 95% statistical significance either. 2 stdev below the mean is pretty bad when it's only one position. I've never set up my spreadsheet to be mathematically consistent, just reasonable.

The normal approximation isn't very good here. Setting aside the 14 pitchers, I investigated the uniformly random allocation of 86 position players to 8 positions. And I found that the minimum count for the 8 positions is so low as three only 3% of the time.

By the way, the minimum is so high as nine only 4% of the time, so an allocation such as
. 9 10 10 10 11 11 11 12 12 .
"doesn't pass the smell test" either, in the same 95% sense DL and PC have used here. It suggests, for example, that the allocation has been designed partly with attention to positional balance. (That may be a sweet smell, not objectionable here.)

--
14 pitchers in 100 baseball players is a small number that will occur only about once in 1000 samples from a population with three pitchers for every eight others. But no one's ballot should be interpreted this way now, for it is not the ballot but the ballot plus the elected HOM that represents a voter's rankings.

1997 Prelim

Keith Hernandez, Vic Willis, and Don Sutton make my PHOM.

1. Hugh Duffy - It isn't that I really love Duffy but that with guys like Keller, Childs, and Jennings gone I have no real teddy bears to hold onto.
2. Dick Redding
3. Buck Walters
4. Elston Howard
5. Pete Browning
6. Gavvy Cravath
7. Dizzy Dean
8. Rollie Fingers
9. Alejandro Oms
10. Vic Willis
11. Roger Bresnahan
12. Larry Doyle
13. George van Haltren
14. Urban Shocker - He has been in my top 25-30 since I Started in 1935 but this is his first vote. Very nice career DERA puts him above soem competitors.
15. John McGraw - While he has leapfrogged about 6-7 other guys in recent years, this has more to do with 5 guys coming off my ballot in the last two years than anything else. Very nice peak, would be at the top of my ballot (and probably already in the HOM) with better durability.

As I look at my ballot I realize that there isn't a single player that I would really go to the mat arguing for. My #1, Hugh Duffy, will probably be in the bottom 20-30 players of my PHOM and it goes downhill from there. I really ahve trouble arguing passionately for anyone in that group.

However, this does not mean that I support a noticeabley smaller hall (I mean 2-5 players isn't a big deal). It is just that there will naturally be 20-30 question marks in a 230 person hall. I am only convinced that about 200-210 players really belong in a 230 person hall and the last bunch are tough to differentiate from the next 20-30 guys.

But this is because I am starting with the contruct of a 230 person hall. If we made the HOM a 200 person hall, I would still have questions about the bottom 17-25 guys. Starting from that point I would only be convinced that about 180 guys are definitely in the top 200 players of all-time. If we make the hal 180 or 175, then I may only feel comfortable about 160 guys. This would happen right on down to the point where I am wondering if Ty Cobb is one of the ten best players ever, which I find pointless.

So I guess I am saying two things. One, dont 'expect to really fight for the guys on my ballot right now because I have tons of questions about them, probably the same that everyone else does. It is a question or which questions bother me most at this point. This may seem like a cop-out but it is the truth.

Second, is that this does not mean that I think the HOM is really too big. It's just the nature of the project, combined with the fact that my teddy bears, guys like Keller, Childs, Jennings, and Ferrell are all gone now. I mean if we were to make the HOM bigger I may throw down for Duffy, Redding, or Walters. I think they belogn in a hall this size, but I don't feel passionate about that.

Paul Wendt *splatted* me (#134) with ....TomH #23 called for rotten tomatoes.

McGraw, Chance, Bresnahan, Monroe, Elliott, and Walters on the ballot (top 15 backlog) constitute a warm welcome for heavy hitters with playing time issues. One rotten tomato for Charley Jones down around number 50 with Easter and Fregosi.

how does Bancroft not rank at least high 30ish with Leach? One rotten tomato for leaving him out of this company and a second for leaving him out of the fifty-man backlog you have listed.

I will spare a fourth rotten tomato for placing Laughing Larry so low.
--
Charley Jones loses pts from me for weak league competition (1870s).
Doyle - I believe WARP's interpretation that he cost his team 10 to 15 runs/yr with the glove. Weaker (NL circa 1910) league doesn't help either.
Bancroft - okay, call me tomato-faced. Not sure how I didn't have him listed among my top 50. He belongs about even with Rizzuto as my #2 or #3 backlog shortstop.

oh, and.....

<u>PLAY BALL!</u>

Prelim:

Ballot loses the three electees plus Perez who I overrated by about 10 slots.

I like Evans well enough, but he is very similar to a bunch of guys from his era, none of whom have hit my ballot or particularly impress me (e.g., Staub, Bonds, Parker, Dawson). To be a career candidate in this era, you need to have Winfield/Molitor type numbers. As a prime candidate, he is a contender but not especially appealing. He's in the 15-25 range.

That leaves me 4 slots for new backloggers. I'm adding 2 former staples who work their way back up, one new guy who debuted at 15 but slid off for half a decade, and one guy I simply overlooked.

What I'm left with is a bunch of guys who are truly borderline. Roush is the only one who I am certain is in the All-Time 200 and I probably have him about 180. Here goes:

(1) Roush (3rd)
(2) Bridges (4th)
(3) Oms (5th)
(4) Cash (6th)
(5) Leach (10th)--If pushes comes to shove, I'd rather have him in then the nect few.
(6) Reggie Smith (7th)
(7) Bob Johnson (9th)
(8) Elliot (11th)
(9) Cravath (12th)
(10) Beckley (13th)
(11) Shocker (14th)
(12) Ben Taylor (unranked)-- I think I missed the boat on im.
(13) Fingers (unranked)
(14) Chance (unranked)
(15) Duffy (unranked)

Tommy Corcoran and his 103 WARP 1 score give Willie Davis a run for his money. Corcoran's the reason why I cross-check with WAAP to see how "empty" that replacement play was and why I don't take WARP at face value.

BOOKS FOR FREE

I'm clearing out my bookshelf, and I found a few baseball (and other) books. If anyone's interested drop me an email. If I don't hear by next Monday, they go off to the used book store or goodwill or something.

Baseball
-Some of My Best Friends Are Crazy by Jay Johnstone
-The Niekro Files by Phil and Joe Niekro
-The All Star Game (goes through 1988) by Donald Honig
-Men at Work (featuring "Who Can It Be Now"...) by George Will

History
-King William IV by Philip Ziegler
-The Boer War by Thomas Packenham

-The Niekro Files by Phil and Joe Niekro

Two hardboiled gumshoe brothers refuse to knuckle under any intimidation or obstacles for their clients.

