Required disclosures (top returnees): Curt Schilling, Sammy Sosa, Vic Willis, Phil Rizzuto, Ben Taylor, Dick Redding

2014 Prelim

1) Greg Maddux - top 25 player all time, 8th on my list of pitchers between Seaver and Spahn
2) Tom Glavine - top 60 player all time, ranked between Blyleven and Robin Roberts around 16th among pitchers
3) Curt Schilling - drops to 20th among pitchers. Ignoring early pitchers he's behind Blyleven and Roberts but ahead of Carl Hubbell, Gaylord Perry and Dazzy Vance. Top 70 player in baseball history.
4) Mike Mussina - top 100 player - around even with Gaylord Perry for 26th among pitchers
5) Frank Thomas - About a 50th percentile Hall of Merit inductee. 10th among 1B behind Buck Leonard and ahead of Mark McGwire
6) Tommy Bridges - Have been supporting Bridges since the 1970 ballot. Still think he's great.
7) Urban Shocker - gets WWI credit
8) Jeff Kent - 15th among 2B just ahead of Bobby Doerr, Joe Gordon and Roberto Alomar
9) Bus Clarkson - NGL and Mexican league credit
10) Phil Rizzuto - WWII credit
11) Gavy Cravath - minor league credit
12) Luis Tiant
13) Bob Johnson - on every ballot since I started voting in 1968
14) Ben Taylor - how do we induct Palmeiro and Beckley but not Ben Taylor? Taylor has the advantage of being the best 1B in the league and they don't. Great fielder during an era where it mattered quite a bit.
15) Bert Campaneris

16-20) Tony Mullane, Bucky Walters, Dave Bancroft, Hilton Smith, Norm Cash
21-25) Johnny Pesky, Dick Redding, Wally Schang, Sammy Sosa, Don Newcombe
26-30) Dave Concepcion, Babe Adams, Tommy Leach, Dizzy Dean, Jack Fournier

70) Vic Willis - too many contemporaries I like better - Babe Adams, Burleigh Grimes, Dick Redding, Wilbur Cooper

70) Vic Willis - too many contemporaries I like better - Babe Adams, Burleigh Grimes, Dick Redding, Wilbur Cooper

With the exception (barely) of Adams, none of these guys were contemporaries. Willis was done by 1910.

Not arguing with the placement, by the way. Willis got a good boost from Pittsburgh's defense, which from what I can tell wasn't far behind that of the Cubs. The Pirates of the aughts had a lot of one-year and two-year wonders who never did much anywhere else; their pitching staff was mostly Willis and a bunch of guys who would have been lucky to get 10 starts for the Giants or Cubs, they were usually near the bottom of the league in pitching strikeouts - and they were still usually about even with the Giants (who were usually the strikeout leaders) behind the Cubs.

-- MWE

Thanks for the comment, I was going mostly from memory.

While certainly Willis did benefit from the Pirates defense, he held his own pretty well too. For example, his FIP-based WAR from Fangraphs is between 15 and 20 wins higher than his teammates Leever and Phillippe. Of course, Willis' BBref and BG WAR is also way higher than his teammates, indicating that no matter how you look at it Willis was better than Leever and Phillippe. Willis' WAR numbers also rate favorably among the pitchers DL likes better as well.

Some RA+ Pythpat career records, with a "big years score" in brackets. Organized in order of Fibonacci win points on the equivalent record.

Clemens 359-187 [124]
Maddux 343-213 [106]
R. Johnson 275-162 [98]
P. Martinez 218-96 [91]
Glavine 284-206 [30]
Schilling 227-135 [50]
Mussina 236-147 [34]
Smoltz 234-152 [16]
Smoltz (as a starter only) 211-143 [16]
K. Brown 216-146 [46]
Tiant 224-164 [35]
Cone 190-132 [19]
Appier 154-119 [27]
Morris 226-199 ]9]
Rogers 194-173 [5]
Radke 154-119 [4]

I would describe Kenny Rogers as a poor man's Jack Morris, or as Morris without as much inning-eating durability. Rogers shares with Morris the fact that his actual W-L record is much better than his RA+ equivalent record, and by about the same amount.

Mussina I see as closely comparable to Schilling. Between the two of them, I'll put Schilling ahead for a slightly higher peak and slightly more post-season value. But it's a close call.

Johnson, Martinez, and Smoltz aren't on the 2014 ballot - they all become eligible in 2015.

Now, here is another table: by RA+ equivalent record, the best consecutive 7 year stretch of a pitcher's career. Going back to Lefty Grove. Ordered by Fib.WinPt. on that equivalent 7-year record. (Note that the 7-year interval is chosen to make Maddux look good.)

Grove: 151-63
Maddux: 139-47
Hubbell: 150-72
Martinez: 122-38
Seaver: 144-69
Clemens: 138-62
Gibson: 144-73
Johnson: 124-48
Feller: 145-75
Newhouser: 144-77
Roberts: 151-89
Koufax: 134-67
Marichal: 142-82
Perry: 151-98
Blyleven: 138-82
Schilling: 119-56
Jenkins: 144-94
Palmer: 137-84
Brown: 113-53
Spahn: 133-83
Pierce: 121-68
Bunning: 132-86

Probably no one needs to be reminded of this, but I'll say it anyway: Maddux had a helluva peak. Those seven years all by themselves are probably good enough for the HoM. Even more so if you extrapolate 1994 and 1995 out to normal season length, which I haven't done.

If you want to push that 7-year equivalent record thing back another quarter of a century, you get Walter Johnson with a 198-78 and Pete Alexander with 181-96. But times really were different back then.

DL from MN, I have two questions about your preliminary ballot--

First, why Glavine over Schilling/Mussina? I agree that those three are all close in value, and all in the top 5 on this ballot, but it looks to me like the only things Glavine has over Schilling/Mussina are IP and batting wins, which aren't enough to make up the difference in pitching quality. In many years, Glavine would be a top-of-the-ballot guy, but in this context, he looks to me like the one who should wait until 2015. What puts him over the top for you?

