DL from MN asked how epoc's Personal Hall of Merit (PHOM) would stack up.

Based upon Chone Smith's WAR, I did an analysis of how each position fared against one another, and found the following breakdown, based upon career statistics only:

Top 50 Top 100 Top 150 Top 200
1B 7 1B 17 1B 25 1B 31
2B 7 2B 13 2B 18 2B 24
3B 6 3B 10 3B 18 3B 23
xC 0 xC 5 xxC 8 xC 13
CF 7 CF 12 CF 18 CF 27
LF 6 xLF 13 LF 24 LF 32
RF 10 RF 14 RF 20 RF 26
SS 7 SS 15 SS 18 SS 23
DH 0 xDH 1 xDH 1 xDH 1

If this is any indication of a PHOM based upon WAR, then the SS and C position would appear quite light, although catchers receive a bonus from the electorate for playing the most demanding position, while 1B and corner outfielders are pretty high.

All right. I'll try to get to everything now.

About my evaluation of pitchers generally, I do use FIP, but I balance the WAR numbers I derive from it with Rally's WAR numbers, which are based on RA. I think this is important, as they both measure things of great importance to our understanding of a pitcher's value. It doesn't make a big difference for most pitchers, but you really notice the ones for whom it does. Gooden and Friend are chief among them, as are Stieb and Glavine in a negative way.

Hugh Duffy - I have Duffy worse than the 19th C. outfielders Paul Wendt mentioned. I'm persuaded that we don't need any more outfielders from that period. He'll be dropped from my final ballot, though there's a good chance he'll be replaced by Guidry, which I'm sure won't go over very well either. I don't like Williamson as much as Duffy, but he might be more deserving based on his position.

Tiant - I do not credit the seasons that bookend his injury seasons as consecutive.

I haven't done a personal HoM by year. That is on my agenda for the coming year, but for now I can say that my straight rankings don't differ all that much from what you guys have done so far, in general. Aside from the guys on my list (except for Duffy, who, as I said, I've reconsidered), some other guys I like a lot more than the consensus are Guidry, Vida Blue, Larry Jackson, Mickey Lolich, Addie Joss, Steve Rogers, and Wilbur Wood for pitchers, and Gene Tenace, Ron Cey, Vada Pinson, and Rocky Colavito. I'm not sure those guys are all PHoM, and there are others who probably would be who you guys give some respect to (Munson, Rueschel, Murphy, Bus Clarkson, and Bill Monroe, to name a few).

Dwight Gooden - I guess firstly I'd like to try to dispel the rumor that there's nothing more to him than his peak. Rally's WAR numbers for him aren't very flattering, but his FIP was way better than his RA. I posted this a month or so ago in Gooden's player thread. He was a top ten pitcher in baseball by FIP in each season from 84-91 (except 89, when he was injured), and top 20 or so in 92 and 93. My personal FIP-derived WAR numbers for him for that period are 9.5, 10.0, 5.7, 5.2, 6.8, 1.9 (injured), 7.8, 4.3, 3.9, and 3.9. That's 84-93 respectively. As I said, I balance that against a WAR based on RA, but still, that is why he looks way better to me than to most of you. Secondly, his peak is really, really amazing. Someone suggested I should like Rosen if I like Gooden, but in addition to having a better, longer prime, Gooden's peak is way, way better. Even without considering FIP, their three best years are very close - 22.5 for Rosen, 22.3 for Gooden - as are their five best years - 30.4 to 30.1. Their three- and five-best consecutive years are similarly close, but Gooden at his best, in 1985, was so much better than Rosen at his best, in 1953. By Rally's WAR, that's 9.7 for Rosen and 12.5 for Gooden. The number of position players who've had a 10-WAR season since 1900 is at least 30 (I stopped counting). The number of 12-WAR pitcher seasons (not just the number of pitchers) since 1900 is five (two by Johnson). If you lower the threshold to 11, you only add three more. I'm not saying that one season is enough to warrant HoM status, but if you're a peak voter (which I'm not, strictly) that's a very formidable peak right there. Rosen's pales in comparison, in my opinion. After adjusting for FIP, Gooden's very similar to Cone or Saberhagen (the latter is my comp of choice, as they were contemporaries and had very short careers), but a little better than both (because of the peak).

Bob Friend v. Reuschel - Rally has friend at 40.8 WAR (including hitting) from 55-65, his best 11-year stretch. Reuschel's best is 47.8 from 72-81 (he missed all of 82; if you want to add 83 it's just .3 WAR more). But over the same respective stretches, I have Reuschel's FIP-derived WAR at 46.9, while Friend's is 65.7. That launches Friend from completely uninteresting to just ahead of Reuschel (who, I believe I mentioned, I still like quite a bit). Altogether, I have Reuschel just .6 WAR behind for best 5 consecutive, but 5.9 behind for best 11 year consecutive.

I'm grateful for Bleed the Freak's latest post (#300) because it does a good job of clearing up where I'm coming from. My answer to his question about my preference for Bonds and Cedeno over Alomar should help to explain things further. To whit, I place a lot of emphasis on consecutive seasons. Given two players with equal WAR, I will like the one who is 5 WAR every season more than the one who alternates 3 and 7 WAR seasons. I think from a practical standpoint, the former is slightly more valuable when constructing a roster that will compete for a pennant. From a theoretical standpoint, the error bar around our estimate of the former player's true (demonstrated) talent is much smaller than for the latter. We may say our guess for both is 5 WAR true talent, and the former is almost certainly between 4.4 and 5.6 while the latter is between 3.8 and 6.2 (or something like that - I just made up those numbers for discussion's sake), so I have more confidence that the former is a great player. There's the slight possibility that the latter is merely solidly above average. I think that makes sense. Anyway, if we look at Cedeno, Bonds, and Alomar again, but look at consecutive 3, 5, 7, and 10 year periods, we get:

Cedeno 21.4 32.5 39.3 47.5
xBonds 18.1 29.7 39.3 50.5
Alomar 20.6 27.5 34.8 49.1

Hopefully that makes it clear why I'd have Bonds and Cedeno higher than Alomar. To pre-empt the question, I have Bonds higher than Cedeno because I give the most consideration to 5 and 11-year consecutive stretches, and Bonds really pulls away from Cedeno in that 11th year, 54.3 to 49.0. While I'm discussing it, I also give bonus points for reaching certain WAR thresholds within a season. Alomar and Cedeno each get 19 and Bonds gets 18. They're really all very close in my system. Hopefully that makes sense now.

Bleed/epoc: It doesn't appear that you're adjusting for differences in season length. Ned Williamson's 4.3 WAR in 1879 was earned in a league that played an 84 game schedule; Hugh Duffy's 7.6 WAR in 1894 was from a league that played 132 games. Almost all voters here think it's important adjust for differences in season length--both when comparing 19th century players to more recent players, and also to take account of the effects of strikes in recent years.

My mistake Brent. I did not realize that when I posted. If I have time, I'll run a new comparison with adjusted games played for Duffy.

Brent: You're right. I haven't yet decided on a way to adjust for season length. In lieu of that, I've been trying to compare short season players to their peers, but I think on Williamson in particular I may be missing the boat. I'll reevaluate.

302. epoc Posted: November 17, 2009 at 09:04 PM (#3390057)
<i>Secondly, [Gooden's] peak is really, really amazing. Someone suggested I should like Rosen if I like Gooden, but in addition to having a better, longer prime, Gooden's peak is way, way better. Even without considering FIP, their three best years are very close - 22.5 for Rosen, 22.3 for Gooden - as are their five best years - 30.4 to 30.1. Their three- and five-best consecutive years are similarly close, but Gooden at his best, in 1985, was so much better than Rosen at his best, in 1953. By Rally's WAR, that's 9.7 for Rosen and 12.5 for Gooden. The number of position players who've had a 10-WAR season since 1900 is at least 30 (I stopped counting). The number of 12-WAR pitcher seasons (not just the number of pitchers) since 1900 is five (two by Johnson). If you lower the threshold to 11, you only add three more. I'm not saying that one season is enough to warrant HoM status, but if you're a peak voter (which I'm not, strictly) that's a very formidable peak right there. Rosen's pales in comparison, in my opinion.

This a little mind-boggling, right?
First read the bold. Then read the underscore.
Back to the top and bottom lines of the bold:
Gooden's peak is way, way better. ... Rosen's [peak] pales in comparison,

Based upon WAR, I hope you give Robin Ventura and Willie Davis a look before settling on Hugh Duffy with your 15th ballot placement.

Ned Williamson was also mentioned as a 19th century candidate...who I have as a bubble candidate, outside of PHOM. Hugh Duffy falls short of my PHOM. Other 19th century players include: Fred Dunlap, George Van Haltren, and Jimmy Ryan.

Unadjusted
xxxDuffy - 18.0 27.5 36.3 44.4 49.6
Williamson 16.0 24.3 30.8 36.5 39.3
xxDunlap - 16.9 23.6 28.7 34.2 34.6
VanHaltren 13.7 21.1 28.1 37.5 46.2
xxxxRyan - 16.4 23.9 30.0 35.0 44.9

OPTIMISTIC:
Extrapolating Duffy and Williamson's WAR from average games played in a season to 162:
xxxDuffy - 22.0 33.6 43.6 52.8 58.7
Williamson 26.7 40.5 51.1 61.8 64.0
xxDunlap - 23.8 35.5 43.3 50.0 51.5
I have taken 20% from Dunlap's 1884 UA season, a clearly inferior league. Suggestions on an appropriate discount are welcome.
VanHaltren 15.7 24.6 33.0 43.9 54.1
xxxxRyan - 18.8 27.2 34.6 43.0 52.3

Taking 5% from extrapolated projection:
xxxDuffy - 20.9 31.9 41.4 50.1 55.8
Williamson 25.3 38.4 48.5 58.8 60.8
xxDunlap - 22.6 33.7 41.1 47.5 48.5
VanHaltren 14.9 23.4 31.3 41.7 51.4
xxxxRyan - 17.8 25.8 32.8 40.8 49.7

Taking 10% from extrapolated projection:
xxxDuffy - 19.8 30.2 39.3 47.5 52.8
Williamson 24.0 36.4 46.0 55.7 57.6
xxDunlap - 21.4 31.9 39.0 45.0 45.9
VanHaltren 14.1 22.2 29.7 39.6 48.7
xxxxRyan - 16.9 24.5 31.1 38.7 47.1

xxxDavis - 19.1 29.2 37.7 45.7 57.1
xVentura - 18.6 29.2 38.3 47.4 55.1
Ventura's 3rd best season came in 1998 expansion, with a 5.8.
Davis after discounting expansion seasons 1962 and 1969 by 15%, his 2nd and 3rd best seasons.
xxxDavis - 18.1 27.6 36.0 44.0 55.9

Kick Davis out of the conversation...Ventura is comparable to Duffy, depending on extrapolation method. Williamson looks awesome in this comparison! Dunlap has Duffy on peak, Duffy is better on career. Ryan and Van Haltren fall short in this comparison.

Does anyone else have thoughts on how to use Rally Monkey's WAR to reflect the value of 19th century ballplayers?
Am I in the ballpark, for extrapolation purposes?

Unadjusted Chone Smith's WAR, best 3, 5, 7, and 10 year stretches:

xxxLarkin - 19.5 31.2 42.1 54.8
xMartinez - 20.1 32.2 43.4 58.1
xxAlomar - 22.1 33.4 42.5 53.5
xxxxxBell - 19.2 30.6 38.1 48.8
xxxBando - 22.5 34.5 45.5 57.2
xxxBonds - 19.5 29.8 39.6 51.8
xVentura - 18.6 29.2 38.3 47.4
xCedeno - 21.8 32.5 42.0 47.5
xxxxCey - 18.1 28.5 36.5 46.2
xChance - 20.9 31.4 38.5 45.1

Does anyone else have thoughts on how to use Rally Monkey's WAR to reflect the value of 19th century ballplayers?

I guess the first step is to articulate the reasons you'd like to make adjustments to reduce the results from the simple extrapolations. Is it because you think the short-season data aren't representative and you want to regress them? Or is it an adjustment for perceived league quality (in which case you'd also want to adjust data from longer seasons)?

My reasoning would be that short-season data is a smaller sample size, and might not be fully representative, so regression may be necessary. Sure, the 1884 UA needs regression since the league quality was weak, along with the last couple or few seasons of the AA, but league quality has been fairly constant or similar in almost all other years...Epoc's take on using Rally Monkey's WAR is a new HOM voting stance to my knowledge, and one that may allow for broadening of our knowledge for a players merit.

Pennant is a pennant

I always discounted the non-NL 19C leagues as inferior in their own time. Not a timeline, not inferior to today. The best baseball at any time and place is a 1. But inferior leagues in the 19C < 1. The AA varies from 1 (no discount, equal to NL in '85-'86) to as much as a 20-25 percent discount the first two years and the last year. The PL is a 1. The UA is a .35 (yes, a 65 percent discount, Dunlap is the only player that matters, discount 1884 by 65 percent and you'll get a perfectly normal Dunlap season). This BTW is exactly the same logic that applies to the NeL and the WWII MLs. Other than that, however, I don't discount for league strength. The NL and AL are indisputably MLs and are each a 1. There needs to be a really obvious and compelling reason for a discount. (I never did anything with the FL though probably should have.)

But obviously all of that comes along with an extrapolation to 162 games. I don't do a full and direct extrapolation. The assumption is a player who plays 81 games in an 81 game season is not necessarily going to be able to play 162 games. So I take off about 10 percent from that.