Oh yeah, WARP1 is downright batty on prewar middle infielders.

The Niekro Files by Phil and Joe Niekro

Two hardboiled gumshoe brothers refuse to knuckle under any intimidation or obstacles for their clients.

Niekro Files... please tell me there isn't an emery board on the cover. ;-)

We used to call it the CF glut. But now, of course, we have gluts everywhere. Still, with Edd Roush as the leading backlogger (finally; he's #1 on my 1997 prelim) and Jim Wynn just recently elected and with my PHoM selections getting deep into the backlog, I decided it was time to re-eval. the CF glut.

And also to cross-check something I should know by now. Is OPS+ park adjusted? I assume that it is, but then I have trouble understanding the CF glut.

Setting Pete Browning aside for the moment as more of a corner type, the leading candidates AFAIAC are Roush, Reggie Smith and Jim Wynn (Wynn a candidate for PHoM).

WS
Roush 326/38-33-30-29-28 peak
Smith 325/29-29-26-25-25
Wynn 305/36-32-32-31-28

For career you've got Roush and Smith, for peak Roush and Wynn, with Roush seeming to be the big winner if you consider both. But where does this pecking order come from?

Defense
Roush A-/62 career dWS
Smith A-/52
Wynn B-/46

OK, this maybe helps to explain the career totals.

OPS+
Roush 126/162-53-49-47-45
Smith 136/168-64-58-48-42
Wynn 129/168-58-54-47-46

How in the world does this explain Smith's weak peak/seasonal numbers? Without even looking, we would assume that there are playing time issues that affect Smith. With his glove being better than Wynn and his OPS+ being better, how can he trail Wynn's peak 36-29 (1 year), 100-84 (3 years), etc?

Playing Time
Roush 13 yrs OPS+ ? 100, 14 yrs ? 100 games--average games 14 years = 135, missed games = 19
Smith 14 yrs OPS+ ? 100, 13 years ? 100 games--average games 13 years = 140, missed games = 22
Wynn 11 years OPS+ ? 100, 12 years ? 100 games--average games 12 years = 145, missed games = 17

Does 5 games explain Wynn's huge edge in peak WS? Not even close.

So I have to conclude that since the uber-stat doesn't add up, that Reggie was better with the glove and just as good or better with the bat as a rate, and 5 extra games a year for Wynn just can't make up for that. So Reggie clearly beats Wynn. I can make a better case for Wynn vs. Roush, since Wynn's slight edge on offense probably cancels Roush's edge on defense. So, for peak, it's a wash and for career it's Roush.

Then there's Roush and Smith. Smith has the same OPS+ edge over Roush, which ends up having no affect whatsoever on their WS. For career it is Roush. All of the above numbers are unadjusted except for WWI. IOW Roush is rated at 154 games and the others for 162. So just eyeball that. And Roush has a big peak edge over Smith. Is this credible?

Still, I end up pretty much Roush > Smith, Smith > Wynn, Wynn = Roush, which doesn't make any sense. And the one and only thing I know as to why this doesn't make any sense is that Wynn's WS peak is nonsense. But maybe Roush's is too.

Discuss.

(PS. For me, in CF the glut after these 3 [of 4, counting Browning] is Leach, Oms, Duffy, GVH, [CP Bell] and Hack Wilson. There's another 10 in the wings if we should happen to elect these 10, which of course ain't gonna happen.)

1. Roush's peak advantage (WS per yr) is dependent on Reggie missing more time. If someone were to use some measure above AVG, Smith would look better.
2. If I were a WS voter, Roush would be high on my ballot.
3. WARP has Roush's defense much much worse than WS does. I personally don't buy WARP's take, but it does temper my opinion of Roush's glove.
4. Roush played in what many think was the weaker league at the time. If were comparing OPS+ (yes, it is park adjusted), I'll take a 1975 AL or NL OPS+ of 130 over a 1915 NL OPS+ of 130.

One interesting point about the Roush/Smith/Wynn comparison is the sabermetric evaluation of their offensive contributions. In this case, OPS+ undersells Roush, even though he drew fewer walks than Smith and Wynn. Roush reached base a little more often relative to park and league in his prime compared with Smith and Wynn. Smith and Wynn obviously had more power.

Smith, of course, was a centerfielder for only 5 years, and had his best seasons in the late 70s as a rightfielder.

1. Well, Reggie only missed about 3 games a year more than Roush in their primes. But your larger point--Reggie was further above average--probably holds.

4. Most people thought Reggie was in the weaker league at the time, too, though I'm more of a pennant-is-a-pennant guy. I mean, yes, I discount Roush's FL season and Browning's AA, but where AL vs. NL is concerned, no, not really.

And the fact that OPS+ is park adjusted (which I thought was the case) only makes the Wynn-Reggie comparison all the more inscrutable. I mean, yes, Wynn played more games per prime season, but again that was about 5 games a year.

And Mike, yes, thanks for reminding me that Smith (and Wynn) were not career CF, while Roush was. Advantage Roush. For Smith we're talking 5.5 years and about 770 games, for Wynn it's about 7 years and 1180 total. Advantage Wynn (over Smith), though the idea that Wynn was a regular CFer for X number of years based on 1180 games is not accurate, either (see below). I mean, basically, he never had even 2 consecutive years as a "regular" CF. Some of that was injuries, some of it was time spent in the corners, which is consistent with his B- WS rating. Smith OTOH was an A- glove according to Bill.

Games in CF

Smith 6-144-155-136-145-87-4-104-0-1-3-1-4-9-1-5-7
Wynn 10-51-155-104-157-93-149-87-48-12-10-148-107-50

More on Smith-Wynn:

For peak seasonal value, though, you have to look at hitting quality, fielding quality, and durability on a seasonal basis. If you compare Wynn's and Smith's best seasons with the bat, you find that Wynn was signficantly more durable in those seasons, and he was generally playing centerfield, where Smith was generally playing right field in his best offensive seasons.