Second, what's your analysis of Jeff Kent? I like Kent, but my initial take puts him more on the borderline than solidly in--I see him as closer to Bancroft/Cash in the just off-ballot group with the current glut at the top than as the best available position player after Frank Thomas! What gets Kent up that high in your estimation?

DL, I'm surprised to see Thomas come in below the trio of pitchers.

My guess for the electorate is a whole is that we'll see Maddux elected in first (possibly unanimously), Thomas in second and one of the trio of similar pitchers in third (Glavine, Mussina and Schilling). Kent should also get decent support and join Sosa at the top of the backlog.

It's the extra bulk for Glavine (1200 innings). I'm a career voter, I don't look at peak at all except how it adds to career totals. Glavine has fewer PWAA but enough WARP2 to make up the difference. I haven't even looked at Glavine's 218 postseason innings for my initial ranking.

Shouldn't surprise you to see Mussina above Thomas. 26th among pitchers is better than 10th among first basemen.

Jeff Kent just barely scores higher than Bus Clarkson. He's the best available position player after Thomas because we've elected all the position players. He's waiting behind Biggio for my PHoM. That said, Dan R's defensive numbers for Jeff Kent give him a positive contribution over his career (+3 FWAA). Not sure that's correct and any adjustments downward could slide him down the list quickly. For example, changing that to -3 FWAA drops him below Norm Cash.

2014 Prelim

1. Greg Maddux, P: Led his league five times in ERA+ and five times in IP, including four years when he led in both. Easy #1.

2. Frank Thomas, 1B: 156 OPS+ is 20th all-time. Led league 3 times in OPS+ and 4 times in Runs Created.

3. Tom Glavine, P
4. Curt Schilling, P
5. Mike Mussina, P: As an Orioles fan, it pains me to put Moose third in this group but Glavine beats him on career (4413 innings to 3562) and Schilling beats him on rate (127 ERA+ to 123).

6. Ben Taylor, 1B: Renowned as a gloveman during the deadball era, Taylor was a better hitter than previously thought. Seamheads gives him a slash line of .339/.400/.467 for an OPS+ of 163. Those come down a little in MLE translations but Taylor's got room to spare.

7. "Cannonball" Dick Redding, P: I’ll keep making the case even though some are jumping off the bandwagon (while others shoot flaming arrows at it). #1 pitcher in 1914/15 (Cuban League), ‘17 and ’19. #1 player in 1917 (25.9 Win Shares in 153 innings). Top three in ‘12/’13, 1915, and ‘15/’16. Top ten in ’12, ’16 and ’21. Great peak, long prime.

8. Jeff Kent, 2B: Best available infielder. 123 OPS+ in 9537 plate attempts. Longer career keeps Kent ahead of Bando despite Bando's better defense.

9. Sal Bando, 3B: Best available third baseman. 119 OPS+, +36 fielding runs.

10. Vic Willis, P: 117 ERA+ in 3996 innings. Big peak, solid prime.

11. Don Newcombe, P: 114 ERA+ in 2154 innings not enough but not the whole story. Negro League, minor league and military service fills out a solid career around his big years.

12. Luis Aparicio, SS: My hobby horse. His 51.7 WAR is no illusion. +92 baserunning, +28 reaching base by error or avoiding double plays, +149 fielding. Did everything well that isn't included in OPS+.

13. Tommy Bridges, P: Top ten in ERA+ 10 times, including 2nd in '32, '33 and '43.

14. Fred McGriff, 1B: Scratching my careerist streak but a 134 OPS+ in over 10,000 plate attempts is nothing to sneer at.

15. Sammy Sosa, RF: Edges out Bob Johnson for the last ballot spot and the best available outfielder.

Required Disclosures

I've voted for Phil Rizzuto has in the past but at this point I prefer the career numbers and consistent prime of Aparicio and Bancroft respectively.

Looking further on Kent, DL from MN wrote:

That said, Dan R's defensive numbers for Jeff Kent give him a positive contribution over his career (+3 FWAA). Not sure that's correct and any adjustments downward could slide him down the list quickly. For example, changing that to -3 FWAA drops him below Norm Cash.

What is your source for these Dan R defensive numbers? The latest Dan R WARP numbers that I have (which go only through 2005 and may have been superseded), show Kent at -1.2 FWAA through 2005. Most of the fielding assessments that I am aware of show Kent as an above average fielder for his career, though the extent to which he is below average varies significantly:

Dan R: -1.2 FWAA (through 2005)
BBRef WAR: -42 fielding runs
Fangraphs WAR: -18.4 fielding runs
DRA: -1.9 fielding runs

The one exception is

Baseball Prospectus: +3.6 FRAA

As you note, the quality of Kent's fielding is key. In my way of looking at merit, Kent shows as clearly but narrowly above the in-out line in BP's WARP, right around the in-out line in Fangraphs and Dan R, and solidly below the in-out line in BBRef.

The fielding assessment you have from Dan R is more positive by two full wins than any others I have seen, so I am particularly curious about its source.

Kent provides a good occasion for considering the relative merits of the various fielding metrics available.

Dan R FWAA2
1992 -0.3
1993 -1
1994 1.4
1995 0.5
1996 0
1997 0.9
1998 0.5
1999 -0.2
2000 -0.8
2001 0.4
2002 -0.3
2003 -0.2
2004 1
2005 -0.1

Then I zero out 1992-1993 because they don't contribute to his overall numbers. Total is 3.1. I have him as -0.1 post 2005.

I'm certainly interested in a thorough assessment of Jeff Kent's defensive contribution. It is critical to his placement.

Taylor was a better hitter than previously thought.

How much? I have him in the spreadsheet comparable to Jake Beckley offensively - another low peak, long career first baseman. If he was 10% better than that he moves to 7th (behind Bridges) on my ballot. Is anyone planning on redoing the MLEs?

Is anyone planning on redoing the MLEs?