But the answer to Brent is "both." You've got league quality adjustments for the non-NL and you've got season length adjustments. But they're two different things and you need to think about each one separately before you combine them into your calculations. I'm not saying my numbers are right, BTW, though they work for me.

Then there's also the pitchers. Before 1893, I also give the defense half of the value of the pitching. It's fairly arbitrary (the number) though conceptually defensible, and it gets 19C pitchers into a realm where they can reasonably be compared with post-1893 pitchers. You end up with a fairly normal looking career for Tommy Bond, e.g., though it's all compressed into half the seasons.

The bottom line for the 19C BTW is that I am still voting for Tommy Bond and Ed Williamson. Most of my other 19C faves have been elected, though Hugh Duffy has also been kicking around the edges of my ballot as well.

Which Davis is that, BTW?

On positional balance:

I've come out strongly in favor of adding more pitchers to the Hall of Merit. We've elected fewer pitchers to the Hall of Merit than the Hall of Fame (30% of HoF, not HoM are pitchers). I'd like to see a roughly 1/3 split between gloves, bats and pitchers. This is for many reasons that I have put forth before, but one of the strongest is probably roster construction. Major league rosters are 1/2 pitching nowadays. It doesn't make sense to me that 1/4 of the elected players would be pitchers if 1/2 of the roster slots and 55% of payroll goes to pitchers.

If you count CF/3B as half glove and half bat, here's the breakdown

LF/1B/RF + 1/2 CF/3B: 21+22+23+(22+19)/2 = 86.5
C/SS/2B + 1/2 CF/3B: 21+22+22+(22/19)/2 = 85.5
P: 65 (27%)

I've allocated 5 extra slots to pitchers in my PHoM which brings my percentage to 30%. That is still probably low.

The election of pitchers has actually slowed down in recent years with one pitcher (Saberhagen) elected in the last 5 years and three in the last 9 (11%). Here's a running percentage over history:

1900 2/8 (25%)
1910 7/23 (30%)
1920 11/37 (30%)
1930 15/56 (27%)
1940 22/75 (29%)
1950 27/95 (28%)
1960 30/116 (26%)
1970 36/135 (27%)
1980 45/157 (29%)
1990 52/181 (29%)
2000 60/210 (29%)
2010 65/240? (27%)

This election is certain to add 2 players to the "glove" count - Alomar and Larkin. The 3rd slot is certain to be a fight between a new pure bat candidate and a pitcher. I would just like to say I hope we err toward pitching.

65/240?
63/237 is now the official count, where Caruthers and Dihigo are pitchers; Ward is not.

I've allocated 5 extra slots to pitchers in my PHoM which brings my percentage to 30%. That is still probably low.

Why is it low? For what league are you building a balanced roster?

If 6 pitchers before Rusie (debut 1889) is above norm and/or 8 pitchers from the "Negro Leagues" is above norm, then the remainder probably falls below some reasonable norm. I would proceed along these lines and disaggregate at least once more.

For what league are you building a balanced roster?

I thought consideration of all positions was in the consitution: "the most worthy candidates from all eras and positions"

I wouldn't just ignore catchers, for example. We've balanced positions so only 3B is really underrepresented with respect to peers and even then it's just a guy or two short. However, for most of baseball history there have been 4 starting pitchers. In modern times that has expanded to 5 starting pitchers and an ace reliever. If you look at positional balance our "roster" currently has an even amount across all positions, and not quite enough pitchers to fill 3 starting slots and 1/4 of a bullpen slot. Right now I'm just encouraging people to consider completing those 3 starter slots (5 guys short), I haven't even pushed for the 4th starter (24 more pitchers). I think there's 5 guys as good as anyone elected, we don't even have to lower standards. Adding 29 guys would certainly lower the bar of what a HoM pitcher looks like.

For example, on my ballot I have PHoM Tommy Bridges, David Cone, Luis Tiant, Urban Shocker, Rick Reuschel and Dick Redding. I think all 5 are pretty darned close to or better than players we have already elected. However, to add 25 more pitchers means not only those guys but Dizzy Dean, Dutch Leonard, Jim McCormick, Virgil Trucks, Leroy Matlock, Hilton Smith, Nap Rucker, Bucky Walters, Vic Willis, Babe Adams, Ron Guidry, Dolf Luque, Tommy John, Frank Viola, Chuck Finley, Lefty Gomez, Wibur Cooper, Dizzy Trout, Frank Tanana, Jack Quinn, Wilbur Wood, Orel Hershiser, Lon Warneke, Noodles Hahn, Burleigh Grimes, Lee Smith and Addie Joss. At that point a "HoM caliber pitcher" and a "HoF caliber pitcher" become two different things. It think there is still value defining a "HoM" pitcher as approximately the same % as a "HoF" pitcher.

Right now we're in danger of being _more_ restrictive than the Hall of Fame for pitchers, not less.

sunnyday, the Davis I mention is Willie...solid career, very limited peak.

Prelim time, more than likely the final ballot but let me put it out in the discussion thread before the move to the other thread.


2010 ballot. Two new middle infielders, two new hitters worth mention.

Methodology in brief: The system used for my ranking entails a little bit of everything including WS, WARP, OPS+/ERA+ with Dan R’s WARP based material and Joe’s PA for pitchers helping to round out the picture. Of course you have positional adjustments, additions to one’s playing record for minor league service, war, and NeL credit and for our real oldtimers some contemporary opinion thrown in. Weighting the various measures smoothes any outliers and helps get my ordering. The results of this work tend to favor prime/peak players over career types but that is not 100% tried and true.

1. Barry Larkin (-). Face of the Reds franchise for well over a decade. Nice little run there between Concepcion and Larkin patrolling SS. 30/30 guy in 1996 and that wasn’t his MVP year.

2. Roberto Alomar (-). 2724 hits, 474 stolen bases, .300 career average out of a 2nd basemen is plenty to crow about. The number of Gold Gloves might be stretching it slightly but during the prime of career was a durable, hitting, fielding middle infielder.

-----There is a pretty decent gap here to the rest of the ballot-----

3. Dick Redding (2). Career was long – decent peak along the way. Outstanding fastball in his day according to James/Neyer book. So he didn’t get into the Hall of Fame; maybe the information collected by HOF committee wasn’t pertinent to Redding’s prime years. He deserves some WWI credit, thus patching up a bald spot in his prime years as 1918 and 1919 were affected. The last NeL pitcher I’d deem as worthy of induction.

4. Tommy Leach (3). Combination hot corner/centerfielder could field a little, hit a little. Second all-time in inside-the-park home runs to Wahoo Sam Crawford. Someone else stated he was uniquely valuable in his particular era and I agree he meant more in the particular era he performed in. Useless trivia: Still holds World Series record with 4 triples in a single series.

5. Bobby Bonds (6). Run on players from the 70s. Even with the constant trades, drinking problem and whatnot his combination of speed/power made him a very valuable player. He wasn’t the next Mays, or as good as his son, but we’re backlogging here. All five tools on display.

6. Norm Cash (4). Nice run from 1961-66 in terms of placing among the OPS+ leaders in the AL. Seems to be a decent glove to go with good on-base skills.

7. Phil Rizzuto (7). Probably wasn’t giving him enough credit for the WWII absence so I tried to adjust accordingly. Glove first but the offense during prime years was nothing to sneeze at either. Holy Cow!

-----My personal line where I’m no longer sure the folks below it have a HOM feel -----

8. Tony Mullane (8). Old time pitcher who threw plenty well, a good hitter to boot. Had some playing time issues since he missed seasons due to being blacklisted. He’s amongst the best of his era when accounting for the time outside of baseball due to conflicts with different leagues. Goes on the all-Nickname team as well.

9. Mickey Welch (9). 300 game winner in the house. Was it due to luck, run support, bad opponents? Still a feat to accomplish, sometimes I need to remind myself that and not totally overlook Smilin’ Mickey. Seemed to pitch well against the other front line starters of his day.

10. Fred McGriff (-). I see a nice prime 1988-94 before the silly ball era takes place. Adds on plenty of career length (60th all-time in games) who didn’t DH much. A very good hitter in the playoffs over many series, slight bump for that.

11. Bob Johnson (10). Argument in brief:

Batting Win Shares misses the mark on his value due to quality of teams he played on. They were horrible and likely cost 20-25 win shares over his 10 year prime with the A’s.

The teams he played on underperforming pythag wins vs. actual, thus a hit to Win Shares. Additionally his teams would end up leaving 2-4 decisions short per year. These incomplete games outcomes shorten Win Shares to divide up.

His career has war years that need discount. But also a couple years at the beginning of his career were in the PCL where he was more than major league quality. MLEs for 1931-32 show a player worthy of starting in the bigs. The tail of his career is nonexistent since the 1946 avalanche of returning War players pushed him back to the minors.

When he retired, Bob Johnson ranked eighth all-time in home runs. thebaseballpage.com lists him as having the strongest arm among left fielders, a sentiment echoed by Bill James in his historical Abstract.

For me he goes ahead of electees like Medwick, Averill, and Willard Brown from his era. Sorry Indian Bob, so close the one year but you’re not getting elected by this group in the near future.

12. Edgar Martinez (-). Some see the 147 OPS+ and go WOW! It deserves notice as does the fact he played very little in the field, wasn’t that level of hitter when fielding, and didn’t own an exceptionally long career. A favorite to watch hit, true pro to the craft.

13. Luis Tiant (13). Was less than the Carlton/Seaver/Niekro grouping of his time but got by on his funky delivery to merit seeding. Check out his 1964 PCL record in Portland: 15-1 with a 2.04 ERA. That deserves a callup I guess.

14. Bucky Walters (14). Short time NL prime starter who has the numbers downgraded a little by the defensive support which was above normal standards. Still high quality work, had a good stick to help his own cause.

15. Bus Clarkson (15). Both Clarkson and Rizzuto were SS who have extraneous factors to account for. This year I lean toward Scooter but both probably worthy.

The rest of the top 100. Its more for my purposes to keep names in my mind for each year’s consideration list.

16. Jack Clark
17. Orel Hershiser
18. Vic Willis
19. Hugh Duffy
20. Spotswood Poles
21. Lance Parrish
22. David Cone – What a mess we have with the pitching backlog. Lots of similar values out there. Cone was nice, pitched for a series of really good clubs along the way.
23. Tommy John
24. Carl Mays
25. Luke Easter
26. Lou Brock
27. Don Newcombe
28. Urban Shocker
29. Ron Cey
30. Ed Cicotte
31. Ben Taylor
32. Jose Cruz Sr.
33. Tony Perez
34. Lee Smith
35. Ed Williamson
36. Gene Tenace
37. Rick Reuschel
38. Jimmy Ryan
39. Thurman Munson
40. Bob Elliott
41. Burleigh Grimes
42. Tommy Bridges
43. Pie Traynor
44. Dolph Luque
45. Tony Lazzeri
46. Jack Quinn
47. Larry Doyle
48. Dizzy Trout
49. Kevin Appier
50. Cesar Cedeno
51. Wally Schang
52. Dave Bancroft
53. Bill Monroe
54. Orlando Cepeda
55. Wilbur Cooper
56. Ken Singleton
57. Bruce Sutter
58. Sam Rice
59. Addie Joss
60. Johnny Evers
61. Dizzy Dean
62. Dutch Leonard
63. Robin Ventura
64. Wally Berger
65. Leroy Matlock
66. Kirby Puckett
67. Buddy Bell
68. Jimmy Key
69. George Burns
70. Fielder Jones
71. Lave Cross
72. Lefty Gomez
73. Mike Griffin
74. Vern Stephens
75. Ernie Lombardi
76. Willie Davis
77. Kiki Cuyler
78. Chuck Finley
79. George Van Haltren
80. Lon Warneke
81. Albert Belle
82. Frank Chance
83. Tony Fernandez
84. Waite Hoyt
85. Gavvy Cravath – I think its coincidence that he’s next to Hondo but you make similar arguments for both. Sluggers who would’ve filled the DH rule very well. I’m probably not as liberal with the minor league credit as others are with Cravath.
86. Frank Howard
87. Fred Dunlap
88. Dwight Gooden
89. Sal Bando
90. Bobby Veach
91. Matt Williams
92. Frank Viola
93. Harry Hooper
94. Ron Guidry
95. Darryl Porter
96. Boog Powell
97. Ed Konetchy
98. Jim Fregosi
99. Dave Parker
100. Don Mattingly

New guys

Beside the four on ballot you’ve got Ventura who I have as similar quality to Buddy Bell. That was good when I oversold on the value of Bell but that gets you deep into the backlog now.

308. Brent Posted: November 19, 2009 at 09:30 PM (#3391965)
>> Does anyone else have thoughts on how to use Rally Monkey's WAR to reflect the value of 19th century ballplayers?

I guess the first step is to articulate the reasons you'd like to make adjustments to reduce the results from the simple extrapolations. Is it because you think the short-season data aren't representative and you want to regress them? Or is it an adjustment for perceived league quality (in which case you'd also want to adjust data from longer seasons)?</i>

309. Bleed the Freak Posted: November 19, 2009 at 10:39 PM (#3392000)
My reasoning would be that short-season data is a smaller sample size, and might not be fully representative, so regression may be necessary.

That may be reasonable regarding "peak" credit for seasons not supported by neighbors of about the same quality; that is, reasonable in "non-consecutive peak" analysis. I think it is generally unreasonable, along two lines (1,2). [Emigrate to the Uber-stats thread, probably the best place here?]