Wynn
Year -- OPS+ -- Games -- Position
1969 -- 167 -- 149 -- CF
1968 -- 157 -- 156 -- CF/LF
1974 -- 151 -- 150 -- CF
1972 -- 146 -- 145 -- RF
1965 -- 143 -- 157 -- CF
1970 -- 141 -- 157 -- CF/LF
914 games total

Smith
Year -- OPS+ -- Games -- Position
1977 -- 167 -- 148 -- RF
1978 -- 161 -- 128 -- RF
1974 -- 157 -- 143 -- RF
1980 -- 153 -- 92 -- RF
1973 -- 150 -- 115 -- CF
1969 -- 143 -- 143 -- CF
1972 -- 143 -- 131 -- RF
900 games total

So, although Smith has seven 140 OPS+ seasons to Wynn's six, he had fewer games total in these seasons. If you drop the 92-game season from his record, Smith has six seasons at this offensive level, averaging 135 games per season, where Wynn averaged 152. That's a 10% advantage for Wynn. Are playing time differences and positional differences in their peak batting seasons a sufficient explanation for the differences in their peaks, according to win shares? A more detailed accounting is still needed to get a definite answer to this question, but I think there is enough evidence here to conclude that win shares may well have it right when it quantifies Smith's and Wynn's peaks as

Smith 29-29-26-25-25
Wynn 36-32-32-31-28

In some cases, Wynn is 10% more valuable, just as their PT differences would suggest he should be. In other cases, he is 20% more valuable, which might reflect the difference between being a centerfielder and being a corner outfielder.

Chris, good points. OTOH in 1996 you've got Wynn 3rd on 32 ballots and Reggie 29th on 10. The measured difference is a bit slimmer than that. Of course, I understand how the balloting works. A 16th place ain't got that swing. Maybe Reggie has got some 16-17-18-19-20th place votes that will translate soon...?

Now that I've finished my WARP for all post-1893 position players, I'd like to take some shots at the preferences of the group that seem the most outlandish to me. Hopefully I can make people reconsider some assumptions, or take a look at players they may have been overlooking.

1. John McGraw: Everyone and their mother agrees that both WARP and WS have replacement levels far below the true MLB level. This means that McGraw--and all low-playing-time guys--will be severely underrated by them, since they give way more credit for "just showing up" than it deserves. WARP furthermore underrates McGraw by continuing to hand 2B way more FRAR than 3B into the 1890s, when the positions were about equally deep (WS gets this right, but underrates position scarcity in general). Give McGraw (and all 1890s 3B) about 12 extra FRAR per 150 games to properly reflect the difficulty of his position, and then subtract about 2.8 WARP1/10WS per 150 games (more WS for 1B/OF, less for SS), and see where McGraw winds up. With a proper replacement level, he has the same amount of career value as Brooks Robinson! (and more before docking him for standard deviations). Key questions: Are you adjusting for the difficulty of prewar 3B? Are you adjusting WARP and WS replacement levels? If you do both, and John McGraw isn't on your ballot, why not?

2. Reggie Smith: We just elected Jimmy Wynn. Smith and Wynn appear to be clones, with similar MLB career lengths, similar amounts of time played at CF, similar peaks, etc. The differences are both in Li'l Reggie's favor: first, WS thinks he was a notably better fielder than Wynn, and second, he has a nifty Japan year that adds to his prime (and his peak, by rate). Question: If Wynn, why not Reggie? (I see we're already talking about this).

3. David Concepción: Is Ozzie Smith a no-brainer for you? Compare David's prime to Ozzie's--they're *identical*. Same hitting, same fielding, same baserunning, basically the same league context. The only difference is that Smith has four more years at that level than Concepción does. Concepción had a bunch of crappy years at the beginning and end of his career that weigh down his career averages; pretend he never played in 70-72 and 83-88 and make him a short-career SS from 1973-82. How does he measure up? You have to pay attention to his context--a low-scoring, hard-to-dominate league where Mario Mendoza was a full-time player. From that perspective, the guy was a superstar. Questions: Are you actually taking into account positional averages or replacement levels, or just guesstimating them? The gaps are bigger than you think particularly from about 1965-85. Are you counting for how difficult-to-dominate the 1970s and 80s were, when a 155 OPS+ was good enough to lead the league in many years? Is 70% of Ozzie Smith good enough for your PHoM?

4. Phil Rizzuto, Johnny Pesky: Both of these guys had three peak years ripped out of the heart of their careers. Question: Are you giving full war credit to all deserving candidates? Rizzuto looks fairly close to Sewell after accounting for the high stdev of the 1920's AL and giving him war credit.

5. Graig Nettles: He had a career OPS+ 6 points higher than Brooks Robinson's. He played as long as Robinson. He played just one less season than Robinson. His peak defense, according to WARP and WS, was actually better than Robinson's, although he tailed off earlier. He played in leagues that were much more difficult to dominate. Question:If Brooks, why not Nettles?

6. Edd Roush: He was the basically the same hitter as Kiki Cuyler, a bit worse after deducting for the Federal League. And he was a merely average fielder, while Kiki was an excellent one. I have Cuyler's advantage on fielding quality far outstripping Roush's extra relative time at CF. Question: If Roush, why not Cuyler?

7. Nellie Fox: This would be, I think, the biggest mistake in the entire HoM, and I wish I could do more than argue passionately against him. There are tons of guys on the ballot--Concepción, Rizzuto, Bancroft leap to mind--who hit like Fox, fielded their positions better than Fox fielded his, AND PLAYED SHORTSTOP, which is WAY more demanding and harder to replace than 2B--most 2B are failed shortstops, after all. The difference between playing 2B and SS is over a full win per season. If Concepción, Rizzuto, and Bancroft had played 2B, they'd be comparable to Fox. BUT THEY DIDN'T. THEY PLAYED SHORTSTOP. Question: Why the 2B Fox over the numerous SS who hit as well as he did and fielded better?

Given the make-up of the backlog, the big questions for me are in CF (see above) and at the corners (to come). E.g. if anybody, why not the Baby Bull. (But, again, later.) But for now, Dan raises a bunch of good questions, though I don't agree with all of his answers.

1. How do you adjust for "difficulty"? Maybe that's a matter of placement among peers. But even McGraw's rates are not that spectacular:

McGraw OPS+ 135/165-48-31-29-26-10-7
Ezra Sutton 119/166-64-51-47-41-36-16-13-10-7
Al Rosen 139/181-62-48-46-46-25-4-3
Fred Dunlap 132/159-47-43-40-24-22-21-19-12-7

And if playing time is such an impossible task in McGraw's day, how do you explain Lave Cross? Besides, if it was so hard for a 3B or IF generally to stay in the line-up in those days it was McGraw's own damn fault as much as anybody.