Yes. I'm working on converting my rankings of major-league players to be based on the second- and third-generation WAR stats now available. Once I finish that conversion, I plan to start re-doing MLEs for Negro League players based on the improved statistical record. I'll start with Ben Taylor and Dick Redding, because Gary Ashwill's research project has been most significant for the pre-1920 stars.

How much? I have him in the spreadsheet comparable to Jake Beckley offensively - another low peak, long career first baseman. If he was 10% better than that he moves to 7th (behind Bridges) on my ballot.

The problem with describing Taylor as a low peak, long career first baseman is that MLEs tend to flatten a player's peak as they often incorporate regression to the mean.

In my case it doesn't really matter. I don't look at peak.

2014 Prelim:

1 Greg Maddux - Top 10 pitcher - in the discussion with Pete Alexander, Lefty Grove, and Christy Mathewson.
2 Curt Schilling & - Fergie Jenkins, Phil Niekro, Nolan Ryan, Ed Walsh territory
3 Mike Mussina - Top 25 pitchers
4 Frank Thomas - Top 70 position player - Well below Hank Greenberg and Jeff Bagwell, well ahead of Jim Thome and Willie McCovey
5 Tom Glavine - Top 30 pitcher - Kevin Brown, Hal Newhouser, Red Ruffing, Juan Marichal territory
6 Sammy Sosa - near HOM median line, a touch below Sam Crawford and Tony Gwynn, a notch ahead of Harry Heilmann and Enos Slaughter
7 Phil Rizzuto - last no brainer from the deep backlog with war/malaria credit.
8 Tommy Leach
9 Buddy Bell
10 Don Newcombe
11 Hilton Smith
12 Bert Campaneris
13 Gavvy Cravath
14 Johnny Pesky
15 Urban Shocker
Vic Willis - PHOM, but off ballot
Ben Taylor - potential for PHOM, waiting for MLEs
Dick Redding - close to PHOM, but seamheads hasn't helped his case

Check out the mainland thread for Jeff Kent, particularly Dan R's comments:
http://www.baseballthinkfactory.org/newsstand/discussion/mlbcom_jeff_kent_retires/

My no vote for Jeff Kent prevented him from being elected to a different shadow hall of fame:


Some Dan notes:
67. David Concepcion de la Desviacion Estandar (Dan R) Posted: January 21, 2009 at 08:03 PM (#3057611)
Kent definitely falls short for me. Here's how he looks in my WARP.

Year SFrac BWAA BRWAA FWAA Replc WARP
1992 0.50 -0.1 0.0 -0.3 -1.0 0.6
1993 0.79 0.5 -0.2 -1.0 -1.3 0.6
1994 0.93 1.4 0.0 1.4 -1.6 4.4
1995 0.84 1.2 -0.2 0.5 -1.3 2.8
1996 0.69 0.6 0.3 0.0 -0.8 1.7
1997 0.95 0.5 0.2 0.9 -1.4 3.0
1998 0.86 3.4 0.0 0.5 -1.4 5.2
1999 0.84 2.0 -0.3 -0.2 -1.4 2.9
2000 1.00 6.0 -0.1 -0.8 -1.5 6.7
2001 1.02 3.7 -0.3 0.4 -1.5 5.3
2002 1.00 4.3 0.4 -0.3 -1.5 5.8
2003 0.81 1.6 0.0 -0.2 -1.3 2.7
2004 0.88 1.7 -0.4 1.0 -1.5 3.9
2005 0.93 3.1 0.1 -0.1 -1.5 4.7
2006 0.69 1.8 -0.3 -0.9 -1.1 1.7
2007 0.81 1.6 -0.1 -1.0 -1.4 1.9
2008 0.69 -0.2 -0.2 -0.9 -1.2 -0.1
TOTL 14.23 33.1 -1.1 -1.0 -22.7 53.8
TXBR 13.54 33.3 -0.9 -0.1 -21.5 53.9
AVRG 1.00 2.3 -0.1 -0.1 -1.6 3.8


In my salary estimator, that comes out to $141 million, where $150M is roughly the Hall in/out line. The combination of 2B being easier now than it was before 1980 and the high standard deviation of the high-scoring, doubly-expanded NL after 1998 just sets him back too far. There are a lot of supposedly borderline modern players I support--not just Larry Walker and Jim Edmonds who I think are no-brainers but Brian Giles and Jason Giambi--but Kent isn't one of them.

Dan explains the defensive metrics he uses for Kent:

69. David Concepcion de la Desviacion Estandar (Dan R) Posted: January 21, 2009 at 08:41 PM (#3057620)
It's my fielding wins, which are a weighted average of Zone Rating and Retrosheet data that shows an overwhelmingly high correlation (r > .85) to an average of modern PBP metrics.

That said, Kent wasn't bad with the glove; he was average. He just didn't hit well enough for long enough to meet the Hall standard for average-fielding, average-running, 1990's/2000's NL 2B.


Dan muses later in the thread, post 79:

1. 2B is not what it used to be. In the 1970's, 2B was truly a "middle infield" position, second only to shortstop in terms of how poor hitters replacement players at the position were. Scrap heap catchers and third basemen were far better offensively than their 2B counterparts. After 1985, that was no longer the case--replacement 2B improved their hitting substantially, to the point where today they are indistinguishable from 3B and CF (as Tangotiger also notes). As a result, Kent (and all modern 2B) need to be held to a higher standard offensively than their historical predecessors, because the position has evolved into a mid-defensive spectrum role.

2. Kent compiled his highest OPS+ marks in one of the easiest eras to dominate in major league history, the 1998-2002 NL. The combination of two expansions and extremely high run scoring meant that hitters' performances were spread out substantially more than they were in, say, the early 1980's NL. A given OPS+ mark (100, 125, 150, whatever) in 1998 "bought" notably fewer pennants than it would have 15 years before. (Note that there is no mention of steroids in this statement).