Before getting there, let me simply state what is a generally reasonable concern about the number of games in the championship schedule, or the "length of the season", for anyone who cares about "pennants" as well as games and runs. For the same league, same teams, winning percentage .625 may generate the same probability of winning a 126-game pennant race as does .600 in a 162-game pennant race. I presume that handling this point is a big part of "pennants added" analysis by Joe Dimino, following Michael Wolverton, if I understand correctly.

Al Peterson - I generally like your ballot but I'm concerned you're undervaluing modern pitchers. Your top holdovers are REALLY old and Cone/Tiant are just sniffing the end of the ballot. If you were a modern general manager would you trade Cone or Tiant for McGriff or Edgar? It's a lot easier to find a 1B/DH guy recently than it is an ace starter.

DL,

Yeah the holdovers are pretty long in the tooth. I regard my top holdovers are misses by the group, not huge misses but I would have selected differently. So they still score well in my system, thus high positions on ballot. I'm not saying we're missing any no-doubt HOMers, all have questions on the resume.

As for the pitcher question positions 10-20 are a lot of coin flips in terms of value. Tiant is next to McGriff and Edgar so that would be a fair trade if you want one-for-one. Cone, well we have Saberhagen, Stieb, and soon a whole slew of HOM pitchers just now retiring. I'm not going to rush to bring him in when there are the sure things and I include Hershiser rated higher.

DL...I agree that the pitcher distribution is light...I have 69 in my PHOM, including Dihigo and Caruthers as full hurlers. Regarding Paul Wendt's question, I see a handful of players from the 1880's as sniffing my PHOM, Dunlap, McCormick, Williamson, and Whitney. I consider their cases from all angles, and they have fallen just short. If I were to rate them based upon Chone Smith's WAR, I will fully extrapolate a players value over an entire season and evaluate the value, although from sunnyday's comments, it appears reasonable to deduct 10% from a straight extrapolation, and I will view a players career under that light too.

If/when I have something more substantial I will use the "Pitchers" thread.

Several participants one year ago, and several again this month, have observed that the Hall of Merit is short on pitchers. I don't recall much concern about this during the tri-weekly elections of 2002-2007. The Hall of Fame honors 74 pitchers among 236 players as we count them (official Players plus Wright, Spalding, Griffith, McGraw, and Foster); the Hall of Merit 63 of 237.

Here are some data on HOM pitchers.
63 of 237, HOM members
08 of 30, " " Negro Leagues and pre-NeL
55 of 207, " " major leagues and pre-majors

That's as close as possible 0.267 (=8/30) pitcher share. If the 2010 class is three major leaguers including one pitcher then we will have precisely sixteen times ^15 players including 4 pitchers^ with precisely two lots from the Negro Leagues and fourteen from the majors.

Hall of Merit pitchers by debut decade (majors or high level)


We may anticipate the prompt election of about seven starting pitchers from the mid- and late 1980s, two from the early 1990s.

I looked through the entire Discussion thread again --November in contrast to real time.

#18 and #21 are upbeat and downbeat notes from HGM and JWPF13 on the matter at hand, what's daunting or not about getting started.

Rusty Priske and DL from MN cast the first preliminary ballots (#37,38). DL is here and Rusty is here in a way. The two prelims from last winter/spring who are not here in November are HGM #154-159 (should get email notice, maybe not on the main list) and Chris Cobb #57. I'm inclined to sign up and cast CC's prelim ballot.
:-) or :-( ??



The ballot from JohnQ and it's discussion this morning really should be here in Discussion. For one thing or two, the same give and take with a new voter looks harsher over there, and its contrast with the rest of the Ballot thread may discourage other readers. Over here it fits. For example see #286-306 above.

I think the group should stick to this policy for newcomers and never or almost-never voters like Mike Emeigh or me.

The ballot from JohnQ and it's discussion this morning really should be here in Discussion. For one thing or two, the same give and take with a new voter looks harsher over there, and its contrast with the rest of the Ballot thread may discourage other readers. Over here it fits. For example see #286-306 above.

Paul, I agree. Let me see if I can figure out a way to 'move' them over here that is reasonable.

From JohnQ on the ballot thread:

As far as Jack Clark goes, I see him as very close to Will Clark who was a first ballot HOM. Both got stuck in Candlestick, both had career 137ops+, both had about the same win shares, Jclark-316, Wclark-331, both had about the same Warp-3, Jclark:58.2, Wclark:65.8. How does one guy (W.Clark) get in on the first ballot and the other guy (J.Clark) not get a single vote in last year’s ballot? That doesn’t make sense.

I've got a vast difference between Will Clark and Jack Clark. For one, there's a lot more to evaluating these guys than OPS+ - and those Win Shares/WARP numbers are bigger differences than you think - especially the WARP difference which is a full very good season. Will has 13% more WARP than Jack. That's a lot.

Here's how the Rosenheck system sees them season by season (I'll toss in McGriff too):



Crime Dog, and especially Jack never had years like Will did in 1988-89. That's the difference between them. Will has 16-18% more career value, with bigger years. That's the difference between HoVG and HoF.

And if you care about that kind of thing (I don't), you could also add that Will still had some good years in him when he left, while Jack and Fred played until they had nothing left.

I do think that JohnQ raises a valid point re: Jack Clark. He may not be the same category as Will Clark. But he has a very similar case to Bob Johnson. "Indian" Bob has been a top twenty returnee while Jack isn't even in the top one hundred. That discrepancy is pretty surprising (and I say that as a Bob Johnson voter).

Sorry to post and run this morning but I had, umm Sunday to do. I tried to be as constructive with the criticism as possible - this was my specific point:

"I'm guessing the biases in his sources of data against short-season players have not been compensated for properly when combined into his ranking system."

That's in direct response to JohnQ's explanation:

"I use Chone’s War, Warp-3 and Win Shares to make my list of players"

I know that all three of those systems are biased toward giving extra credit to players with longer season length. WARP3 has a blatant timeline. In general I agree with John Q's rankings for modern players (except having Tony Perez so darned high) and support several of the same guys.

My comments about the HoM voting this year in general deal more with losing the discussion factor (Why did they all wait until there were 3 days left? Why didn't they explain in more detail how they ranked the guys off ballot?) along with a big uptick in new voters using Chone's WAR. I like Chone's WAR, however, like everything else I think there are some systematic flaws that need to be corrected to use it well in a HoM context. None of these flaws have been discussed lately. The give and take is what makes this process work. Scrutiny from others has made my ballots better. I guess I expect new voters to challenge longtime voter ballots if they think they have something to contribute which is why I post so many prelims. Dan R and Brock Hanke haven't been shy about challenging the status quo since joining the project.

"they’ve appeared on 80-100 ballots already and they still haven’t finished in the top 3"

The 'top 10 returning guys' did finish in the top 10 though, which is more than you can say about Jack Clark who was pretty resoundingly rejected for the past several years.

Oh, and we've all tried to change karlmagnus but we pretty much just accept that there's going to be an outlier opinion based on a consistent but baffling process he has.

Regarding the 100 ballots thing - yes it's been 100 ballots for a couple of these guys - but at the same point, it's also been just 7 real life years - we've learned a lot during that time, and as new methods emerge, all players need to be reevaluated by those methods.

for newcomers, this is a good vibe of the "discussion" we seek.

Thankfully we've had dozens of robust new voices over time, who not only make good points, but who further inform our own analysis.

I suspect we might seem too concerned about "process" as in when a ballot is submitted, but the real observation is on trying to maintain a consistent approach.

I'm the old guy who cast the first vote in the first election (by fluke), but still I enjoy the new approaches. And whatever agita might or not arise in this voting, I am not concerned about it affecting this overall process. More likely to me, it improves next year's voting.

The 'top 10 returning guys' did finish in the top 10 though, which is more than you can say about Jack Clark who was pretty resoundingly rejected for the past several years.

Not that I represent the group's consensus by any means (nor have I remained active since we moved to once a year, but I have no problems with Jack Clark as such. Prior to JohnQ's ballot, I may have been Clark's best friend, placing him in the 20-25 spots on my extended ballots. Much higher than Leach, BTW.
That being said, it would be nice to get more as to JohnQ's reasoning for so thoroughly denouncing the old-school players.

First off, I didn’t know that the protocol was to discuss these topics on this thread, my bad.

If I had known this was going to be this much aggravation I wouldn’t have bothered with a ballot. I thought this was a pretty open-minded group of people open to a lot of different perspectives and opinions. I’m in my 40’s and I have less patience for this type of stuff. This is the kind of stuff that turns people off to sabermetrics. If this is the way you greet first time posters then your greeting skills have much to be desired.

The instructions for the ballot were pretty open ended: pick 15 players, top 3 get in, keep an open mind to all time periods, and elect the 15 best eligible players. There was no mention that WAR or WARP 3 or Win Shares or Ops+ etc were crap stats and shouldn’t be used unless I was willing to spend 80 odd hours making esoteric adjustments. There was no mention that I should contact DL-MN and ask him who I should put on my ballot or that I would ruin the integrity of this institution by putting Fred Mcgriff, Jack Clark or Robin Ventura on a ballot. But after reading several posts that was the message I was receiving.

The first post was from DL-MN, who basically implied that I clicked on the Prospectus web site and just listed the names that appeared under WARP-3. Then there was an implication that I and other first time posters were going to severely ruin the integrity of the HOM . Greetings from the Baseball Think Factory.

As far as Jack Clark goes, I not saying he was as good as Will Clark, Will Clark was clearly better but I just wanted to point out that Jack Clark isn’t that far away. Take career WAR, WClark is ranked 121, J.Clark is ranked 138. Or Take career Win shares, WClark is ranked 114 in the 2002 book and JClark was ranked 138. Will Clark gets in on the first ballot and Jack Clark didn’t even receive a vote last year out of 100 odd players? To me that doesn’t make sense. I don’t even like Jack Clark.

As far as Ventura/Leach: I made a list of 60 candidates and ranked them by WAR-WARP-3 and Win Shares. I took each player’s career value added it with his best 7 years and divided by 2. Then I ranked all the players. Ventura finished 17/60 in War, 5/60 in Warp-3, and 32/60 in Win Shares. Leach finished 39/60 in War, 44/60 in Warp-3, and 9/60 in Win Shares. Warp 3 was the big difference for Ventura. I then used ops+ to make slight adjustments or considered 5 year spans in the case of Bobby Bonds. Most of you probably think its crap; sorry that’s what I used I didn’t have a slide rule, abacus, or weegie-board handy to make adjustments.

Again to clear something up, I never denounced any of the pre-1920 players. I simply made a list of the top 15 eligible candidates from my point of view and no one eligible from pre-1920 made my list.

To me, there should have been more people chosen per year so you wouldn’t have such a backlog of players. It’s the 2010 ballot and there’s still players from the 1890’s-1900’s on many ballots. I’m not saying that these players don’t deserve to be elected; I’m saying that after 80-100 ballots their fates should have been determined already. Maybe there should have been 5 choices per year or a run-off election where everybody picks 25 players from pre-1920 and the top 12 players get elected. Then after that people can only choose players from post-1920.

The problem with your current system is it created too much of a backlog of players from the 60’s-90’s because your still trying to decide the fate of 20-30 players from the dead ball era. Seriously, why did it take John Mcgraw 100 odd-ballots to get elected?? What are guys going to be 70 year old men arguing whether John Olerud belongs in the HOM or not??

Again it was an open invitation to interpret the ballot by my own method. If you don’t agree with other peoples’ opinions, perceptions, methods or the measurements they use, then just keep it a closed system with 20-30 voters that meet your criteria. I have no problem with that.

If I had known this was going to be this much aggravation I wouldn’t have bothered with a ballot. I thought this was a pretty open-minded group of people open to a lot of different perspectives and opinions. I’m in my 40’s and I have less patience for this type of stuff. This is the kind of stuff that turns people off to sabermetrics. If this is the way you greet first time posters then your greeting skills have much to be desired.

John, you are taking this way too personally. The whole purpose of this group is discussion. If you aren't willing to discuss your methods, if you think that's 'aggravation' then maybe this isn't the place for you. That's the whole purpose behind the requirement of explaining a ballot. It's not some personal vendetta against you.

And people aren't just going to take your word for it, they are going to challenge you, and expect a reasoned response. That's not aggravation, that's discussion.

The first post was from DL-MN, who basically implied that I clicked on the Prospectus web site and just listed the names that appeared under WARP-3. Then there was an implication that I and other first time posters were going to severely ruin the integrity of the HOM . Greetings from the Baseball Think Factory.

I'm pretty sure you couldn't possibly have mischaracterized DL's post more than you just did. He simply pointed out something that looked suspect (not having any pre-WWII players on your ballot) asked for an explanation. That's the way things go around here.

When that happens, you are expected to provide some evidence and explain yourself. It doesn't have to be an uber-stat, but a specific reason for why you chose this player over that player. Something more than a generalized nothing comment like I think he was a better player. Why do you think that?

As far as Ventura/Leach: I made a list of 60 candidates and ranked them by WAR-WARP-3 and Win Shares. I took each player’s career value added it with his best 7 years and divided by 2. Then I ranked all the players. Ventura finished 17/60 in War, 5/60 in Warp-3, and 32/60 in Win Shares. Leach finished 39/60 in War, 44/60 in Warp-3, and 9/60 in Win Shares. Warp 3 was the big difference for Ventura. I then used ops+ to make slight adjustments or considered 5 year spans in the case of Bobby Bonds. Most of you probably think its crap; sorry that’s what I used I didn’t have a slide rule, abacus, or weegie-board handy to make adjustments.

Seriously, your tone is WAY out of line. This isn't some flame board.