2. If Wynn, then Reggie Smith. With ya there. But if Wynn and Reggie, then Roush.

3. I have to accept that it was somehow harder for Davey Concepcion to hit OPS+ 90 than all those other guys you mention in order to rank him ahead of them--meaning Bancroft, Rizzuto, even Pesky--and I'm just not there yet (setting aside Fox for the moment). And is 70 percent of Ozzie good enough? Not even close. I can wait for Barry Larkin for a Red-legged SS.

4. Rizzuto and Pesky: I'm with ya there. Rizzuto is on my ballot, Pesky is in the second 15. At least a full SD above Cecil Travis ;-)

5. Nettles. Well, "if Brooks, then Nettles" doesn't work for me. I slipped down the slippery slope for Brooksie but I have great remorse about it. I'm just not usually a sucker for the career arguments. And so with Nettles. I like him a lot better than Buddy Bell, but what about Reggie (er, I mean Tommy) Leach?

Nettles 321/28-27-26-25-22-21-21-19-18-17 (16 yrs in double figures)
Leach 328/31-29-28-27-26-25-24-21-19-17 (15)

Or: Nettles 110/135-28-24-21-20-15-12-12-10-9 (12 yrs 100)
Leach 108/135-32-32-25-21-16-10-6-0 (9)

And: Nettles A- and 93 dWS, Leach A+ and 95; and Leach was out there at 3B when it was so hard to stay in the line-up. I'm not sayin' Tommy is a slam dunk over Nettles, but "if Nettles, then Tommy."

Darrell Evans was a lot more compelling and I didn't really like Darrell all that much. And Ed Williamson is still hanging out there. A better glove than Nettles, and a better hitter. A shorter career, sure, but not relative to his peers.

My 3B: #11 Williamson, (#16 Da. Evans and Ezra Sutton--both PHoM last year), #18 Tony Perez (only 1/4 of a 3B of course), #20 Leach (half a 3B), #23 Rosen, #28 Bando, #29 Pesky (half a 3B), (#32 Sewell, 1/3 of a 3B), (#32 Ken Boyer, not a HoMer IMO), #35 Elliott, #44 Clarkson (half a 3B), #50 Nettles, #51 Traynor, #65 McGraw, #85 Cey, #100 Buddy Bell. That is a hell of a glut and I sure could be mis-underestimating somebody. But if so, I still wonder about Traynor more than I wonder about Nettles or McGraw. I also wonder if I've mis-overestimated Tony Perez but that will get washed out among the "hitters."

6. I am re-checking my "hitters" now and Cuyler didn't even make my list of 24 guys to re-eval. (He is #101 on my ballot, after Buddy Bell.) I will add him in, but I will be surprised if he doesn't come in 25th. As to Cuyler being a better fielder than Roush, I just don't believe that. I mean, stupidity has been rampant in MLB at times but if that were true how did Roush stay in CF all those years?

7. OK, now this is a good question. Fox is on my ballot, he is already PHoM (as is Rizzuto) but your argument makes sense. The guy I really worry about, though, is Dick Lundy. We now have reason to doubt the argument that blew his HoM chances away--I mean, he dropped from OPS+ 110ish to about 90ish when it was decided he didn't walk enough, and now we don't know if that was an accurate assessment.

I've got Rizzuto in my top 10, Fox between 10 and 15, Leach around #20, Pesky about #30, Lundy and Bancroft at #55 and 57. This is among the good-field, no-hit candidates, there are other IF in there. Bancroft is probably too low, but at least I know that if he is too low, it's not by a ton. I mean, maybe I could see him around #25-30. But with Lundy, if he's too low he could be around #25-30 just if I decided to value defense more highly. Then if I figure he's a 110 rather than a 90-95 (OPS+) then he's on my ballot. We gotta figure him out.

David Concepción: Is Ozzie Smith a no-brainer for you? Compare David's prime to Ozzie's--they're *identical*.

Maybe by some measures, but WS sure doesn't see that.
Concepcion, Dave WS by year for career ............ ...total prime16 prime13 prime10 prime7 prime4
..5 4 6 16 25 19 23 19 25 24 17 20 17 .8 11 12 8 8 2 ....269 ....258 .....236 .......205 .....152 .....91
Smith, Ozzie ........WS by year for career ............ ...total prime16 prime13 prime10 prime7 prime4
20 7 17 8 19 18 19 25 23 33 22 20 11 25 20 19 9 2 8 ...325 .....306 .....262 .......217 .....159 ...103

Given that WS is known to UNDERrate superior defense, one would generally think that the greatest defender in the history of the universe might be shafted by WS, but here Ozzie's prime, regardless of how many years are used, is better than David C's.

(Not that you have to have the prime of Ozzie to be in the HoM, tho!)

Like fellow peak voter Sunny, I'm finding Dan's arguments most convincing--at least where I didn't already agree with him to begin with (McGraw, the wartime SS candidates)--concerning Fox. I'm beginning to feel I went a bit too much with consensus on him, and didn't hold to my peakster principles quite enough.

I'm leaning heavily toward dropping Fox off my ballot and down quite a bit for this election. Do any of the hardcore Fox supporters want to counter Dan's arguments against him, though?

Thanks for taking my "challenges" seriously! This definitely makes all the time I put into calculating these stats worthwhile. Here are my responses:

Sunnyday2:

1. OPS+ will *criminally* underrate McGraw. First, it doesn't count his steals, which is a major, major part of his offensive production. Second, he is perhaps the most OBP-heavy player ever, and moreover, he played in high-scoring leagues, when the relative importance of OBP to SLG is at its highest. On a rate basis, his offensive production in 1899 is more like a 205 OPS+ in the modern game than a 168 (before docking him for standard deviations)--I can show you the math if you want. And third, while an average defender overall, McGraw's fielding was quite up-and-down, which means when it was good--as in 1899--it brought him to Ruthian levels of dominance. The problem is of course the playing time, but the rates are stratospheric. McGraw is an extreme player where OPS+ will miss drastically--you have to use a run estimator to calculate the value.

2. Roush only follows from Wynn and Smith if you ignore the fact that Roush played in the teens and twenties, when standard deviations where far higher than they are today. Roush's WARP or WS "bought" notably fewer pennants than Smith's did.

3. That's really stacking the deck against Concepción. Concepción added a good chunk of value on the basepaths that OPS+ doesn't capture. Moreover, about half of his career was worthless filler at the beginning and end, while Rizzuto had virtually no decline phase and Bancroft's wasn't nearly as long. He had a 101 OPS+, plus substantial baserunning value, for his best decade (1973-82). Bancroft was 106 with no baserunning value from 1918-27, while Rizzuto was just a 95 with a bit of baserunning value from 1941-53 (10 MLB years). Then you add in the standard deviation factor, and I think Concepción comes out on top. That said, I have the three of them within $6M of each other, so it's really a virtual tie.