3. Given these contextual factors, you simply can't compare Kent's 9,537 PA at a 123 OPS+ to, say, Bobby Grich's 8,220 at 125 and conclude that they had similar offensive value, or that Kent was slightly superior. (In fact, if we ignore baserunning and fielding, I have Grich at 65.2 wins above replacement and Kent at 54.9). Given how easy it was for teams to find a "decent" OPS+ at 2B during Kent's career, he simply didn't separate himself enough from that benchmark to meet the Hall's overall standard. (For the same reason, I am much less friendly to Alomar than most, although I still think he deserves enshrinement).
He further explains his method for calculating replacement in post 93:


the *overall* replacement level doesn't change, just the *relative* replacement levels of the different positions. My system only attempts to measure players' contextual value as accurately as possible. I don't know for sure why middle infielders hit so much better now than they used to, although I have a pretty strong hunch it has something to do with the reduction in turf fields and the changing emphasis in the game on power over speed. But it indubitably has happened, and that's what counts when we are trying to determine how many wins a player was worth to his club.

I trace the evolution of the defensive spectrum over time by looking at the performance of the worst 3/8 of MLB starters over rolling nine-year periods. That means that for Kent, we're talking about literally 100 player-seasons (which exclude the performance of all above-average players that might skew the curve with their once-in-a-generation excellence). The odds that a sample this large would be influenced by fluctuations in the talent are sufficiently small that I see no reason to doubt the results.

I'll have to update with the new 2006-2007 numbers.

Mike (#3) has "The Pirates of the aughts had a lot of one-year and two-year wonders who never did much anywhere else; their pitching staff was mostly Willis and a bunch of guys who would have been lucky to get 10 starts for the Giants or Cubs...."

I have GREAT respect for Mike's analyses, but this simply is not true. The Pirates in the oughts had Sam Leever and Decaon Phillippe, who would certainly have been in the rotation of the 1906 Cubs, although Three-Finger Brown would still have been the ace. They also had some people like Ed Doheny and Lefty Leifield, who would have been on the 1906 Cubs' rotation in their good years. Besides that, Willis only pitched 4 years for the Bucs: 1906-1909. He did pitch a little better for the Pirates than he had pitched previously for Boston, where he actually led the league in Losses twice. The Pirates of this period had few strikeouts because they were ground ball curveball pitchers with tremendous control. That meant that everything went down, and was a strike, but you could hit a grounder off of it any time. Fred Clarke collected curveball pitchers, including Willis, Deacon and Sam, Jesse Tannehill, and Babe Adams. He did have very good infield gloves out there, but the pitchers he collected knew how to use that - induce grounders. That is probably why Willis pitched a bit better in Pittsburgh. He knew how to induce grounders. He also may have suffered from some large workloads in Boston. - Brock Hanke

here's my top few as of now. i would personally vote for 8 in hof:

1) maddux - i have him as 4th best pitcher ever (johnson, young, clemens) both in raw terms and adjusted for historical epoch

top four stay the same by both calculations but after that they differ:
raw 5-10 alexander, mathewson, nichols, pedro, grove, randy (seaver)
adj 5-10 pedro, randy, grover, grove, seaver, mathewson (nichols)

so as you can see, i have a consensus group of four, then a consensus group of 7 for a consensus 11. big gap after #11 in both cases

2) thomas - my deduction for 1b is low compared to bad fielders at other positions so he doesn't drop off like in bbref war. 1b is my basis b/c highest # chances, lowest error rate. also he quit fielding after a while. i have him 16th in my hitting score b/w wagner and lajoie, with a gap to lajoie (and another big gap after 18th)

for my top hitters if you are wondering, i have ruth, gap, bonds, gap, cobb, williams, gap, usual suspects

3) schilling - i have him as 15th best pitcher, between gibson and perry

4) mussina - i have him as 21st best pitcher, b/w walsh and marichal

5) babe adams - i have him as 40th best pitcher, led league in whip 5 times (1 was only 160 IP though)

6) tom glavine - i have him as 41st best pitcher, and he does have 300 wins

7) sosa - i have him 83rd on my hitting score, way below mcgwire, walker, mcgriff, palmeiro. surprisingly low. above general hof standard still, but does not look good compared to contemporary candidates. haven't evaluated him for fielding or baserunning yet.

8) jeff kent - have him in 100-150th range for best hitter score, good for 2b, don't see a huge penalty for fielding or baserunning but haven't evaluated him on those yet.

other notes:

tommy bridges - have him as 61st best pitcher
urban shocker - have him as 70th best pitcher
bucky walters - have him as 86th best pitcher
vic willis - 96th best pitcher
--brad radke line--
don newcombe - 123rd best pitcher

don't really want to vote for any of these pitchers if i can help it but i'll just put them on the end in order if i have to (unless one of them was a really terrible fielder or something). newcombe is below my brad radke hellz-2-tha-nah line, though. (i am reading colbert's book)

and yes, morris is also below all of those guys. but he's still above marquard.

fred mcgriff - 53rd hitting
norm cash - 67th hitting

these 2 are below sosa/kent to my fan-heart but should i put these 2 above sosa/kent? that's what i'll be considering. i'll do fielding/baserunning evaluations on these 4 guys to see if it enlightens me at all.

aparicio - top 5 running score (so far), good fielder except first (one) and last two seasons, obviously bad hitting score (ahead of only maranville and vizquel at this point and barely - yikes)

pitcher rank is just pitcher score, fielding not considered. score curved for historical epoch and anomalous UER percentage in a season. taken around league averages. i also do a historical curve for fielding that is less steep. but i am only up to the 1890s as far as doing actual ratings (plus a few modern test cases). i do not curve historically for baserunning or hitting, since i use stats that are already accounting for year to year fluctuations and there is no obvious "worse" state in the past to base said curve on (lower fpct for fielders, high uer/er ratio for pitchers).

disclaimers: i do not use pbp data, i do not consider leagues outside of na, nl, al. no ua, aa, fl, pl, nnl, nal, cl, wl, etc. credit from me

sorry if you are swamped by my commentary, i just started commenting on here again and i was eviscerated last time i tried to vote on one of these (i had only done a pitching score then) so i feel compelled to give more detail

jdennis - for a hall of merit ballot you just rank your top 15 players. Make sure you look at the required disclosures.

disclaimers: i do not use pbp data, i do not consider leagues outside of na, nl, al. no ua, aa, fl, pl, nnl, nal, cl, wl, etc. credit from me

Just a clarification: you'll eventually need to look at players in these leagues, in order to be fair to all players from all eras. The PL is pretty easy since it was nearly NL quality (or maybe a bit better); the AA/FL/UA are a bit tougher but I think there was some work done on AA/NL and UA/NL conversions on an old thread somewhere. The Negro Leagues are tougher still but the Negro Leagues players' threads are a very good source of info and opinions; additionally both BBREF and Seamheads.com have (incomplete) Negro Leagues statistics.