When you post something like that it's great, but it doesn't end there. You could expect someone to point out that hey, WARP3 has a pretty nasty timeline (which kind of goes against the being fair to all eras concept) and has a very poor adjustment for shorter schedules (ditto). They'd ask what version of WAR you are using and again - does it adjust for schedule length? They'd also mention that WS has a super low replacement level and ask if you thought about any of that at all.

And of course WARP3 is going to favor Ventura, it's got a massive timeline and poor schedule adjustment. Did you take that into account? If not, any reason why?

Those are entirely reasonable questions to ask, and points to bring up. No one is expecting you to write your own uber-stat from scratch. No one is going to say your adjustments are crap. But they are going to ask if you made any, and if not why?

Again why are you taking this so personally? How dare we question you? Seriously?

I’m not saying that these players don’t deserve to be elected; I’m saying that after 80-100 ballots their fates should have been determined already.

No, their fates haven't been determined. John McGraw got in last year, for example.

Seriously, why did it take John Mcgraw 100 odd-ballots to get elected??

Because he was a borderline candidate. Combine this with the fact that some improved metrics were developed that put him in a more favorable light, the electorate changed a little, and it was a weak year. Perfect storm. But his getting in isn't a bad thing for the Hall of Merit it's a good thing.

The problem with your current system is it created too much of a backlog of players from the 60’s-90’s because your still trying to decide the fate of 20-30 players from the dead ball era.

Some of us consider that a virtue of the system. Are there more players (per team, don't forget about expansion) from the 60's-90s in our backlog than from the 00's-30's? Especially when you consider that they've had far fewer years on the ballot, so you'd expect a few more? And if John McGraw is better than that backlog he goes first, and they wait. Reggie Smith went in with him, so it's not like we aren't electing borderline players from the 60's and 70's also.

Again it was an open invitation to interpret the ballot by my own method.

Sure, and by posting a ballot, that's an open invitation to have some of us ask questions about that methodology.

If you don’t agree with other peoples’ opinions, perceptions, methods or the measurements they use, then just keep it a closed system with 20-30 voters that meet your criteria. I have no problem with that.

It has nothing to do with 'agreeing'. It has everything to do with understanding those opinions and asking them to be explained.

We generally have around 100 players get votes, with only 15 player ballots, so it definitely isn't that we all agree with each other. We all have very different opinions and when someone asks how we came to a conclusion, we explain it.

I hope you stick around, and if you do, you need to tone down your attitude a little bit, and expect to discuss things and have your assumptions challenged.

I think you're overreacting, Mr Q. These people are just trying to have a discussion. And the HoM been an evolving process over the course of years; they're trying to explain it. Furthermore, I highly doubt you're the only person over 40 around here.

I'll admit that the HoM is an esoteric thing, a bit of a difficult nut to crack. I've never participated specifically because the standards are high: I don't have the energy to quantify older players, and I feel that there's a certain duty to try to do that, and abide by the rules as they've evolved and as they are stated. I think, however, that it's good that the HoM is a bit intimidating. It's an enormous project, encompassing the entire scope of professional baseball, from long before any of us were born, and including leagues and players that no one here can possibly have seen -- let alone the dozens and dozens of players that even the oldest of us are unlikely to be familiar with even from youth, for reasons structural (Negro Leagues, for instance) and personal/locational (given the paucity of posters over 60, we're unlikely to have representatives of all baseball cities pre-1950). It's a difficult task, best tackled by people who have the time, energy, and interest to do so.

The HoM was conceived as a bit of a corrective to the Hall of Fame, with it in mind that anybody can vote -- but not completely without barriers or gates; though all a person need demonstrate to vote is the dedication to fairly consider all eligible players (more, by the way, than the Hall of Fame requires), that consideration is not easily performed. That is as it should be. Otherwise the HoM would be full of Jay Buhners and Jim Rices.

Not that I am trying to imply that you are a dunderhead or a dimwit who would simply pack the list with your favorite players, and I think that that is not what anyone here, even those by whom you might feel attacked, is trying to do. HoM discussions can be heated at times, but they usually lack the ad hominem silliness that must be waded through (and to which, no doubt, I contribute more than my share) in your average BBTF thread. I think you've misinterpreted a certain understandable care for standards arrived at through a years-long process and genuine, constructive criticism -- criticism which does, in fact, brook disagreement and acknowledge the possibility of differing opinions -- as an assault on your intelligence or understanding of the game.

The truth is that if you, like I, do not have the time or energy to engage the subject as deeply as the people who participate in this process, then perhaps it is best that you, like I, remain a spectator. It would be a shame, however, if you abandoned the idea altogether without considering investing the effort it requires, because it is obvious from your posts that you have invested thought and discernment in your ballot. If I made one, it would look much like yours, though of course agreeing with me on player evaluation is not necessarily the hallmark of intelligence or correctness. I would urge you, however, to re-read these posts with it in mind that this not the tubes-flooded environment often found elsewhere on this site. A lot of them are done in a calm and reasonable tone, to wit:

I suspect we might seem too concerned about "process" as in when a ballot is submitted, but the real observation is on trying to maintain a consistent approach.

I'm the old guy who cast the first vote in the first election (by fluke), but still I enjoy the new approaches. And whatever agita might or not arise in this voting, I am not concerned about it affecting this overall process. More likely to me, it improves next year's voting. --Howie Menckel

And:

I know that all three of those systems are biased toward giving extra credit to players with longer season length. WARP3 has a blatant timeline. In general I agree with John Q's rankings for modern players (except having Tony Perez so darned high) and support several of the same guys. . . . The give and take is what makes this process work. Scrutiny from others has made my ballots better. I guess I expect new voters to challenge longtime voter ballots if they think they have something to contribute which is why I post so many prelims. Dan R and Brock Hanke haven't been shy about challenging the status quo since joining the project. --DL from MN

Those are just two examples, one from the man who you seem to feel has attacked you. I would argue that these are guys seeking to keep the process vigorous as well fresh, not make you feel dumb or unwelcome. The fact is that the HoM is not built to be the kind of thing that turns people on to sabermetrics; it is a difficult and esoteric undertaking that seeks to combine the best of what expertise has to offer with the best of the old wisdom of crowds axiom. What it certainly does not do, however, is adhere to any orthodoxy, or discourage reasoned disagreement. One might say that reasoned disagreement and discussion is the foundation of the whole idea.

Very well said Voxter. You would have saved me some time if you'd gotten that up there a few minutes earlier :-)

...then just keep it a closed system with 20-30 voters ...

I had a longer post in mind, but Joe and Voxter already said most of it. Just accept that being challenged is not the same as being attacked. We've all been challenged over something or other, some time.

During our two weeks or three weeks per year run, we normally had 45-50 voters, and with some people moving in and out, I have a list of 95 different voters that participated after 1921, so the whole list might be 100. Yes, it seems to have dwindled some since we reached "real time" and slowed down. That's only to be expected, as habits change and people move on to other interests. (But where are you, Chris Cobb? And KJOK? Would you consider voting again, Dag Nabbit?)

If it's really down to 20 or 30 (and I'm hoping to do better than that), then that's a bad thing - it concentrates our eccentricities (and face it, we've got some pretty eccentric long-timers) and it narrows our viewpoints.

I do still have some 1890's outfielders on my ballot - but I can see that in a couple more years, enough strong new candidates will appear to swamp them off the back of my ballot. We're about to start to develop a new backlog heavy in recent players.

Edit: And Brent's ballot makes 25. We're going to at least crack 30 - can we go for 40 or is that too much to ask? But really - Edgar in the #1 spot? That's, uh, different.

Regarding Brent's ballot - you've got Edgar and 67.2 and Larkin at 68.8 WAR. I'm curious as to how you got there.

DanR has 85.2-56.6; WARP 121.8-104.4 and WS 346-305.

Those very different metrics have Larkin 13.7, 17.4 and 28.6 wins better. But you get them almost even. Seems strange, just curious as to how.

Brent is taking CHONE's WAR numbers straight. The massive difference between my evaluation of these two players and CHONE's has entirely to do with our positional weights (and I am stunned to see its magnitude).

I have Martinez at 50 wins above overall league average for his career, CHONE has him at 54--close enough. Similarly, I have Larkin at 33 wins above overall league average for his career, while CHONE has him at 31; even closer. But here's the whopper: I say that a replacement player with Martinez's mix of positions and playing time would have been 7 wins below overall league average, while CHONE says he would have been 14 wins below overall league average. And I say that a replacement player with Larkin's mix of positions and playing time would have been 51 wins below overall league average, while CHONE says he would have been 37 wins below league average. So we're talking about a 21-win swing just based on how you weight SS versus DH.

To remind voters how I derive my weights (and thus, implicitly, argue that mine are correct and CHONE's are not), Nate Silver found in his 2006 Freely Available Talent study that the minor leagues are just about overflowing with guys who could be MLB average hitters but can't play the field, while replacement shortstops not only are the weakest hitters among replacement players at any position but also are below-average with the glove. Thus, replacement level for DH is league average, while for SS it's about 3 wins a season below average (2.5 with the bat, 0.5 in the field). Moreover, remember that Larkin played the first half of his career before the SS offensive explosion--he was a contemporary of the Alfredo Griffins and Felix Fermins of the world, guys who would never get a look today.

Take a look at the AL league positional splits at bb-ref. DH almost always has lower average offensive production than 1B, often significantly lower. There may be a lot of minor league players who can hit, but there's also evidence that most players don't hit as well when they DH. During Martinez's tenure the Mariners were getting a lot more offense from their DH than most other teams.

the minor leagues are just about overflowing with guys who could be MLB average hitters but can't play the field

But Martinez isn't an average major-league hitter. Martinez is almost half as good again as an average major-league hitter, going by OPS+.

There are two statements here - 'Martinez is a good hitter at a position with a high replacement level'. 'Martinez has a 147 career OPS+'.

I'm as big a fan as anyone around here for 'good-glove, no hit' types, probably moreso. If a guy gives otherworldly defence, and something like league average for his position hitting, I'm interested. (Thus, I would put more emphasis above on 'overflowing' and 'but can't play the field'. The majors know something maybe we don't?)

But on an absolute scale, Martinez is a bloody good hitter. For me, the worrying words that follow are 'in a strong era for offense' not 'at a position with a high replacement level'. But Larkin also inhabits much of the same era. (He gets a little boost, for me, from playing in the NL's 'little deadball era' of 1988-91.)

On positional balance:

Not picking on Rob Wood in particular but he makes my point. He posted a ballot with 1 pitcher in his top 20. The guy he supports (Tommy Bridges) is one of my favorites, however it is unreasonable to believe this reflects positional balance. Maybe the next 5 guys are pitchers in his system but 5% is too low of a percentage.

Also, with Chone's WAR there are a couple issues with positional balance. I haven't looked into the sort of catcher bonus it requires, not that it is a factor this year. However, his system loves modern 3B. I don't think we've been unfair to 3B post 1970; we've elected Boggs, Brett, Evans, Molitor, Nettles and of course Mike Schmidt. Personally, I also have lukewarm support for Ron Cey who rounds out my PHoM. If you go strictly with Chone's WAR though it spits out that Sal Bando, Buddy Bell, Ron Cey and Robin Ventura all belong. This is also one of the reasons it really likes Edgar Martinez. I think voters using this system need to ask themselves if they believe we're in the golden age of third basemen or if Chone's WAR might be a little off kilter.

Sure players hit a little worse as DH's, but that's because those guys aren't regular DH's, and generally the guys that occasionally DH are doing so when they are already tired as a way to get a day off while keeping their bats in the lineup.

And IMO, the main reason that DH's hit a little worse than 1B is because they aren't as good as baseball players overall. Most that can hit can also field, at least well enough to play a passable 1B or LF. To say that a player is more valuable to a team as a DH than he would be as a 1B, even a bad defensive 1B makes no intuitive sense.

So what do you guys think? Would it be beneficial to extend this a week? In the past that has worked out well, are we really losing much in terms of getting out in front of the Hall of Fame balloting? If this would make the difference between you submitted and not submitting a ballot, please let us know . . . I could go either way here.

I will say that adding previously elected pitchers like Brock did is not the way to add more pitchers to your ballot. Thanks Brock for the most unintentionally funny ballot of the year.

I don't see any big reason to extend a week unless we get a couple people who say they're using the holiday weekend to catch up.

It is worth noting that Sean's system applies a "DH bonus", of I believe around 5 runs per "season". I don't remember how he defines seasons, but that could account for around seven wins worth of discrepancy.

But Martinez isn't an average major-league hitter. Martinez is almost half as good again as an average major-league hitter, going by OPS+.

Of course. Dan isn't really talking about Martinez here; he's talking about the replacement level for DH, and how it affects Martinez. Combining the position adjustment and replacement columns, it looks as though AROM uses a DH replacement level that's about 7 runs below average per full season (and since DHs don't field, that's all in hitting/baserunning), while Dan R uses one that's exactly average. That makes a big difference in Edgar's case.

Dan, did Silver's study consider how well the non-fielders run the bases, or avoid double plays? I imagine they'd be worse than league average at both.

It remains curious to me, to say the least, how DH is considered to be a full position, in and of itself. I am not at liberty at the moment to check the actual numbers, but IIRC, no more than 5-6 AL teams used a more-or-less fulltime DH this year. The rest used the lineup spot as a repository for regular hitters getting a day off from the rigours of fielding. Like with relief pitchers, there would seem to be a sort of chaining effect vis-a-vis replacing DH's.

JohnQ:

To me, there should have been more people chosen per year so you wouldn’t have such a backlog of players. It’s the 2010 ballot and there’s still players from the 1890’s-1900’s on many ballots. I’m not saying that these players don’t deserve to be elected; I’m saying that after 80-100 ballots their fates should have been determined already. Maybe there should have been 5 choices per year or a run-off election where everybody picks 25 players from pre-1920 and the top 12 players get elected. Then after that people can only choose players from post-1920.