4. Well, if you now don't think Brooks was deserving, then of course "if Brooks, then Nettles" won't sway you! I'd definitely vote for Leach if he had had his career in the 1970s and 80s; he'd be at $96M. But he played in an extremely high standard deviation era, where his WARP/WS didn't buy nearly as many pennants as Nettles's did.

5. Roush wasn't a bad fielder; he was basically average (-14 FRAA for his career). But Cuyler was a good one (+41 FRAA). And their offensive profiles are identical.

6. Again, why Fox over Bancroft? Bancroft had a HIGHER OPS+ (98 to 94), and was a superstar defender, while Fox was merely a good fielder. AND BANCROFT PLAYED SHORTSTOP. Seems to me Bancroft has every argument but career length...

TomH,

I find WS so opaque I barely pay attention (although I do reluctantly use Fielding WS in my system). If you're interested, I'll be happy to walk you through how I have Ozzie and Dave's best nine years basically equal, step by step.

Dan R is arguing that Nellie Fox would be our worst mistake. Someone else has been posting for years that Pete Browning would be our worst mistake. I haven't seen any specific posts on this but I'm sure we could find a voter who would think that Edd Roush would be our worst mistake. We're deep into the backlog and anybody that we elect at this point would be "our worst mistake."

And Dan R, what about Dick Lundy? He is Bancroft's equal when seen in his (Lundy's) worst possible light--i.e. the light in which he has been cast here all these years. If you give him any allowance at all as not being a complete, total and utter slug when it came to take a pitch out of the strike zone, he can do nothing but elevate.

I'm also wondering if Dan R is guilty of time-lining and not being fair to all eras. In post #164, he's basically said that he would vote for Roush if he played in the '70s instead of the '10s and '20s, and Leach if he played in the '70s and '80s instead of the '00s and '10s.

Dan R is arguing that Nellie Fox would be our worst mistake.

Well, that's not the part of his post I was talking about, anyway--it was the bit comparing him to the various SS candidates.

I'm also wondering if Dan R is guilty of time-lining and not being fair to all eras.

I don't think he can be accused of traditional timelining--his is based entirely on his standard deviations, so it's not that all old eras=bad and new ones=good. He's not high on the expansion '60s, for instance.

And of course, timelining wouldn't affect his take on Fox much.

I suppose I can let Dan defend himself, though. :)

sunnyday2, the consideration set for this work is just MLB position players from 1893 to the present. I will do a comprehensive review of Negro Leaguers (and pitchers, and pre-1893 players, of course) before voting in 1997.

Chris Fluit, no timeline at all--I am indeed being fair to all eras. That includes being fair to low standard deviation eras like the 1970s and 80s, when it took fewer WARP or WS to win a pennant. A pennant is a pennant, and you can't effectively convert from wins to pennants until you know the standard deviation. That is what I am correcting for. This is no timeline--I regress the 1998 NL more than the 1908 NL.

Chris, good points. OTOH in 1996 you've got Wynn 3rd on 32 ballots and Reggie 29th on 10. The measured difference is a bit slimmer than that. Of course, I understand how the balloting works. A 16th place ain't got that swing. Maybe Reggie has got some 16-17-18-19-20th place votes that will translate soon...?

He'll be #15 next week on my ballot, Marc.

Dan R is arguing that Nellie Fox would be our worst mistake. Someone else has been posting for years that Pete Browning would be our worst mistake. I haven't seen any specific posts on this but I'm sure we could find a voter who would think that Edd Roush would be our worst mistake. We're deep into the backlog and anybody that we elect at this point would be "our worst mistake."

There is no such thing as a consensus "worst mistake," anyway. Peak guys will have their suspects, prime guys their own targets and career voters will peg different candidates.

And Dan R, what about Dick Lundy? He is Bancroft's equal when seen in his (Lundy's) worst possible light--i.e. the light in which he has been cast here all these years. If you give him any allowance at all as not being a complete, total and utter slug when it came to take a pitch out of the strike zone, he can do nothing but elevate.

If the HoF's data is correct, Lundy's career OPS+ is about the same 92ish that it was under the old MLEs. His plate discipline is a lot better than the old evidence had suggested, but it looks like his power numbers were significantly exaggerated in the old data also, so that his OPS+ comes out about the same as before. He doesn't quite come up to Dave Bancroft as a hitter.

I found the data...I'm about to head out of town in just a second to have no internet access for the weekend, so I'll email you all the data I have when I return..

KJ,

Just about ready to post up Starvin' Marvin, were you able to locate the data for the leagues in question? Thanks!!!

1953 Laredo, G-147, AB-5235, H-1383, 2B-219, 3B-35, HR-82, W-71, L-76
1954 Vancouver, Won - 74, Lost - 49
1955 Columbia, Won - 89, Lost - 51

Sorry don't have anything more for 1954 and 1955.

1953 Gulf Coast League, G-580, R-57772, AB-38130, H-10170, 2B-1630, 3B-331, HR-631, BB-4287
1954 Western International, G-566, AB-38024, R-6201, H-10723, 2B-1807, 3B-332, HR-723, BB-5243
1955 SAL, G-559, AB-37383, R-4862, H-9524, 2B-1420, 3B-411, HR-552, BB-4526

7. Nellie Fox: This would be, I think, the biggest mistake in the entire HoM, and I wish I could do more than argue passionately against him. There are tons of guys on the ballot--Concepción, Rizzuto, Bancroft leap to mind--who hit like Fox, fielded their positions better than Fox fielded his, AND PLAYED SHORTSTOP, which is WAY more demanding and harder to replace than 2B--most 2B are failed shortstops, after all. The difference between playing 2B and SS is over a full win per season. If Concepción, Rizzuto, and Bancroft had played 2B, they'd be comparable to Fox. BUT THEY DIDN'T. THEY PLAYED SHORTSTOP. Question: Why the 2B Fox over the numerous SS who hit as well as he did and fielded better?

I am somewhat bothered by this line of argument. Simply put, you cannot play baseball without a 2B. In the pennant is a pennant system of merit, it follows that the pennant winner needs 1 starting SS and 1 starting 2B. Naturally, the skill set requirements for each position are somewhat different, but they are both absolutely essential. I cannot, in good conscience follow an argument that seemingly downgrades a player for what he wasn't. Not every 2B is a failed shortstop just as not every RP is a failed SP.