Thanks for those numbers. I was overestimating his 2006-2007. Kent now drops just behind Sosa for me. Borderline 2B candidate but beats the hell out of Nellie Fox.

Kent was pretty good in the postseason - .840 OPS in 189 PA. A lot of these new candidates have an extensive postseason record to consider.

I just had to comment somewhere that the next HOF vote is going to be CRAZY.
It was already a super crowded ballot and no one was elected.
Now you add into that Maddux, Glavine, Thomas, Mussina, and Kent.
That's unreal!

So, Maddux is the obvious #1 and I’m assuming he’ll be unanimous. But the separation between Schilling, Glavine, and Mussina is far less obvious. Watching them in their prime, I would’ve taken Mussina, but then again, I was biased then because I’m an Orioles fan. So, how do the stats compare among the three? We’ll progress from basic to advanced.

Innings pitched: Glavine (4413.1), Mussina (3562.2), Schilling (3261).
200+ Inning Seasons: Glavine (14), Mussina (11), Schilling (9).
Top Inning Season: Schilling (268.2), Glavine (246.2), Mussina (243.1).
Innings per game: Mussina (6.6), Glavine (6.5), Schilling (5.7).
So Glavine has a big edge in innings pitched, although his rate per game is virtually the same as Mussina’s. In Schilling’s big years, however, he put up slightly more innings than Mussina or Glavine.

Strikeouts: Schilling (3116), Mussina (2813), Glavine (2607).
200+ K Seasons: Schilling (5, with 3 being 300+), Mussina (4), Glavine (0).
Top K Season: Schilling (319), Mussina (218), Glavine (192).
Schilling with the big edge in strikeouts, although I’ll admit I was surprised Glavine never reached 200 K in a season.

Career ERA-: Schilling (80), Mussina (82), Glavine (86). Thought Glavine would be closer to Schilling and Mussina, but let’s take a look at each pitcher’s 5 best ERA- seasons and how they compare to each other:
1. Glavine (59)
2. Mussina (63)
3. Mussina (64)
4. Schilling (65)
5. Schilling (66)
6. Schilling/Glavine (67)
8. Schilling/Schilling/Mussina (69)
11. Glavine (70)
12. Glavine (71)
13. Glavine/Mussina/Mussina (72)

If you add the rankings up (so lower the number the better), Schilling ends up with 31, Mussina with 39, and Glavine with 43. Overall, though, no one stands out too much as the numbers are all pretty close. Let’s do the same thing with FIP (Schilling career 74, Mussina career 81, Glavine career 94)-:

1. Schilling (54)
2. Schilling (58)
3. Schilling (60)
4. Schilling (63)
5. Schilling/Mussina (68)
7. Mussina/Mussina (71)
9. Mussina (76)
10. Mussina (77)
11. Glavine (79)
12. Glavine (80)
13. Glavine (81)
14. Glavine/Glavine (82)

As you can see, Schilling dominates, with a score of 15, Mussina follows with 38, and Glavine finishes with 64. It is pretty amazing that Mussina’s best seasonal FIP- is the same as Schilling’s fifth-best, and that Glavine’s best is worse than Mussina’s fifth-best.

So to sum up this section, it’s clear that Schilling > Mussina > Glavine based on rates, but that Glavine > Mussina > Schilling when it comes to career innings pitched. So how to weigh this information to see who to rank where? Well, let’s see how the various WAR systems illustrate it:

B-R: Mussina (78.1), Schilling (76.9), Glavine (69.3)
FG: Schilling (86.1), Mussina (85.6), Glavine (68.5)
BG: Mussina (78.6), Glavine (76.4), Schilling (73)

B-R JAWS: Schilling (62.3), Mussina (60.3), Glavine (53.7)
FG JAWS: Schilling (69.7), Mussina (63.9), Glavine (51.4)
BG JAWS: Mussina (61.3), Schilling (59.6), Glavine (59.2)
AVG JAWS: Schilling (63.9), Mussina (61.8), Glavine (54.8)

This ranking is the same as the rate rankings, although there is not a lot of separation. Rather than listing the top 5 seasons for each pitcher for each metric, here is the total top 5 season WAR for each pitcher via each metric:

Schilling: 36.2 B-R WAR, 41.8 FG WAR, 35.4 BG WAR
Mussina: 32.1 B-R WAR, 31.3 FG WAR, 33 BG WAR
Glavine: 29.6 B-R WAR, 25.4 FG WAR, 31.4 BG WAR

When also considering playoff performance, it would seem Schilling also adds to his advantage although Glavine and Mussina aren’t slouches either:
Schilling: 133.1 IP/2.13 ERA/13 starts 7+ IP, 2- ER/120 K
Mussina: 139.2 IP/3.42 ERA/8 starts 7+ IP, 2- ER/145 K
Glavine: 218.1 IP/3.30 ERA/16 starts 7+ IP, 2-ER/143 K

However, there are some other considerations. Both Glavine and Mussina were great fielders, while the same cannot be said of Schilling. Mussina always had to face a DH, which Glavine didn’t and Schilling only had to do at the end of his career. Schilling and Glavine’s postseason opportunities coincided with their best seasons, something not always true for Mussina. But most importantly, two of Mussina’s best years according to ERA- were the strike years of 1994 and 1995, something that can’t be said for Schilling or Glavine (Schilling in fact was below average in 1994). I haven’t adjusted for this in the above, as people disagree on whether to flat adjust or adjust with some regression. While I’m not sure it’s enough to put Mussina ahead of Schilling, right now I’m thinking of ranking Mussina just barely ahead of Schilling, with Glavine bringing up the vaunted rear, even though I greatly respect his innings total.