The problem with your current system is it created too much of a backlog of players from the 60’s-90’s because your still trying to decide the fate of 20-30 players from the dead ball era.

Yes, the Hall of Merit should do more to welcome newcomers and introduce them to HoM protocol. The pertinent information is there now, but you have to dig for it. I suggest you, or anyone, spend some time in the HoM archives.

Regarding eligibility limits for candidates, there was a discussion of that back in 1951: Term Limits or Perpetual Eligibility for Candidates? With that proposal for fallowing players' candidacies, most players retiring before 1935 would not presently be eligible; only the better candidates from before then would be the players still eligible now. Lesser candidates would still come up for eligibility every 20 years. That proposal was scarcely considered.

I would recommend for newcomers who are unfamiliar with the deadball and 19th century players, you could simplify the job by confining your studies to those candidates doing well in the voting. These seven: Gavvy Cravath, Tommy Leach, Dick Redding, Hugh Duffy, George Van Haltren, Vic Willis and Mickey Welch.

All else having been said, I think we should accept JohnQ's ballot. Snark aside, he did what we asked him to do. Whether he wants to participate again in the future is up to him.

I would also support moving epoc's ballot over for him if he fails to do so.

The other new voter that I'm aware of, mystikx20, did not post a prelim but he responded to questions to obviously his counts.

My point with epoc is he did it more the right way than any of the other new voters. They missed stuff and their ballot will count. If epoc misses the transfer, he still did it more the right way and it should count.

PS. I'm OK with extending the ballot for a week. Getting some more ballots is more important than a rigid adherence to protocol.

PPS. But then can we do the MVP Project ;-)

Finally on positional balance.

Pitching today is equally important as it was any other time. That hasn't changed. However, pitching value is spread amongst more pitchers who pitch less. And since we're electing individual pitchers, not "pitchers" as a class, I think it is entirely reasonable that we might elect fewer pitchers from the modern era than we did from previous eras.

By analogy, catchers are no more important now but we're electing more modern catchers than old-timers. Why? Because individual catchers are on the field more and are accumulating more value.

If we should have to elect modern pitchers in the same numbers as before, but at a lower threshold, well, shouldn't we have elected more catchers, at a lower threshold, from 1890-1945?

positional balance?
I can't agree with fra paolo and DL from MN that every ballot should reflect positional balance, certainly not as a matter of policy. If two people disagree 10% on relative value of catchers and shortstops, say, then after 100+ years the HOM may have elected all of the shortstops from the would-be top ten of one person's backlog and may have passed over five catchers who now appear on the other person's ballot.

I would like to learn more about fra paolo's pitcher system, including how one does Fielding Independent Pitcher Earned Runs (that's the foundation, I infer).

I do think it's unusual that Rob Wood's ballot now "leans" as it does (one pitcher in twenty) but that's empirical not principled. I observe that the HOM is near the "few pitchers" end of the spectrum. Cooperstown and other shadow halls on the web all recognize pitchers at a higher rate than we do here. Someone whose PHOM would have several fewer pitchers is therefore a "surprise" and someone who would go the other way is "no surprise". In the New BJHBA (2001), however, Bill James put only twenty pitchers in his top 100. If we had a lot of voters who followed Bill James rankings (not the same as using win shares somehow), maybe we would have 50 pitchers instead of 63 and Rob Wood would have ten pitchers in his top twenty at this point in the process.

Brent--that's because most teams don't have a full-time DH, but instead use the slot to rest players who are tired and/or hide players who are semi-injured. If you limit your search to the aggregate production of *full-time* DH's, which is what Edgar was, you will see it is far higher than any other position and right in line with what you would expect from a replacement level of league average (around 2 batting wins above average a season).

Blackadder--Oh really, I didn't know that. Well, I *very* strongly disagree with giving DH's a bonus for the supposed difficulty of having to hit without playing the field. That will make CHONE's system far, far more favorable to DH's than mine is.

Eric J--It did include the SB/CS of DH's, but not their non-SB baserunning and DP avoidance. Yes, that should drag it down a bit, but we're talking what 1-2 runs a year tops? It's a feather on the scale, nothing more.

I never said every ballot should balance positions, just that your ranking system overall should. I think it's a constitutional requirement. It takes a while to get to a catcher or a centerfielder on my rankings (for example) because we've elected all the marginal ones. Once you hit that C or CF they start appearing about every 10-15 slots though. The normal distribution is pretty amazing.

I admit I boosted my ranking of Roger Bresnahan precisely because he was an outlier wrt his era. If he had put up the same value in 1970 I wouldn't have supported him.

the supposed difficulty of having to hit without playing the field

I bet a portion of this is using the DH to rest players. Players who are being rested are usually fatigued or injured and are therefore more likely to have a bad set of plate appearances.

The language I've always liked is that DH is a role, but it's not a position.

If you were forming some hypothetical all-star team from some time or some locality over time, and you decided to make it a team for a DH league, most of the time your DH would be an extra corner OF or extra 1B - the criterion would simply be how well he could hit. That's a role, not a position.

I similarly think that relief pitcher is a role, not a position. The position is pitcher.

For both DH and relief pitcher, I don't think we're under any affirmative mandate to include so-and-so many such players. If a DH hits enough to stand up to comparisons to 1B and corner OF, overcoming some allowance for lack of defensive value, then we can include him. And if a RP's better rate performance and leverage makes him as valuable as a HoM starting pitcher, then we can include him. But we don't have to include either just to say we have the role represented.

All else having been said, I think we should accept JohnQ's ballot. Snark aside, he did what we asked him to do. Whether he wants to participate again in the future is up to him.

I agree - I don't think we were seriously considering not counting it (assuming he responded, which he did), were we?

I am fine with extending a week . . . I've had one member specifically send in a request, and it will give epoc (and others more time).

Any opposition?

Pitching today is equally important as it was any other time. That hasn't changed. However, pitching value is spread amongst more pitchers who pitch less. And since we're electing individual pitchers, not "pitchers" as a class, I think it is entirely reasonable that we might elect fewer pitchers from the modern era than we did from previous eras.

I tend to disagree. While this is true, amongst their peers, great pitchers are just as important - and they have just as much importance in the postseason as ever (see Schilling, Johnson 2001, Sabathia or Cliff Lee in 2009, Cole Hamels in 2008, etc).

It's a different time and pitchers are asked to do different things and they are paced differently. But they are still just as important and the top level pitchers should still be recognized as they always have been.

You don't see teams talking about breaking the bank for a SS at the all-star break. You do hear this when it comes to a star pitcher.

By analogy, catchers are no more important now but we're electing more modern catchers than old-timers. Why? Because individual catchers are on the field more and are accumulating more value.

I admit I boosted my ranking of Roger Bresnahan precisely because he was an outlier wrt his era. If he had put up the same value in 1970 I wouldn't have supported him.

As a follow up to my last post, I think we may have been a little too tough on the old-time catchers also. I would not have a problem if we elected someone like Wally Schang, for example (although he hasn't quite moved up my ballot recently either).

So I agree with the latter approach.

For both DH and relief pitcher, I don't think we're under any affirmative mandate to include so-and-so many such players. If a DH hits enough to stand up to comparisons to 1B and corner OF, overcoming some allowance for lack of defensive value, then we can include him. And if a RP's better rate performance and leverage makes him as valuable as a HoM starting pitcher, then we can include him. But we don't have to include either just to say we have the role represented.

I'm with you on DH, but not on relievers, or relief aces more specifically.

The key difference is that DH's are there because they cannot do something (field or stay healthy).

But relief pitchers are specifically put in that role because their skill set is suited to it. Relief Ace hasn't meant 'failed starter' for a very long time. Heck, they take good starters like Papelbon and make them closers.

DH however, is 'failed fielder'. Relief Ace is different than relief pitcher. I feel like DH is much closer to a role than a position - I think that relief aces (and good setup men) are the opposite. I realize there's room for interpretation here, and it's hard to put the difference into words (for me anyway), but it's there.

For one, they don't give out a Rolaids Designated Hitter Award. :-)

Hey guys,

I'm not sure where to ask these questions, sorry if this is the wrong place. I have a few questions about what is and isn't allowed. I'm not trolling, I swear, and sorry if any of this sounds accusatory -- am thinking about voting in 2011 (was thinking about this year but did not have time to do the work I'd want to do ahead of time), but not going to do various kinds of work if I'm not allowed to. I'm sure much of this has been covered in a thread somewhere, and apologize for repeating it. Anyway, without further ado:

The constitution seems rather open-ended on the issue of what one is and is not allowed to consider: a reference to "best", and then this sentence: "Voters are strongly encouraged to consider only a player’s on-field accomplishments and other factors which had an impact on the outcomes of the player’s baseball games," along with some notes about eras. Perhaps this is on purpose. Succinctly: is the point of this to find the "best" players, or the most "valuable" players?

For example, it seems likely that Barry Bonds was one of the best players on the planet, certainly above replacement level, in 2008. The same is true, possibly for very different reasons, for, say, Will Clark, 2001, or Joe Jackson, 1921. These players obviously accumulated no "value," but is it allowed to give them credit for being among the "best" players on these years (leaving aside the question of the extent to which that was actually true)? Obviously something like war credit and Negro League credit is very commonplace here, but I'm not sure if those are viewed as special cases, or if they are just an implication of the desirability to find the "best" players.

A similar question is true for minor league credit, especially contemporary minor league credit. It seems likely that Ryan Howard in 2004 would have been above replacement level in the majors, regardless of whether or not it was reasonable to expect him to have been in the majors going into 2004, not to mention Edgar (who I noticed did get MiL credit from a few people.) What about Sadaharu Oh? Is he getting no votes because it's not allowed, or because he is not viewed as as-good-a-baseball-player as the other people on people's ballots?

Some other questions about whether or not something is allowed:

-- Is it allowed to rely on contemporary (subjective, though possibly formed by individuals based on statistics) opinions of a player's value? This could be quantified in many ways: (actual) all-star appearances, awards voting, who the player was traded for, salary, HoF voting, usage patterns.

-- Is it allowed to give players in pennant races extra credit? In the extreme, would it be allowed to use a version of PPA (Pennant Probability Added)? What about a rate stat for this? (This would have the practical effect of weighting players' early-season stats more if they played for bad teams or dominant ones, and less if they played for a team in a pennant race at the end of the season.)

-- Is it allowed to consider manager usage or context-dependence of stats? For instance, consider the Koufax/Drysdale Dodgers. There were times when both were rested but only one could start. Perhaps over the course of their years, either would have accumulated a few more starts if on another team. Similarly, a low IP total could be adjusted for for a SP on a team with an exceptionally good bullpen. Another example of this is a player who batted low in the order due to teammates (e.g. Davey Concepcion); is it allowed to adjust such a player's counting stats upwards? What about downgrading walks for #8 hitters in non-DH leagues? (Not that many #8 hitters are serious HoM candidates, but you get the idea.)

-- The HoM timelining has a recommendation built in based on the number of elected players each year. Is it allowed to believe different things about desirable relative representation of era? Obviously this is a controversial topic; proportional representation could be done in any number of different ways (e.g. constant # per year, proportional to number of players in MLB, proportional to size of "talent pool," proportional to interest in baseball (e.g. attendance numbers), etc..) I'm not talking about "are contemporary players just better than old players as per WARP3", but rather "what should the representation of the HoM be" -- is that something that an individual can vote based on their personal opinions of?

-- Suppose that a player is believed to have an actual effect on his teammates' performance. Is one allowed to consider this? Does it matter if it is a subjective opinion or a quantitative one (e.g. a study of performance before-or-after)? Does it matter if there is a known mechanism (e.g. sign-stealing) involved? What about a negative effect, e.g. introduction of cocaine to the clubhouse, or recruiting them to a game-fixing organization? I'm not talking about character issues (e.g. steroids) -- I'm talking about situations where a player affected by his actions the performance of his teammates, translating into an effect on his team's record. (Indeed, if a player were responsible for steroid-enhanced performance of his teammates, this would be a plus.)

-- Is it allowed to evaluate a player based on the postseason structure of the time (implicitly axiomatizing that the point of baseball is to win the World Series)? What I mean by this is the following. Compare 1993 to 1995. In 1993 there were 2 playoff spots for 26 teams. In 1995 there were 4 playoff spots for 28 teams. This could have a couple of implications: 1) a win might be worth less in 1995 than 1993 (because it is less likely to affect whether or not a team makes the playoffs -- I haven't evaluated this claim but it could be true), or 2) replacement level might be set differently for 1995 than 1993 (because worse teams are making the playoffs, thus providing 162 games of below-average performance might be more valuable in 1995, since it can more reasonably be part of a playoff team.) Which if either of these would be allowed?

Thanks for your patience and understanding, and sorry for the long post. Again, thought I'd throw these questions out there before going to the effort of doing thorough research. (Some/many of these things are things I do not actually expect to use, but would like to check out to see if they are worthwhile/significant/useful.)

Athletic Supporter

Wow AS a lot there. We'd love to have you join us. Would giving an extra week help you to get a ballot in for 2010? :-)

I don't have time right now to go point by point - but I don't think the things you propose are interesting.

In terms of what is allowed - I would say that Will Clark 2001 no. But Ryan Howard 2004 could be reasonable - think more like Charlie Keller 1938. General practice around here is that the player has to have had at least one season in the minors that was so good that you could make the case he should have been promoted after that, but wasn't due to blockage or organizational incompetence. Those types of things tend to be more in the past than recently.