The "failed SS" point was just rhetorical. The real point is simply that a league-average hitter at SS is worth far more than at 2B, because you can get say an 85 OPS+ guy for free at 2B but only a 70 OPS+ guy for free at SS. Thus, a 94 OPS+ with good defense at second (Fox) is not nearly as valuable as, say, a 98 OPS+ with great defense at short (Bancroft). I haven't seen any counterargument to this yet.

For my money (and I may be underrating him a bit), I give them equal value with the glove as well as roughly equal value offensively. The separation comes from career length and in-season durability.
Also, I personally rank players by position before I set overall rankings.
Among eligibles, Bancroft is one of many. He doesn't stick out of other eligibles in any one category. I can see how he would be thought more of than someone like Rizzuto or Pesky, but I'd take Stephens over him any day.
Among 2B, no one really comes close to Nellie (according to my world view). Larry Doyle would if I could be convinced that he was at least league average as a 2B. As the 2B with the best overall package, I will continue to have Nellie Fox high on my ballot and as an unabashed member of my PHOM.

The real point is simply that a league-average hitter at SS is worth far more than at 2B, because you can get say an 85 OPS+ guy for free at 2B but only a 70 OPS+ guy for free at SS.

Out of curiousity, where do teams get "free" 85+ OPS second basemen? Do they go to MLB offices, and say, "may I please have an 85+ OPS second baseman? Mine got hurt, and I don't have a replacement."

Sorry for the snark, but one of my biggest disagreements with Dan's methods is his insistence that you can measure "replacement value" with some level of precision. Replacement value is a theoretical concept, not one that can be exactly measured. Every team has different options - I don't think you can say with certainty that a replacement level second baseman is x% over a shortstop.

That said, I agree that Fox was a little overrated, and I agree with the given point that, all else being equal, I value shortstops over secondbasemen. Fox has other advantages over the guys Dan mentions - durability, for example. I also have nowhere near as much confidence as Dan does in defensive statistics - its not a given that Fox, a gold glover, was only "good".

rawagman--but there has to be some sort of absolute standard...just hypothetically, let's say that the HoM had already elected 50 first basemen, something ludicrous. Then let's say that there was a HUGE gap between the #51 and #52 available 1B. The #51 1B might stand out in relation to other first basemen...but would still be inferior to the masses of backlog players at every other position. I don't think our real HoM has grossly overelected 2B by any stretch, but I do think that the best avilabile 2B (who I would say is Tony Lazzeri) is well behind the top remaining candidates at other positions.

Dizzypaco--They call one up from the minors, generally. Someone like D'Angelo Jiménez would be a freely available 2B in today's game. You want him? He's yours.
OK, but even anecdotally, Concepción was certainly considered to be an all-time great fielder; I imagine Bancroft was as well although I'd have to dig into the archives. Would Nellie Fox be on anyone's list of the all-time greatest fielding 2B? I mean, you don't have to be Maz or Bid McPhee, but I certainly wouldn't think of him in the next tier down. (Then again, my "anecdotal" sense may be influenced by my knowledge of defensive statistics).

I voted for Reggie Smith a couple of times, and may again. I never voted for Wynn until a 15 slot last time. Theoretically, Reggie may grab that spot.

Dan - that's a nice hypothetical, but not very useful here.
We currently have 17 2B in the HOM and 20 SS's.
Of the upcoming candidates, I would hazard a guess that among 2B, we will elect Sandberg and probably one or both of Randolph and Whitaker.
Same idea for SS - we will almost surely elect Yount (approx. 50% SS), O. Smith, Trammell and Ripken.
This leaves us with 19-20 2B and 23.5 or so SS.
Not an enormous discrepancy, but enough that I'm not going to go out of my way to distinguish among the backlog SS. (But I'll still vote for Vern Stephens).
Seeing as how I still feel that Fox stands head and shoulders above the remaining backlog 2B, and they are at somewhat of a positional disadvantage to SS, there is more of a case there for Nellie.

They call one up from the minors, generally. Someone like D'Angelo Jiménez would be a freely available 2B in today's game. You want him? He's yours.

Not every team has a decent secondbaseman hanging around in the minors just waiting for an injury. I don't think it is true that there are always easily identifiable players available to pick off waivers who are guaranteed to put up an OPS+ of 85.

There are lots and lots of examples of teams playing a player at second base who was a terrible hitter - Doug Flynn played for years. When a guy gets hurt, he is often replaced by someone really bad. That there is someone else out there who is theoretically available is irrelevant. I'm guessing, but I believe this may be what leads your system to underrate players who are durable and overrate those who miss large numbers of games.

OK, but even anecdotally, Concepción was certainly considered to be an all-time great fielder;

Actually, I'm pretty sure he wasn't. He was certainly considered very good, and won some gold gloves, but I really don't think he was considered at the level of, say, Ozzie Smith or Marty Marion. I think he was thought of as pretty similar to Nellie Fox, actually - very good, but a clear step below the all-time great fielders.

To be fully consistent, I really should reconsider how backlog 3B stack up against the other positions, which I have been meaning to do for some time.

rawagman--I guess I don't have positional "affirmative action" in my PHoM besides a 20% catcher bonus. I don't see why we have to assume that the greatest players have been distributed equally across positions over time. I just use standard deviation-adjusted wins above contemporary positional replacement level as the basis for my evaluation, and leave it at that.

Dizzypaco--no, what causes my "system to underrate players who are durable and overrate those who miss large numbers of games" by your estimation has nothing to do with my WARP and everything to do with my salary estimator, which I explicitly designed to reflect my preference as a voter for peak rate. The actual salary equation given by Nate Silver is $212,730*WARP^2 + $402,530*WARP. I apply that equation to WARP2 *per full season played* and then multiply the result by the player's percentage of the season played. If you apply the estimator to simple seasonal totals, you get, for example, 2003 Albert Pujols with $26,937,351 and 2003 Barry Bonds with $24,536,199, since Bonds' 231 OPS+ in 130 games was worth slightly fewer wins above replacement than Pujols' 189 OPS+ in 157 games. But I would have voted for Bonds as the 2003 NL MVP over Pujols. To reflect this, I apply the salary estimator to the rate and then multiply by playing time, which gives me $26,955,590 for Pujols and $29,607,509 for Bonds. This "feels" right to me. But your mileage may vary. The effect is quite large: to take the biggest outlier and the #1 player on my ballot, John McGraw, he's $125,400,221 in my system and just $107,985,702 if you apply the estimator to seasonal totals, so we're talking a 15% valuation difference in some cases using the exact same WARP2 numbers.