Glavine was the best hitter of the three. How significant is that going to be in the consideration (not much, but it had to earn the odd win here or there).

Normally, I don't really care about pitcher hitting, but with Glavine it looks like it worth really incorporating. That being said, I don't know if it's enough to change the rankings of these three. I also didn't want to include it in my analysis above since for Mussina, there really isn't any hitting data since he spent his whole career in the AL.

One good thing is that pitcher hitting replacement level is that of the average hitting pitcher. No one chooses pitchers for their offense, so the replacement level is league average. Which means (luckily) that AL pitchers aren't hurt or helped by their not being allowed to hit in comparison with NL pitchers, on average.

I've emailed Sean Forman about how B-R calculates replacement level for pitcher offense . . . I don't think they are doing this. I get Jim Kaat at 29 RAR hitting, but B-R has him at 56, so my first inclination is that they don't use average pitcher hitting as replacement level. Haven't checked the other sites.

Regarding comment 26, i performed an analysis in the maddux thread that included hitting stats, showing mussina and schilling as a coin flip weighting equally sean smith, baseball reference, baseball gauge, and fangraphs war.

Glavine pulls up the clear rear using sean smith and fangraphs, fangraphs is based upon fip, and glavine made great use of weak contact.

With war and integration credit, i have don newcombe at the top of the backlog.
with wwi credit, urban shocker falls next in line.
a cluster forms after these 2 two.

The most baffeling to me is how well george uhle comes out, no one has voted for him, but he comes out hall of merit worthy with the 4 systems, thanks to hitting value...does anyone else like uhle?

JoeD/29:

BR describes their positional adjustment for pitchers here: they set replacement level such that pitchers sum to zero WAR for their position-player contributions (batting, baserunning, GIDP), rather than zero for hitting alone. There's also a list of the yearly pitcher position adjustments. It seems like this shouldn't make a huge difference, though, since pitchers are rarely on base enough to do much harm with their terrible baserunning.

As an aside: this assumption may fall apart in very early baseball, when many "pitchers" were in fact two way players, and therefore selected partially on the basis of their non-pitching skills. For instance, Jack Stivetts' BR page shows a weird relationship between RAR and WAR--an RPW of nearly 30 when he was primarily a pitcher, dropping abruptly to 10 when he became a full-time OF. Other pitchers, also, have a much higher RPW than position players during this era. I'll look into this effect more but it may be that BR WAR unduly penalizes two-way players when they were common.

Good commentary alex, bob caruthers was one of the biggest losers in the new baseball reference war.

Tom Glavine is a great argument to ignore FIP entirely.

I don't remember evaluating George Uhle.

Upon further review I'm not really sure what's going on here...I looked at 1922 and 1923 pitcher batting and indeed, the total pitcher* hitting WAR was zero. But for 1876 and 1881, this was not the case. In 1881, total pitcher RAR was 32.3, which is I think as it should be: pitchers who were better hitters received more PT as position players, skewing the averages upward. But total pitchers as hitters WAR was negative 4.7 WAR, which is bizarre. I may email Sean Forman to see if he knows what's going on here, but this is a pretty minor point. Joe, has he responded to your question yet?

*Using BBREF's demarcation of pitcher/not-pitcher, which defines someone as a pitcher if they were a pitcher in a plurality of their games played.

Quick look:

1) Maddux
2) Thomas
(gap)
3) Mussina
4) Glavine
5) Schilling

Yes, Glavine is Tommy John plus. But we're under-rating John (and groundball-heavy pitchers in general, though we latched onto Rick Reuschel).

I anticipate #6-#12 being the same as 2013:
6) Dolf Luque - receives color line credit
7) Wally Schang - mystified why he earns so little support
8) Tommy John - cf. Glavine comment
(gap)
9) Sammy Sosa
10) Hilton Smith
11) Buddy Bell
12) Luis Tiant

Some movement at the bottom:
13) Ben Taylor (new)

I was trying to justify "Norm Cash on, Fred McGriff off", which led me to a general reappraisal of 1B and the conclusion that Ben Taylor stood out farther from his contemporaries than either of the two. This is a cautious placement; he may rise even higher (into the Luque/Schang/John cluster).

14) Lee Smith (was #13)
15) Kenny Lofton (unch.)

Jeff Kent just misses the ballot. He's the Larry Doyle of his era. In the "post-integration 2B/3B" realm, Kent sits above Bando but below Bell.

Rizzuto and Willis are "top 25, not top 15". I'm conservative with war credit and think we made a mistake by inducting Charlie Keller.

Redding is a Bobo Newsom clone; great fastball, bounced around, not a consistent plus pitcher. Hilton Smith (a "Bucky Walters plus" candidate) should be ahead of him in the NeL queue.

I WANT badly to vote for Boomer Wells, but being Herb Pennock 2.0 doesn't qualify you for the Hall of Merit.

33. DL from MN Posted: January 13, 2013 at 10:41 AM (#4346077)
Tom Glavine is a great argument to ignore FIP entirely.

I don't remember evaluating George Uhle.

Tango suggests a 70/30 rWAR/ fWAR split:
insidethebook.com/ee/index.php/site/comments/gwar_the_latest_in_the_war_implementations

Back to Glavine/Mussina/Schilling

B-R: Mussina (78.1), Schilling (76.9), Glavine (69.3)

That includes negative WAR seasons for Glavine at the beginning and ends of his career. Let's zero those out for everyone. I also zero out seasons where WAA + WAR < 0.