I think it would be quite interesting - if one had the capability to do MLE's for everyone and count them for everyone throughout history. I don't think it's fair if you cherry pick one guy and said look at his great minor league season and then compare him to others without giving them the same benefit of the doubt. For the most part the Keller credit folks (of which I'm one) do that. Heck, read my ballot, Jack Quinn is very high, in part because of his PCL work mid-career.

The PPA approach is interesting - but I would think in terms of the fairness to all players, you'd need to do it in such a way that a player who has good/bad teammates isn't helped/hurt (in any combination) in the implementation. Someone like Ralph Kiner shouldn't be penalized (or helped) because his teams were awful - and vice-versa for Mickey Mantle.

I think the ideas you propose are very interesting, and I'd like to see the results.

I also think it would be great for the group as a whole if these things were discussed, so that we could a) add the features we like ourselves (see DanR and Chone's WAR systems, for example) b) give you constructive feedback in things you may have failed to consider.

What I mean is, it's better to join the discussion and help build knowledge, than to just show up with might appear to be a crazy ballot from out of LF with no peer review/chance to change the views of the peers. That's one of the reasons we ask new voters to post a preliminary ballot. It's not just to 'police' them. But it's so we can see their thought process, and have some give and take.

Does that make sense, seem reasonable? By all means, we'd love to have you join the group.

The general feeling is merit is tied to pennants and each season (back to the 1870s) has an equally valuable pennant. Players are generally evaluated in the regular season with the assumption that their contribution is to a "neutral" team and they aren't penalized for having bad teammates. Likewise any negative years are usually zeroed out because the thought is the team would replace bad players if the GM was doing his job. Credit for seasons not in MLB are generally given for the following reasons:

1) stuck in the minors / Negro Leagues due to racial discrimination or financial reasons but performing well at the highest available level to that player
2) not playing but in military service
3) sudden death (very few of us, controversial)

Injury is not usually accepted, early retirement because the player wanted more $$ isn't accepted.

Postseason contribution is haphazardly considered and using a pennants added approach would be welcomed (by me) assuming a poor postseason isn't treated worse than not getting there at all.

Japanese players aren't considered but MLB players have been credited with time spent in Japan as "time stuck in minors".

I almost forgot - a corollary to a "pennant" is Joe Jackson. I've discredited the entire year the 1918 White Sox threw the World Series for those players involved in the scandal and that is a fairly common practice.

346. Blackadder Posted: November 23, 2009 at 10:57 AM (#3394494)
It is worth noting that Sean's system applies a "DH bonus", of I believe around 5 runs per "season". I don't remember how he defines seasons, but that could account for around seven wins worth of discrepancy.

354. David Concepcion de la Desviacion Estandar (Dan R) Posted: November 23, 2009 at 12:38 PM (#3394603)
Blackadder--Oh really, I didn't know that. Well, I *very* strongly disagree with giving DH's a bonus for the supposed difficulty of having to hit without playing the field. That will make CHONE's system far, far more favorable to DH's than mine is.

I suppose that is a popular position, not to say consensus. Tom Tango has argued for the same, in some context where he also asked "Why doesn't Dan Rosenheck do that?" (I think that was in exchange with Blackadder or Bleed, whom I probably conflated until just now, as two pseudonyms that are too similar.)

#362
That is a lot of good questions from Jason that should be reproduced or linked to another thread. I'm not sure we have that thread, actually, for various methodological questions or ones more general than "war credit" and "major league equivalency" and "pennants added"(abandoned?). For most of the questions the answer to Jason must be Yes but it will be good if others take those questions as stimulants, performance-enhancing.

Wow AS a lot there. We'd love to have you join us. Would giving an extra week help you to get a ballot in for 2010? :-)

Thanks for your quick reply. Unfortunately, there's no way I could vote intelligently this year either way -- I'm going to be fairly swamped until the end of the month (hence procrastinating by typing a wall of text :)) and I don't see having enough time to compile resources/spreadsheet-it-up to be as informed as I would like. You guys are smart, I don't think adding a half-assed vote is desirable. :)

The postseason thing is interesting, something where a player is really adding value, or informing us more of his true ability (especially true if you buy "against better competition, true colors show" type arguments), but something where the opportunity is out of a player's control. I haven't really decided how I feel about it, but it seems as though a postseason inning should count at least as much as a regular season inning.

I think this is interesting from DL:

Postseason contribution is haphazardly considered and using a pennants added approach would be welcomed (by me) assuming a poor postseason isn't treated worse than not getting there at all.


Why do you think that a poor postseason shouldn't be treated worse than not getting there at all? If we accept that getting to the postseason is (largely) out of a player's control, it seems to me that if you give credits, you have to give demerits as well, to make the whole thing balance out -- otherwise you are (on average) giving a player credit for his team making the postseason in the first place. Just as as a thought experiment, suppose that player A had a .300 EqA in the regular season and then went 2-for-20 in the postseason, while player B had a .300 EqA in the regular season and his team missed the postseason. Would you not say that player B had a better season? His rate stats are better, and his counting stats are only worse because his team happened to play fewer games by virtue of not making the postseason.

I would anticipate, if I were allowed, at minimum throwing in postseason stats into a season-line, and counting it as a "172-game season" or whatever, adjusting the resulting stats to a "normal" 162-game season, and adjusting for higher quality of competition faced in the postseason. In some cases this would result in a player having a worse value for this season than he would if you just counted regular season stats.

DL, I hope you meant 1919 in post #365.

DH is a role, but it's not a position.

Almost more like "lead-off batter," or "clean-up" hitter, or "2nd batter," that sort of thing, I'd say.

Postseason contribution is haphazardly considered and using a pennants added approach would be welcomed

Pennants Added actually has nothing to do with whether a player was actually in a pennant race or not.

PS. I don't consider post-season. The whole point of giving MLE credit--e.g. for players who are off at war, e.g.--is to level the playing field. Why throw in one particular criterion that goes the other way?

Interesting. AS. I'd love to see a comprehensive post-season accounting for everyone, ever.

I've thought about going back and figuring Win Shares (I've built a spreadsheet that can calculate them) or asking DanR to figure his WAR for every post-season series.

Then, once you have the data, you can figure out how to apply it fairly. It's on the list, but this damned job keeps getting in the way.

Anyone want to give me a research grant? :-)

Seriously though, something like that would be immensely valuable for the evaluating the modern player.

Post season play isn't so important when you've got 1/10 of the league making it and playing just 7 games max, like in the 1960s.

But since 1995, 27% of the league makes the post-season, and they can play upwards of 12% of a season there - heck if the WS goes 7, Sabathia is throwing 18% of a season there in super high (seasonal) leverage.

I think looking at post-season play does make more sense now. As does finding a way to apply it so that you can compare a guy like Jeter to a guy like Ernie Banks; or a guy like Smoltz to a guy like Robin Roberts in a fair way.

I very much like your idea of prorating it into the seasonal lines of players who do make it, assuming you can fairly adjust for the tougher competition, and weather that isn't as conducive to hitting as well. The adjustments for pitchers and hitters won't be the same, etc.

BTW, posts 362-372 are an excellent example of why we have perpetual eligibility. You never know what new items will come to light.

Would you not say that player B had a better season?

Not at all. Neither player contributed anything more to his team's pennant chances than his regular seasons stats would indicate. A small sample size flop in the postseason is meaningless. Performing well against top competition in the postseason is worth a (small) bonus. I zero out crappy/injured seasons for players the same way.

Postseason stats can't be treated like regular season stats because replacement level becomes meaningless in the postseason.

I think the main reason why nobody uses postseason stats is there is no good comprehensive database for them that allows contribution to a pennant race to be measured systematically.

Yeah, I meant 1919.

Then, once you have the data, you can figure out how to apply it fairly. It's on the list, but this damned job keeps getting in the way.

Anyone want to give me a research grant? :-)

Yeah, a lot of the things I've thought about doing (on that list) would probably represent a massive time commitment. I'd imagine at first I would just evaluate any non-HoMer that has received a vote in the last HoM election (plus promising newbies), then broaden to eligible HoMers. Things like adjusting game-by-game (or, probably more realistically/fairly, month-by-month) for the relative importance of that game would definitely be a huge pain.

I'm leaning towards using a rate stat based on that, a sort of WPA/LI on a season-scale rather than a game-scale (this will be fair to players on all qualities of teams.) Of course, calculating pennant probability based on the standings is not a simple thing to do well (probably will end up being just Monte Carlo with 50/50 for each remaining game.) I'll start investigating at some point. :)

Jason Kendall's 6 millionth and then some fan (aka AS, aka Athletic Supporter)

Those are some great questions. This is what I think, bearing in mind that this is only one opinion.

Succinctly: is the point of this to find the "best" players, or the most "valuable" players?

Succinctly: one of the points of the Hall of Merit is to eliminate any differentiation in the idea between "the best" and "the most valuable." The best are the most valuable, and vice versa. A player isn't more valuable because he had better teammates who helped him get into the playoffs. Just my opinion, but the "valuable" in MVP and the "fame" in Hall of Fame are too of the most abused words in the English language.

For example, it seems likely that Barry Bonds was one of the best players on the planet, certainly above replacement level, in 2008. The same is true, possibly for very different reasons, for, say, Will Clark, 2001, or Joe Jackson, 1921. These players obviously accumulated no "value," but is it allowed to give them credit for being among the "best" players on these years (leaving aside the question of the extent to which that was actually true)? Obviously something like war credit and Negro League credit is very commonplace here, but I'm not sure if those are viewed as special cases, or if they are just an implication of the desirability to find the "best" players.

Some voters do award blacklist credit for players whose careers were interrupted or ended prematurely do to such issues. This is a controversial and not universally accepted premise, but it is allowable within the constitution. For example, some voters credited Charley Jones for his 1881 and '82 seasons which he lost due to a financial dispute with one team which led to his being blacklisted from the major leagues. That credit played a role in Jones' eventual induction. However, Jones was elected with votes on more than 20 ballots with many of those voters awarding no blacklist credit.

If you were considering blacklist credit, I would suggest that such credit could help a case, but not make it. The three players you mentioned need no such credit. Curt Flood is unlikely to be a legitimate candidate even with such credit. At least one current voter does give such consideration to Eddie Cicotte, though it is again a controversial decision.

A similar question is true for minor league credit, especially contemporary minor league credit. It seems likely that Ryan Howard in 2004 would have been above replacement level in the majors, regardless of whether or not it was reasonable to expect him to have been in the majors going into 2004, not to mention Edgar (who I noticed did get MiL credit from a few people.) What about Sadaharu Oh? Is he getting no votes because it's not allowed, or because he is not viewed as as-good-a-baseball-player as the other people on people's ballots?

Joe Dimino and DL from MN already answered this, but I'll add my two cents. Minor league credit was a more important issue in the early years of the project when it wasn't always as clear which leagues were "the majors" and when major league franchises didn't have farm systems. For example, Lefty Grove pitched for the Baltimore Orioles of the American Association for 4 and 3/4 seasons before the Orioles allowed the major league Philadelphia Athletics to purchase his contract. Grove's subsequent major league career was impressive in itself, but his full career shows even more value when you consider his 111-39 record with a 2.90 era in 1243 innings in the minors. Most voters who award minor league credit would give Grove consideration for 3 years of minor league service, in essence writing off one season at a high level in order to get noticed.

Minor league credit for early players is one of the way we have of being fair to all eras and widely (if not universally) accepted. Minor league credit for more recent players (post WWII for white players, post integration for African Americans and dark-skinned Hispanics) is a little more tenuous. Some voters do award minor league credit on the same principle- one big year to get noticed, subsequent years credited. A few voters give Edgar Martinez such credit and when Ryan Howard becomes eligible in 20 years or so, a few voters may give him credit as well. However, I would say the same thing that I did regarding blacklist credit. Though minor credit isn't nearly as controversial, such credit should help a case not make it. Edgar Martinez is getting plenty of votes from members who do not give minor league credit to players after 1960.

I'll keep going but that's enough for one post.

I think the main reason why nobody uses postseason stats is there is no good comprehensive database for them that allows contribution to a pennant race to be measured systematically.

That doesn't mean it's a bad idea to count them. It means someone needs to create the database so we can.

Not at all. Neither player contributed anything more to his team's pennant chances than his regular seasons stats would indicate. A small sample size flop in the postseason is meaningless.

I disagree, I think it's at a minimum as meaningful as a bad week during the regular season, and more realistically probably 2-3x as important. It's just a question of how to fairly account for it.

Neither player contributed anything more to his team's pennant chances than his regular seasons stats would indicate. A small sample size flop in the postseason is meaningless.

Is it more meaningless than a small sample size flop from May 20-30, which is counted negatively towards a player's value? I'm not (yet) saying that 2-for-20 in the postseason is necessarily more important, or indicative of a player's value or ability, than 2-for-20 in May. But it seems weird to me that it would be _less_ important. As Joe said, this question is much more important in the modern age, where there are players (who, not coincidentally, tend to be HoM-relevant) who spend a substantial fraction of their career playing in the postseason (Mariano Rivera has pitched the equivalent of almost two full seasons in postseason innings.)

I can definitely see the argument that since not all players have the opportunity for these "extra stats," they should be thrown out; just not my personal preference.

And yes, I would say that going 2-for-20 in the postseason is a substantial negative contribution to a team's pennant chances (where "pennant" here means "World Series win").

At least one current voter does give such consideration to Eddie Cicotte, though it is again a controversial decision.