The salary estimator is just a toy. What would really make me happy is if the group would look at my actual WARP2 data in the same way they use BP WARP or WS, plugging it into their own valuation systems (peak, career etc.) and drawing their own conclusions, since I think/hope it successfully addresses a lot of the issues that people have with those metrics.

Re Concepción: is there any way to go about measuring contemporary opinion? I certainly thought Concepción was as highly regarded as Marion--as good as anyone not named Ozzie.

rawagman, the 3Bs are a quagmire, as is any group you choose to re-eval these days. I tackled 'em a couple cycles ago and, just for fun, threw in the old-time 2Bs, since that was the same place on the spectrum in those days.

The On-Ballot or Nearly So's

1. Larry Doyle--offense trumps defense, an MVP at his peak
2. Tony Perez--like I said, offense trumps, nowhere near an MVP
3. Ed Williamson--great all-around resume, very nearly an MVP type player for a couple years

(...and when I actually cast my ballot I decided to go for a lack of weaknesses rather than for strengths, IOW, for Williamson ahead of Doyle and Perez. And Perez, I should add, is currently slipping into the 16 to 30 crowd anyway. He does well with this group, he does a lot less well if I consider him at the other corner, even with the little bit of added defensive value, because it is just a little bit.)

The 16 to 30 Crowd

4. Tommy Leach--glove trumps offense, more valuable in his particular day than at any other time
5. Al Rosen--peak trumps career, an MVP in his prime
6. Sal Bando--nice peak, a bona fide MVP candidate, the best currently eligible "pure" 3B of the post-expansion era, not that that is super-high praise or anything
7. Johnny Pesky--play at SS only helps, woulda been more valuable almost any other time, and not just because of the war, also it woulda helped if they had put Vern at 3B where he belonged

HoVG

8. Bus Clarkson--played a little 3B
9. Bob Elliott--nice but fading into the crowd
10. Fred Dunlap--very nice even with UA discounted by 65 percent, on my ballot way back when
11. Vern Stephens--looked more like a 3B; WWII hurts, not just the discount but the balata ball; woulda been more valuable if allowed to swing for the fences throughout his career
12. Graig Nettles--sorry, can't get 'im any higher
13. Pie Traynor--also on-ballot once upon a time
14. John McGraw--peak too short even for me

Filler

15. Tony Lazzeri--not sure if he belongs on this list but if you like good-hit/no-field, then Doyle beats the crap out of him
16. Buddy Bell
17. Ron Cey--Bell and Cey were really the same person
18. Johnny Evers--probably underrated but what the hell, can't get him ahead of Lazzeri and sure can't get Lazzeri any higher
19. Billy Nash--nice little player, I should have him and Denny Lyons at 19a and 19b
20. Lave Cross--a tip of the hat to the career guy

There are more apples and oranges comparisons in here than you can shake a paring knife at, and you could just as easily see a different facet of every one. But then, the top 4 are the only ones who really matter. The rest are just for fun.

Dan R, Roush's peak was 1917-1919. He and Rogers Hornsby were arguably the best players in the National League at the time, and the standard deviations do not, as I see them, appear to be unusual. I understand and agree with the argument about evaluating Roush's performance in the 20s due to the decentralizing effect of the environment.

Roush played 2/3 of his career in the high-stdev 20s. I actually have the stdevs for the 1918-19 NL as fairly low (similar to the 2003 NL), so his peak isn't taking much of a hit in the adjustment at all. It's the rest of it where he gets dinged.

For fun, how do your stdevs of the 20s compare with the stdevs of 94-06? It's a little early for you to be thinking about McGwire and Sosa, I know.

Mike, the stdev graphs are available for download in the Hall of Merit Yahoo group.

There is a huge difference in AL and NL stdevs in the post-1993 era. The 1990's NL stdevs are *quite* high--I have the 1998 NL regressed as much as the 1902 AL, for example. The overall level of the NL in the 90s is about similar to the level of the AL in the expansion 60s, or just a tad below the NL in the 20s. The AL in the 1990's is extremely tight, similar to the NL in the 70s and 80s or the integration-era NL.

By 2005, the NL projected stdev has dropped to the level of the AL in the 90s, while the 2005 AL stdev is among the lowest I have recorded.

One of the highest residuals I have--meaning the difference between the projected and the actual stdev--is the NL from 2001 to 2004. A lot of that is just Bonds--another "one-man star glut"--but there are big misses in '01 (Sosa, L. González) and '04 (Beltre, Rolen) even without him. This may be back-door evidence of steroids; it's hard to say.

They call one up from the minors, generally. Someone like D'Angelo Jiménez would be a freely available 2B in today's game. You want him? He's yours.

Except in the 1950s when they had Clarkson, Williams, Easter, Jethroe, Wilson, Barnhill, and dozens of MLB-ready ex-NgLs in the high minors. Dan, you and I have discussed this question off-line, but I wonder what the group thinks about how replacement/SD are or could be influenced by the presence of so many black stars who weren't being called up. They were all options for any MLB team....

My response to this is that that should be part of a league quality adjustment (which I am yet to make), not a replacement level adjustment.

Would the following be a valid way to measure replacement level?

Since a lot of replacement level theory is that one can find 'freely available talent' at a certain level, coudl replacement level be measured by those that are freely available, i.e. AAA or late NeL players? Maybe one could take teh bottom 6 or so MLB players at a position as well as the top 10 outside of MLB but still ready to be called up. This would not, of course, be useful in the era of a strong, independent PCL or for most of NeL history. Josh Gibson and 1924 Lefty Grove would not count. But would those 16 players constitute some sort of replacement level? Is this valid? Would it change things at all?

Mark Shirk, that's close to what I do...My method for calculating rep level is the following:

1. Average the standard deviation-adjusted wins-above-average rate of the bottom 37.5% of starting players at each position for each season in major league history.
2. Take the average of these rates for 1985-2005, and compare that to Nate Silver's Freely Available Talent levels published on Baseball Prospectus (derived by looking at the production of MLB players over age 27 making less than twice the league minimum salary). The gap between these two numbers is the difference between the worst-regulars average and the freely available level.
3. For each season in question, take a nine-year moving average of the worst-regulars average, and add on the gap found in #2.

Note that I am making one giant and unsupported assumption: the relationship between the worst-three-regulars average and the freely available level has remained constant over time. This is clearly not true for 1890s outfielders, all of whom should be knocked down a few ticks.