Altered B-R PWAR: Mussina (78.1), Schilling (77.8), Glavine (69.6)
Altered B-R PWAA: Mussina (48.9), Schilling (55.1), Glavine (41.6)

This makes Glavine look like the player to place at the bottom until...

you add in BWAR: Mussina (0.1), Schilling (-0.8), Glavine (7.3)

That makes the numbers:

WAR: Mussina (78.2), Schilling (77.0), Glavine (76.9)
WAA: Mussina (49.0), Schilling (54.3), Glavine (48.9)

They're within the error bar in WAR. I think it requires a systematic look at postseason records to really separate these pitchers. Also, any peak analysis is going to require postseason records get tacked on to their best seasons.

Overall postseason records:

PSPitcher W  L  W-L%  ERA  G  GS  GF  CG  SHO  SV  IP  H  R  ER  HR  BB  IBB  SO  HBP  BK  WP  BF  WHIP
ToGlavine 14  16  .467  3.30  35  35  0  3  0  0  218.1  191  91  80  21  87  11  143  7  1  3  912  1.273
MiMussina 7  8  .467  3.42  23  21  0  0  0  0  139.2  121  56  53  19  33  7  145  3  0  2  571  1.103  
Schilling 11  2  .846  2.23  19  19  0  4  2  0  133.1  104  37  33  12  25  0  120  3  0  1  525  0.968

I think Schilling pretty clearly bests Mussina (who is no slouch) in around the same number of innings. Glavine (once again) has the bulk argument. Glavine does have 11 UER which is pretty high.

PSBatter G  PA  AB  R  H  2B  3B  HR  RBI  SB  CS  BB  SO  BA  OBP  SLG  OPS
ToGlavine 36  70  59  4  10  1  1  0  5  0  0  5  18  .169  .234  .220  .455
MiMussina 23  4  4  0  0  0  0  0  0  0  0  0  4  .000  .000  .000  .000
Schilling 19  28  24  1  2  0  0  0  0  0  0  0  12  .083  .083  .083  .167  

Glavine provides a lot more postseason value with the bat.

All I can shake out is that Mussina belongs on the bottom of the list.

DL, I'm sold. New prelim, which should stay the same for awhile. Parentheses indicate 2012/2013 ballot placement.

1) Greg Maddux
2) Frank Thomas
3) Tom Glavine
4) Curt Schilling (NA/5)
5) Mike Mussina
6) Adolfo 'Dolf' Luque (1/6)
7) Wally Schang (2/7)
8) Ben Taylor (off/off)
9) Tommy John (3/8)
10) Sammy Sosa (NA/9)
11) Hilton Smith (4/10)

[Break point - I strongly believe that the top 11 deserve HoM enshrinement, but I'm not 100% certain for those below]

12) Buddy Bell (7/11)
13) Luis Tiant (6/12)
14) Lee Smith (10/13)
15) Kenny Lofton (NA/15)

PSBatter G PA AB 
ToGlavine 36 70 59

There are 6 sac bunts in there for Glavine too which I consider a positive.

A peak analysis is going to arrange things differently but I wouldn't consider a peak analysis that didn't take into account postseason innings or hitting. Schilling's peak is going to look pretty great with those postseason innings tacked on.

The way I consider "fair to all positions and all eras" the #30 pitcher is about even with the #10 first baseman. That's what keeps me from ranking Thomas ahead of the pitchers.

Sorry, no unanimity for Maddux... but awfully close.

1. Frank Thomas
2. Greg Maddux
3. Sammy Sosa
4. Jeff Kent
5. Fred McGriff
6. Tommy Leach
7. Mickey Welch
8. Tony Perez
9. Bobby Bonds
10. Luis Gonzalez
11. Vic Willis
12. Buddy Bell
13. Hugh Duffy
14. Tom Glavine
15. Jim McCormick

16-20. Lofton, Schilling, Finley, Grace, Redding
21-25. Redding, Mullane, Brock, Cash, Olerud
26-30. Streeter, Bando, John, Grimes, Greene

Both Rizzuto and Taylor are late in my Top 100

The way I consider "fair to all positions and all eras" the #30 pitcher is about even with the #10 first baseman.

With the much larger number of pitchers required on a team than 1B, shouldn't that be more like #40-50 (Or MAYBE even higher?)

I am not a big fan of relative rankings by position (as could be seen in my relative lack of support for Mike Piazza), but if you are going to do them...

Rusty- no Mussina?

#30/#10 is about 27% pitching in the Hall of Merit, which is around where the % is at the moment. I'm just slightly kinder to pitchers than the overall Hall of Merit. Thing is, I don't really adjust the numbers much to get there. I have thought about making that adjustment but held back partly because there wasn't always a 5 man rotation and partly because the value for pitchers is spread out much thinner after the development of the modern bullpen. I think 30% pitching is a fairly conservative amount and I could see 33% based on the pitching/defense/offense value split. That would make the #33 pitcher even with #10 at any position and would give a quota of 83 pitchers elected. We've elected 67 pitchers and two of those (Dihigo and Caruthers) aren't full-time pitchers. Even I can't quite get to being 26 pitchers short.

I don't understand your ballot in the slightest. Every GM in baseball would have traded Frank Thomas to obtain Greg Maddux. They also would have dealt Luis Gonzalez to get Glavine or Schilling or Mussina (who isn't even listed) too. It's one thing to disagree with ranking between eras (Welch over Glavine) but I don't grok your ranking of modern players.

#30/#10 is about 27% pitching in the Hall of Merit, which is around where the % is at the moment.

Okay, now I understand what you are getting at. As I say, I am not a fan of ranking by position so it doesn't affect me, but all knowledge is good knowledge!


And so you know, Luis Gonzalez is the player I am most 'uncomfortable with' at the moment. He could easily drop.

As for Thomas and Maddux... I would rather have Thomas, but there is a hair's-width between them.

I have Mussina around 50. He could move up, theoretically, but not likely enough to get on the ballot. I think he is overrated.

Scanning the nubmers in the header, another way to look at it is that I think that WS (Where he is below Kent, Gonzalez and Alou) are a better gauge of his value than WAR (where he is above Glavine).

You sure, DL from MN? Given the relative attrition rates for hitters and pitchers, it wouldn't've surprised me at all if most GM's would have preferred Thomas. Maddux certainly has a ton more career value than Thomas does, but are his best 8-9 seasons (arbitrarily endpointing to benefit the Hurt) really that much greater? Frank Thomas was the best pure hitter (at peak) between Ted Williams and late-career Barry Bonds, was he not?