I can't (don't?) read every ballot word for word, but I missed this one. Wow. I'm not sure this one isn't over the line . . . this wasn't a condition of the game, say what you want about the Black Sox, and I'm generally quite sympathetic to them. But giving them credit for the time they missed? That seems awfully tough to justify to me.

I'm leaning towards using a rate stat based on that, a sort of WPA/LI on a season-scale rather than a game-scale (this will be fair to players on all qualities of teams.)

I see this as being very cool if you could pull it off. Like not counting anything a player does after his team clinches HFA or is eliminated from the playoffs (except against playoff contenders maybe?), and prorating the 'meaningful' part of their season to cover the entire season.

I mean if Jeff Bagwell breaks his wrist the day after the Astros are eliminated from the playoffs, does it really matter? Weren't those first 148 games his 'full season' anyway? The rest are exhibitions.

But again the key is covering everybody and not applying something like this selectively. IMO, something like this is probably the next 'quantum leap' in terms of historical player evaluation.

But I'm not sold on doing it in a WPA/LI manner - as game 163 Tigers/Twins is absolutely not any more important than Tigers/Twins June 12. Without what happened earlier, what happened later doesn't get to happen. Games in May when you are 6 games up are every bit as important as games in September when you are 1 game back, if that's where you eventually end up.

A position player whose entire season consisted of 2-20 in May would get zero credit in my system. I treat the postseason the same way, except I treat each postseason series as a discrete event (2-8 in the first series is worth something pushing to the next level, 0-12 in the next series is worth zero).

two from the Ballot thread

73. James Newburg Posted: November 22, 2009 at 03:01 AM (#3393796)
Might be interested in casting a ballot. Does anyone know if Chone Smith's WAR numbers are adjusted for season length?

From Brent, Bleed, and others I infer not. See #301-312 above, especially #303-305.


123. Esteban Rivera Posted: November 23, 2009 at 12:29 PM (#3394594)
2010 Ballot:
...
4. Bill Monroe - Seems to be one of the best second basemen of his time. Has any new info on him (or other NeL candidates) been uncovered?

For a start see the recent contributions to Hilton Smith and Dick Redding, by Brent who seems to be working on wikipedia players pages as an ongoing project. Gary A (agate type) is one of his sources, which does go beyond data previously published in print. If none of it is strictly new here(?), there is nevertheless something new when Gary A comments here.

Anyone who has been away since the last tri-weekly cycle, about two years ago, should also visit at least our threads for Bill Monroe, Charlie Grant, Monte Irvin.

-- Is it allowed to give players in pennant races extra credit? In the extreme, would it be allowed to use a version of PPA (Pennant Probability Added)? What about a rate stat for this? (This would have the practical effect of weighting players' early-season stats more if they played for bad teams or dominant ones, and less if they played for a team in a pennant race at the end of the season.)

Generally, no. PPA would be allowed but as sunnyday notes, it is given to all players and not just those in an actual pennant race.

-- Is it allowed to consider manager usage or context-dependence of stats? For instance, consider the Koufax/Drysdale Dodgers. There were times when both were rested but only one could start. Perhaps over the course of their years, either would have accumulated a few more starts if on another team. Similarly, a low IP total could be adjusted for for a SP on a team with an exceptionally good bullpen. Another example of this is a player who batted low in the order due to teammates (e.g. Davey Concepcion); is it allowed to adjust such a player's counting stats upwards? What about downgrading walks for #8 hitters in non-DH leagues? (Not that many #8 hitters are serious HoM candidates, but you get the idea.)

Yes, it is. We consider not only the stats themselves but also the context in which a player was able to compile those stats. For example, it's fairly common to consider a pitcher's defensive context when evaluating his numbers. Some voters dock Bucky Walters for pitching in front of a good defense which may have had the affect of making him appear to be a better pitcher than he really was. That can be overdone, but it is part of the process.

Managerial decisions and usage patterns are part of the picture as well. I remember the discussions about Whitey Ford and Billy Pierce, two '50s pitchers who were subsequently elected to the Hall of Merit. Some voters were concerned that Whitey Ford's numbers were artificially inflated because he didn't have to face the best-hitting team in baseball- his New York Yankee teammates. However, others noted that Ford's manager adjusted his start dates so that he faced a higher proportion of the other best teams than other pitchers of that era. The same was true for Pierce. He faced a disproportionately high number of top teams (including the Yankees). That helped him get elected as it told us that he was potentially even better than his numbers would indicate. Going forward, I could see that being part of the picture for a future candidate like Roy Halladay, who has a high number of games against the dominant teams of his era, the Yankees and Red Sox.

Usage is even more important in terms of evaluating relief pitchers. How many times did they face the best hitters in the opposing line-up, and not just the three guys who happened to be up to bat in the ninth?

Usage can also come into consideration for hitters. The name that comes to mind is Norm Cash. One of the disagreements over Norm Cash is whether his numbers should be taken at face value or discounted because he often sat down against left-handed pitchers, allowing him to maintain higher rate stats than he would have otherwise.

However, I'm not sure of the examples you use. Concepcion may have batted lower in the order for the Reds than he would have for another team, but his opportunities weren't necessarily reduced by it. He would have received more batting opportunities (plate appearances and at-bats) than other players in his spot in the order because his superior teammates would turn the order over more frequently. The lower-spot/more opportunities probably evens out. It may be worthwhile to look at, but I'd be skeptical of any large adjustments. Also, since most voters use uber-stat systems like Win Shares and WARP rather than traditional counting stats like home runs and RBI, it's mostly irrelevant.

[ok, "Invalid Edit Comment" means that another author has intervened(?) ... yes, that seems to be ... next fortnight we will again have a few hours to work on a few inches contribution, but not this week]

Anyone who has been away since the last tri-weekly cycle, about two years ago, should also visit at least our threads for Bill Monroe, Charlie Grant, Monte Irvin.
Index, HOM discussion of Negro Leagues candidates

He would have received more batting opportunities (plate appearances and at-bats) than other players in his spot in the order because his superior teammates would turn the order over more frequently. The lower-spot/more opportunities probably evens out.

You could easily be right about this, it's an excellent point. However, I'm not sure that the uberstat systems correct for this. Suppose we have two copies of the same player in the lineup of the same team, one batting first and one batting eighth. Over the course of the season the player batting first will get maybe 15-20% more of everything than the player batting eighth (let's ignore RBI and RS for this, and just consider a Linear Weights type system, which I believe is the offense input behind WS.) He will grade out as having 20% more VORP because he has the same rate stats but 20% more appearances. I believe he will also grade out as having 20% more offensive WS.

Maybe the Concepcion example is misleading because in that case, maybe it cancels out for Concepcion, but we should actually downgrade the WARP for, say, Joe Morgan, who was getting the benefit of a high batting slot _and_ the benefit of extra PA's due to teammates. Now, those 10% more PA's (say) due to his teammates _do_ mean that Morgan produced more value, but the fact that Actual-Morgan produces 10% more value than Alternate-Universe-Morgan (with worse teammates) doesn't mean that Actual-Morgan is better and, debatably, doesn't mean that Actual-Morgan should be ahead of Alternate-Universe-Morgan in the inner circle of the HoM.

This argument is somewhat torpedoed by the fact that Morgan's PA/G with the Astros were really no different than his PA/G with the Reds. I wouldn't expect it to necessarily be a huge effect, but given that the players at the top of the backlog are all very similar in value, even a 2- or 3-win difference over the course of a career could make some difference. I don't know; I wouldn't be surprised if the lineup-spot and more-PA's effects were entirely insignificant, much smaller than even that.

I will be a bit annoyed if we postpone the ballot, because I both bust a gut in stayng up late and behaved irresponsibly towards my work in order to get my Hall of Merit ballot finished.

The annoyance isn't so much in the act of postponing the ballot, but because someone earlier asserted that there should be no extension under any circumstances. It would have been better if the extension option had not been dismissed so sternly in the first place. If we're going to be relaxed about deadlines each year, let's not pretend otherwise. It just causes headaches for some people.

But it won't kill me.

-- The HoM timelining has a recommendation built in based on the number of elected players each year. Is it allowed to believe different things about desirable relative representation of era? Obviously this is a controversial topic; proportional representation could be done in any number of different ways (e.g. constant # per year, proportional to number of players in MLB, proportional to size of "talent pool," proportional to interest in baseball (e.g. attendance numbers), etc..) I'm not talking about "are contemporary players just better than old players as per WARP3", but rather "what should the representation of the HoM be" -- is that something that an individual can vote based on their personal opinions of?

Yes. Being "fair to all eras" does not mandate identical representation. There is a general consensus that more recent decades should have higher representation due to a higher talent level coming from a greater geographical distribution and a greater number of teams.

This might be a good time to post the shorthand list of players by decade:

1860s - 1 (Pearce) (SS)

1870s – 9 (Anson, Barnes, McVey, Pike, Spalding, Start, Sutton, White, Wright) (P, C, 1B-2, 2B, 3B-2, SS, CF)

1880s – 19 (Bennett, Brouthers, Browning, Caruthers, Clarkson, Connor, Ewing, Galvin, Glasscock, Gore, Hines, C Jones, Keefe, Kelly, O'Rourke, Radbourn, Richardson, Stovey, Ward) (P-5, C-2, 1B-2, 2B, SS-2, LF-3, CF-3, RF)
{Candidates –Welch, Williamson, Dunlap}

1890s - 18 (Beckley, Burkett, Childs, Dahlen, Davis, Delahanty, Grant, Griffith, Hamilton, Jennings, Keeler, Kelley, McGraw, McPhee, Nichols, Rusie, Thompson, Young) (P-4, 1B, 2B-3, 3B, SS-3, LF-3, CF, RF-2)
{Candidates –Duffy, Van Haltren}

1900s - 18 (Bresnahan, M Brown, Clarke, J Collins, Crawford, Flick, R Foster, Hill, G Johnson, Lajoie, Mathewson, McGinnity, Plank, Sheckard, Waddell, Wagner, Wallace, Walsh) (P-7, C, 2B, 3B, SS-3, LF-2, CF, RF-2)
{Candidates – Leach, Willis, Monroe, Joss}

1910s - 17 (Alexander, Baker, Carey, Cobb, E Collins, Groh, J Jackson, W Johnson, Lloyd, Magee, Mendez, Roush, Santop, Speaker, Torriente, Wheat, J Williams) (P-4, C, 2B, 3B-2, SS, LF-2, CF-5, RF)
{Candidates –Cravath, Redding, Taylor, Doyle}

1920s - 20 (Beckwith, Charleston, Coveleski, Faber, W Foster, Frisch, Goslin, Heilmann, Hornsby, Lundy, Mackey, Moore, Oms, Rixey, Rogan, Ruth, Sewell, Sisler, Vance, Wilson) (P-6, C, 1B, 2B-2, 3B-2, SS-3, LF, CF-2, RF-2)
{Candidates – Grimes, Traynor, Shocker, Mays, Schang}

1930s - 29 (Averill, Bell, R Brown, Cochrane, Cronin, Dickey, Dihigo, Ferrell, Foxx, Gehrig, Gehringer, J Gibson, Greenberg, Grove, Hartnett, Herman, Hubbell, Lyons, Medwick, Ott, Paige, Ruffing, Simmons, Stearnes, Suttles, Terry, Vaughan, Waner, Wells) (P-8, C-4, 1B-5, 2B-2, SS-3, LF-2, CF-3, RF-2)
{Candidates – B Johnson, Dean, Bridges}

1940s – 18 (Appling, Boudreau, W Brown, DiMaggio, Doerr, Feller, Gordon, Hack, Irvin, Keller, Leonard, Mize, Musial, Newhouser, Reese, Slaughter, Trouppe, TWilliams) (P-2, C, 1B-2, 2B-2, 3B, SS-3, LF-3, CF-3, RF)
{Candidates – Rizzuto, Walters, Pesky, Clarkson, Elliott, Stephens}.

1950s – 18 (Ashburn, Banks, Berra, Campanella, Doby, Ford, Fox, Kiner, Lemon, Mantle, Mathews, Minoso, Pierce, Roberts, J Robinson, Snider, Spahn, E Wynn) (P-6, C-2, 2B-2, 3B, SS, LF-2, CF-4)
{Candidates- Newcombe}

1960s – 21 (Aaron, Allen, Boyer, Bunning, Clemente, Drysdale, Freehan, B Gibson, Kaline, Killebrew, Koufax, Marichal, Mays, McCovey, B Robinson, F Robinson, Santo, Torre, Wilhelm, B Williams, Yastrzemski) (SP-5, RP, C-2, 1B-3, 3B-3, LF-2, CF, RF-4)
{Candidates – E Howard, Brock, Cash}

1970s – 24 (Bench, Blyleven, Carew, Carlton, Da Evans, Fingers, Fisk, Grich, RJackson, Jenkins, Morgan, Nettles, Niekro, Palmer, Perry, Rose, Ryan, Seaver, Schmidt, Simmons, RSmith, Stargell, Sutton, J Wynn) (SP-9, RP, C-3, 1B-2, 2B-3, 3B-3, CF, RF-3)
{Candidates- Tiant, Reuschel, Concepcion, Perez, Bonds, Singleton, Campaneris, John, Staub, Tanana, Bando}

1980s- 22 (Boggs, Brett, Carter, Dawson, Eckersley, Dw Evans, Gossage, Henderson, Hernandez, Molitor, Murray, Raines, Randolph, Ripken, Saberhagen, Sandberg, O Smith, Stieb, Trammell, Whitaker, Winfield, Yount) (SP-2, RP-2, C, 1B-2, 2B-3, 3B-3, SS-4, LF- 2, CF, RF-2)
{Candidates- Puckett, Murphy, L. Smith)

1990s- 3 (W Clark, Gwynn, McGwire) (1B-2, RF)
{Candidates- Cone, Belle}

As you can see, the number of electees per decade remained fairly stable from the 1880s to 1950s, with an average of 18 players per decade (ignoring the outlier of the 1930s). However, the number of players elected per decade has been steadily increasing since then. The '90s look to produce at least 23 players and could run as high as 27 or 28. This reflects the expansion of the number of teams and number of players, and possibly the increase in talent level as well. There are disagreements as to whether or not we are accelerating at the correct rate. Some might suggest we're doing so too quickly, others too slowly. But "what should the representation of the HoM be?" is a legitimate question to ask and part of the regular debate.