This is clearly not true for 1890s outfielders, all of whom should be knocked down a few ticks.

i think in general, we'd need to know the whether the FAT in the highest-performing MiLs were (in translation) better than the worst regulars to answer that. Not just Easter but Bilko. Because Bilko might be above replacement, but he's assumed to be if he fits the FAT definition (27+, cheap, etc).

Along with the CF glut, I re-eval the "hitters" this year.

19C--Browning and C. Jones remain the only backloggers of note, and in that order both by me and the electorate, at #7 and #11 which seems appropriate (both are PHoM). I still have Joe Kelley in my personal backlog, where he remains low borderline.

Deadball--Beckley is an imminent electee after all these years (#6) while Cravath is #14 while Chance trails badly. I prefer Cravath to Beckley (Cravath is already PHoM while Beckley is low borderline). Chance is somewhat under-rated but it doesn't matter, he is below borderline for me, too. Jimmy Sheckard remains in my backlog at mid to low borderline, but ahead of Joe Kelley.

Lively Ball/Golden Age--Bob Johnson is over-rated at #13 and Chuck Klein about right at #47. I see them as pretty close and below the borderline. Nobody else stands out from this already over-represented era. Oh, I checked out Kiki Cuyler and, yes, he is nearly Johnson and nearly Klein. Still below borderline.

Negro Leagues (Golden Age, part 2)--We have nobody in contention. Ben Taylor is maybe under-rated at #46 but, again, it probably doesn't matter, he's low borderline though a darn site better than Pete Hill. Luke Easter slides between them.

Early Expansion--We have Perez #9, Cash in the 20s, Cepeda and Brock in the 30s, F. Howard in the 40s. Say what? Cepeda and Howard spread-eagle this group! Both are on my ballot and both are now PHoM. Perez is mid-borderline, Cash is low borderline, Brock well below. See below.

Late Expansion--We have Staub at #22 as the best of this group??? Bonds and Singleton are in the 30s, Jim Rice way down in the 60s. Keith Hernandez remains in my personal backlog, and Dewey Evans and Dave Parker come in as borderliners--Dewey maybe middle border and the Cobra low borderline comparable to Jim Rice. Here again, I think we have it seriously wrong, though I admit that I had Singleton as low as the 70s initially. He is now all the way up in the 10s, clearly a mid- to high borderliner a la Jim Wynn whom I mis-underestimated initially. See below.

Early Expansion

Cepeda 310WS/34-30-29-26-26
F. Howard 297/38-334-30-28-25
Cash 315/41-27-24-24-23
Perez 349/33-32-31-25-25
Brock 348/31-30-30-26-26

In this group, Cepeda has the best WS total for seasons 4 through 10, though all are close through 7 years (all at 21-22-23). And aside from Cash's fluke season, Howard leads in cumes through season 9. Perez has a decent peak story to tell, but Cepeda and Howard just run away from everybody in seasons 4 through 10.

Cepeda OPS+133/166-66-58-48-35
F. Howard 143/180-73-72-54-51
Cash 138/198-48-47-41-40
Perez 122/162-62-48-38-24
Brock fughedaboutdit

Howard takes no prisoners here though Cepeda and Cash catch and surpass in seasons 8 through 10. But Cash's playing time issues are severe. That leaves Cepeda and Howard.

I suppose you could argue that Perez has more defensive value vs. this list and, yes, 49 career dWS vs. Cash at 35, Cepeda 28 and F. Howard 24. No excuses for Frank, he sucked. But Frank leads Doggie 260-249 through 10 years, meaning 234-200 on offense, which is where both earned their bacon. So, yes, Doggie is a credible 3rd here but only that.

Late Expansion

I admit it. I screwed up on Jim Wynn and Ken Singleton. In the case of Wynn, I was pretty focused on Reggie Smith whom I still consider to have been better. But with Reggie safely in my PHoM, I'm now ready to take Wynn, too. In Singleton's case, well, I just screwed up. And I had Dewey #3 on my prelim and now I can't remember why.

Singleton 302WS/36-33-32-28-28
Staub 358/32-30-28-28-27
Da. Evans 347/31-29-26-25-24
Bonds 302/32-32-31-31-24
Parker 327/37-33-31-29-26

Singleton 132 OPS+/168-58-56-54-46
Staub 122/165-55-47-38-36
Da. Evans 126/160-55-46-46-35
Bonds 129/152-46-41-38-35
Parker 121/163-48-46-43-40

So who stands out? WS career is Staub and Evans, WS peak is Singleton and Parker, prime is Staub. OPS+ is Singleton all the way. So I end up with 'em this way:

Singleton
(then Hernandez)
Staub
Evans
Bonds
Parker
Jim Rice

They're all close enough which is another way of saying they are mid- to low borderline. If I'm gonna take a guy with some weaknesses I might prefer to just go ahead and go to an under-represented position and this ain't it.

So put 'em all together and the high borderline and better are these:

Browning
C. Jones
Cepeda
Singleton
Cravath
F. Howard
(Hernandez)
(J. Kelley)
(Sheckard)
Cash

And add in the CF:

Roush
Browning
C. Jones
R. Smith
Cepeda
(J. Wynn)
Singleton
Cravath
F. Howard
(Hernandez)

Bottom line: Roush and Browning and C. Jones are appropriately positioned and Cravath nearly so, Beckley and Perez and B. Johnson and Oms are over-rated, and Reggie Smith and Orlando Cepeda and Frank Howard and Singleton seriously under-valued, at least based on 1996 results.

sunnyday, you left out Alejandro Oms.

one more time for us all to remember: using Win Shares, hitters in DH leagues (AL 1973ff) are underrated. We may disagree on the exact amount, but two players of equal goodness get between 5% and 11% fewer batting win shares in the modern AL than in the NL. Crucial when comparing Nettles/DwEvans/Lynn/ etc to others.

I have Oms listed among the CF in post #154: Roush, R. Smith, (Wynn), Leach, Oms, Duffy, GVH, (Bell). Just a bit south of (Hernandez), (J. Kelley), (Sheckard) and Cash. Borderline, but low borderline. Among the NeLers I am no longer convinced he should be above Ben Taylor but he remains ahead of (Pete Hill) and Luke Easter.

sunnyday2, WS can't handle guys like Frank Howard, because it has no Loss Shares. You'd need negative WS to properly capture Howard's impact with the glove.

Do you really believe that the difference between the best and worst fielding corner OF in the game is just 10 runs (-5 to +5)? 'Cause that's what WS has (corner OF are usually between 1 and 4 dWS).