I wouldn't expect it to necessarily be a huge effect, but given that the players at the top of the backlog are all very similar in value, even a 2- or 3-win difference over the course of a career could make some difference. I don't know; I wouldn't be surprised if the lineup-spot and more-PA's effects were entirely insignificant, much smaller than even that.

It's certainly something worth looking into.

Looking at Chris Fluit's list makes me wonder if the 1950s are underrepresented. On the one hand, segregation is gone (but not forgotten) so the talent level is beginning to rise, while the number of teams stays the same. On the other, integration is proceeding slowly, so the talent level isn't rising all that fast. Also, some '1950s players' may actually be '1960s players'. I need to revisit that Mark Armour article on the progress of integration that was based on the SABR presentation he gave in St Louis 06. That'll have to wait until I get home again in December.

-- Suppose that a player is believed to have an actual effect on his teammates' performance. Is one allowed to consider this? Does it matter if it is a subjective opinion or a quantitative one (e.g. a study of performance before-or-after)? Does it matter if there is a known mechanism (e.g. sign-stealing) involved? What about a negative effect, e.g. introduction of cocaine to the clubhouse, or recruiting them to a game-fixing organization? I'm not talking about character issues (e.g. steroids) -- I'm talking about situations where a player affected by his actions the performance of his teammates, translating into an effect on his team's record. (Indeed, if a player were responsible for steroid-enhanced performance of his teammates, this would be a plus.)

Generally, no. I remember this being a vigorous discussion during Dick Allen's candidacy. There were some who maintained that Allen's attitude negatively affected his teammates' performance and that his numbers should therefore be discounted as an offset. It was generally regarded as unquantifiable and the general consensus was to leave it out of any consideration. Voters are allowed to exercise the one-year boycott in such situations (nobody did) but not to impose an otherwise arbitrary discount.

This would be true for both negative influences (Allen, Jeff Kent, Milton Bradley) as well as positive influences. Pee Wee Reese got no extra consideration for being a good teammate to Jackie Robinson and helping the clubhouse rally around him. He gets into the Hall of Merit on his performance alone (including relevant war credit).

Pulling this over from the ballot thread:

What makes (Edgar) significantly better than Bob Johnson?

Quoting Edgar's career OPS+ figure (147) doesn't really give a sense of Edgar as a player to the same extent that Indian Bob's 138 does. In Edgar's first 7 years as a DH, he posted an aggregate OPS+ of 163, better than every single season of Johnson's career except for his war-aided 1944. If you ignore his three part-time seasons breaking into the lineup and his 2004 swan song, Edgar's career breaks down into three phases:

1990-92: Regular third baseman. 145 OPS+ over 1804 PA.
1993-94: Struggling third basema. 115 OPS+ over 552 PA. Started transition to DH during '94 season.
1994-03: Regular DH. 159 OPS+ over 5487 PA.

On the other hand, Johnson basically remained the same player throughout his career - a guy who would post an OPS+ somewhere between 130 and 150 while playing a corner outfield spot. Conveniently, Edgar's two really good stretches cover the same number of years as Johnson's entire career. I would say that even though Edgar had a little more trouble staying on the field, his typical value during either of his primes was better than Johnson's.

Another relatively minor quibble re Brock's ballot, since he's already corrected the major ones. :)

Re David Cone: That's not a good sign, since pitchers of his quality are hard to find. If one keeps moving on, he must be some sort of clubhouse problem.

You posted the same thing last year, Brock, and there were a couple of responses (OK, one of them from me, I admit) taking issue with this statement. I refer you to last year's ballot thread (http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/2009_ballot), posts 84 to 86, at the end of which you said:

As for right now, I'm going to take your analysis and Howie's personal knowledge to heart and revisit Cone when I'm working out the pitcher ranks without making any deductions, to see what happens.

Believe me, I understand re not having time for these things, and it's likely that even removing what I feel is a misunderstanding of Cone's teams' attitude toward him (as I posted there, I think only the Mets would have had any reason to think of him as a clubhouse problem--and even that's not at all certain) would not put him on your ballot. But I wanted to just bring it up again, since I really do feel that it's unfair to be giving him these "intangible" deductions--he doesn't merit them. In my opinion, obviously. :)

-- Is it allowed to evaluate a player based on the postseason structure of the time (implicitly axiomatizing that the point of baseball is to win the World Series)? What I mean by this is the following. Compare 1993 to 1995. In 1993 there were 2 playoff spots for 26 teams. In 1995 there were 4 playoff spots for 28 teams. This could have a couple of implications: 1) a win might be worth less in 1995 than 1993 (because it is less likely to affect whether or not a team makes the playoffs -- I haven't evaluated this claim but it could be true), or 2) replacement level might be set differently for 1995 than 1993 (because worse teams are making the playoffs, thus providing 162 games of below-average performance might be more valuable in 1995, since it can more reasonably be part of a playoff team.) Which if either of these would be allowed?

I'm going to preface my response by once again asserting that this is only one opinion, but I think this is something we will have to seriously consider going forward.

In the past, it was severely discouraged if not outright prohibited. Postseason credit was considered unfair and unreliable. It's unfair in that postseason participation is unevenly distributed. Some players never got the chance to play in the postseason despite superior careers (such as Ernie Banks) while others had a disproportionately high number of playoff appearances due to the franchises for which they played (such as Catfish Hunter). It was considered unreliable in that an undue amount of influence was placed on a small proportion of games. Even if a postseason game is worth 2-3x a regular season game (as someone suggested in this thread), it was still 7 games worth 21 as opposed to a full season of 154.

However, I think that's changed in the Wild Card era. More teams make the playoffs and so more players have the opportunity. Although there are cases of players not making the playoffs in their career at all or until very late (Carlos Delgado, Roy Halladay, Brian Roberts- basically your AL East teams), those cases are increasingly rare. Plus, the added rounds of the playoffs mean that the sample sizes are significantly larger and therefore also more statistically significant. For example, Bernie Williams played 121 postseason games. That's almost a full season- certainly enough to have qualified him for a batting title in a given year. Andy Pettitte has 249 innings in the postseason, more than in any one year in his regular season career. The longer postseason is also affecting usage. Joe Girardi reduced his primary pitchers' workload down the stretch, taking them out of games early and giving them extra days off so that they'd be fresh for the postseason. That means that Sabathia, Burnett and Pettitte's games started, innings pitched and wins totals were reduced. Only marginally, to be sure. But when we're splitting hairs about borderline inductees, those incremental differences can add up.

I'm not sold on the idea of postseason credit yet. It still unevenly rewards players on good teams like the Yankees, Red Sox and Braves.

Furthermore, the larger sample size does mean that the players with the best postseason numbers are now more likely to be the same as the players with the best regular season numbers. There's a reason why potential (and surefire) HoMers like Bernie Williams, Manny Ramirez, Mariano Rivera, John Smoltz and Andy Pettitte are the ones re-writing the postseason record books (in terms of career numbers, at least). So it's possible that postseason credit is actually less needed as the ones who would most deserve it are the ones who would probably be elected anyway. Either way, we'll have a better idea in a decade or so when we can see what's happened to the candidacies of players from the '90s and '00s.

one item from Jason #362, the excellent list of big items that has taken over the board
-- The HoM timelining has a recommendation built in based on the number of elected players each year. Is it allowed to believe different things about desirable relative representation of era?

It's not a recommendation but a floor, a minimum representation of every initial period defined essentially by major league "finale" dates. Thus,
- at least four players with finales before 1893
- at least six with finales before 1894
...
- at least 234 with finales before 2003
...


That was a conservative policy in two senses,
- anticipating that representation of almost every initial period would naturally be above this floor. (Hence the elections collectively rather than the stipulation would determine the time distribution of HOM players. Although the constraint would be binding every year, it would never be binding with the hindsight of several years.)
- stipulating this much early representation "just in case" the first sense proved wrong.

Of course the stipulation may have ensured fulfillment of the anticipation partly by repelling, throughout the 110 triweekly cycles, those potential newcomers who couldn't tolerate the number of older players already recognized.

In the event Charley Jones, Browning, Beckley, Childs, Griffith, Jennings, McGraw, Waddell, and Bresnahan were elected after the passing of the wave from the 1920s and 1930s, the formal Negro Leagues and one so-called Golden Age of the majors. --elected in the last fifty years

Looking at Chris Fluit's list makes me wonder if the 1950s are underrepresented. On the one hand, segregation is gone (but not forgotten) so the talent level is beginning to rise, while the number of teams stays the same. On the other, integration is proceeding slowly, so the talent level isn't rising all that fast. Also, some '1950s players' may actually be '1960s players'. I need to revisit that Mark Armour article on the progress of integration that was based on the SABR presentation he gave in St Louis 06. That'll have to wait until I get home again in December.

I wonder that, too.

And yes, one likely "1950s player" is listed as a "1960s player." With smoother integration, Elston Howard might have been a '50s player as he turned 31 in 1960. However, as his career actually stands, he accrued more value in the '60s (plus, there are some who argue that he was a late bloomer anyway).

Also, another theory as to the low number of '50s candidates (we have an even number of inductees) is WWII. Potential superstars may have been derailed by military service, injury and death before we even knew who they were. The better way to assess that would be to look at HoM inductee birth years as opposed to decades of accrued value (which is what this chart attempts to measure). A significant dip in players born in 1924-27 (those who would have turned 18 during the war years of 1942-45) could be instructive.

Bob Johnson was a very good fielding outfielder, not all his value came from his bat.

Crikey, another ballot using WARP3, WAR and Win Shares that only has players from the 162 game schedule has appeared.

Joe,

You keep repeating not to take comments personally, how else am I suppose to interpret the comments by DL-MN? He basically said my choices sucked and that I would ruin the integrity of the HOM. In the real world you would have deal with those comments.

You made this comment: “And people aren't just going to take your word for it, they are going to challenge you, and expect a reasoned response.” Why do I have to justify my response?? I made my choices I gave my reasons that’s it. Why do I have to justify by response to DL-MN or anyone??

You seem to want to open this process to new people as long as they vote the way you want them to vote. What the hell is the point of that?? Just for people to rubber stamp your decisions?? What’s democratic about that process?? What you do is turn people off to this whole process.

If that’s the case just keep the HOM closed to 20-30 voters. I have no problem with that.

Sorry, I though Jack Clark, Robin Ventura, Ron Cey etc, were better than any of those pre-1920 players, that’s my opinion sorry if you don’t agree.

My point about John McGraw is that this is the 2010 ballot and you’re still dealing with turn of the century players. Maybe you should have increased the number of members elected each year so you wouldn’t still be dealing with 20-30 members from the dead-ball era.

You keep bring up “Being Challenged”. Why? Don’t have an open invitation to vote. If I vote in a Presidential election I don’t have to justify my vote or accept being challenged by people. I made my choices I gave my reasons, you’ve already elected 237 odd players, I have no problem saying that any of those 15 players I chose

Of course. Dan isn't really talking about Martinez here; he's talking about the replacement level for DH, and how it affects Martinez.

Yes, and I'm talking about Replacement Level Theory. I think it's irrelevant that Martinez can be more easily replaced by someone tolerable, because Martinez' absolute contribution is high enough that it warrants our attention. By contrast, the full value of Bert Campaneris' contribution is obscure, because it's not apparent from his batting stats that Campaneris was harder to replace, unless you know the context.

I like DanR's work an awful lot, but I worry there's too much weight put on his measure of Replacement Level in the case of a player whose career OPS+ would rank among the top 10 in a typical season.

I can't agree with fra paolo and DL from MN that every ballot should reflect positional balance, certainly not as a matter of policy.

The problem I'm trying to solve is how to get an accurate value for the 'glove positions'. The most important player on the field is the pitcher, who is involved in every pitch. The second-most important is the catcher, who is involved in most of the pitches, and often has a short career as a weaker hitter because of the wear-and-tear effect of simply being a catcher. This was the point I raised in the Munson thread, which is that games played as catcher should count as a meritworthy statistic. I don't think strict positional balance is necessarily the best way to solve the problem, but it's one of the simplest. If I stumble upon a better solution, I'll start arguing for that.

I would like to learn more about fra paolo's pitcher system, including how one does Fielding Independent Pitcher Earned Runs (that's the foundation, I infer).

Well, it's derived from Clay Davenport's article on his fielding translations in the 2002 Baseball Prospectus annual. Basically, non-HR hits and non-K outs are divided between pitchers and fielders on a 30/70 basis. Doing this at the league level allows one to work out an 'average team's' number of ERs.

Today, however, I concluded that my era adjustment is inadequate, so it still needs some work. (DanR has mentioned the same problem working with DERA.) When I think that's fixed, maybe that's the time to post some more detailed explanation somewhere.

A significant dip in players born in 1924-27 (those who would have turned 18 during the war years of 1942-45) could be instructive.

Didn't Anthony Giacalone's presentation at SABR this year potentially provide some information that could help with this? Or did he only start after 1929?