Maddux certainly has a ton more career value than Thomas does

Nuff said

Does Joe still calculate his Pennants Added? I would guess that Maddux would be ahead of Thomas, but that it would be fairly close.

So I redid the JAWS figures with giving credit to Mussina and Glavine for time missed in 94 and 95 and for hitting for Glavine. The results below:

B-R JAWS: Mussina (62.5), Schilling (62.3), Glavine (60.1)
FG JAWS: Schilling (69.7), Mussina (64.8), Glavine (57.5)
BG JAWS: Mussina (63.8), Glavine (63.3), Schilling (59.6)
AVG JAWS: Schilling (63.9), Mussina (63.7), Glavine (60.3)

Mussina didn't do as well as I thought, and I don't see how one can rank Mussina ahead of Schilling, as Mussina's better fielding is more than trumped by Schilling's postseason excellence. Adding Glavine's hitting certainly brings him right into the fold, but it's not enough to rank him ahead of Schilling or Mussina for me.

Also, in case anyone's wondering about the other two big additions to the ballot (these are adjusted for 94 and 95):

B-R JAWS: Maddux (80.6), Mussina (62.5), Schilling (62.3), Glavine (60.1), Thomas (59.1)
FG JAWS: Maddux (90.7), Schilling (69.7), Thomas (65.8), Mussina (64.8), Glavine (57.5)
BG JAWS: Maddux (88.6), Mussina (63.8), Glavine (63.3), Thomas (59.7), Schilling (59.6)
AVG JAWS: Maddux (86.6), Schilling (63.9), Mussina (63.7), Thomas (61.5), Glavine (60.3)

As you can see Maddux is really in a world of his own on this ballot. While Thomas does have a better AVG JAWS than Glavine, I'm still going to rank Glavine ahead of Thomas on my ballot due to the positional argument that DL has eloquently explained along with the fact that I also deduct Thomas some in ballot rankings due to the amount of time he spent as a DH (although no where near as extremely as Brock does).

Yeah. I have no idea where Thomas will end up on my ballot, but this is going to be the biggest DH deduction I have ever made. Right now, I'm trying to figure out some way of assigning a negative value to zero performance within Bill James' Win Shares model, where the zero point is half of the league runs scored or 1.5 times the number of league runs scored, depending on whether you're looking at offense or defense. But at minimum, I will deduct Thomas as much as the worst full season in MLB history by a full-time field player, on the grounds that anyone playing the field would have at least one chance successfully accepted in a season, which is more than a DH can say. I tend to regard Tom Glavine as being the modern Pud Galvin, rather than Tommy John plus, but I don't really know if there's any difference between the two viewpoints. Both are essentially saying lots of IP, not a great rate. - Brock

"But at minimum, I will deduct Thomas as much as the worst full season in MLB history by a full-time field player, on the grounds that anyone playing the field would have at least one chance successfully accepted in a season, which is more than a DH can say."

If you can read that sentence again and still reach the same exact conclusion - I fear for you.

Some guys are helping the team by not playing the field. As long as the rules in his league allow a DH he should not have anything deducted beyond the normal positional adjustment.

One thing I once discussed with Bill James was his defensive spectrum. It's not JUST based on the defensive importance of each position. If a player throws LHed, he has no choice but to be an outfielder or a 1Bman. That's also part of the reason why players at 2B, SS, 3B and C don't hit as well as 1Bmen or OFers. Since most players who throw LHed also bat LHed, many of the greatest hitters of all time could not possibly have played 2B, SS or 3B, and were very unlikely to be catchers.

Babe Ruth played SS at the boys school when he was a young teenager, but that wasn't gonna happen in the major leagues.

The best combo is to throw RHed and Bat LHed (or switch hit). Then you can play anywhere in the field, AND have an advantage as a hitter.

Brock, that's not how defensive value accrues. If it did, we'd have a HoM of all catchers, since they get like 100 chances a game for a passed ball that they avoid, and maybe a smattering of first basemen. Defensive value is chances times difficulty of chances times run impact per chance. If you make one HR-saving catch in the outfield, which, say, only 5% of OF would make, then the other team's run expectancy on that play is (.95*1.44) + (.05*-.27) = +1.35 runs with an average OF, and -.27 runs with you, for a contribution of 1.62 runs. If a first baseman has a 99% chance of catching throws from his infielders, then the other team's run expectancy when the ball is thrown to him is (.99*-.27) + (.01*.5) = -.262 runs. After he catches it, it's -.27 runs. So his contribution is worth .008 runs. That means that one HR-saving catch is worth the same as 1.62/.008 = 202.5 throws successfully caught at first base.

Now, since DH's don't field, they are all league-average fielders for their position. What determines their overall value is the replacement level--how much better scrap-heap DH's hit than scrap-heap 1B. Historically, that's been about 10 runs a year. If you don't like that approach, you can ask how badly DH's would play 1B if they had to. I've studied the aggregate fielding at 1B of guys who DH at least 3/4 of the time, and they come out at around 9 runs below average a year. So whichever way you approach it, the magnitude of the DH positional value penalty is about the same.

It is flat out wrong to give the same defensive value to DH's as the worst fielding seasons in MLB history. If DH's were traded to the NL and forced to field, they would NOT be put at a position where they'd cost their teams 35 runs with the glove. They'd be put at first base, where they would cost their teams around 9-10 runs. There's no more reason to give DH's a -35 than there is to give, say, LF a -35 because he'd be -35 at shortstop.

It is flat out wrong to give the same defensive value to DH's as the worst fielding seasons in MLB history. If DH's were traded to the NL and forced to field, they would NOT be put at a position where they'd cost their teams 35 runs with the glove. They'd be put at first base, where they would cost their teams around 9-10 runs.

I usually assess around -25 runs defensively for a DH (positional adjustment plus defense), which is a little more punishing than that, but not outrageously so. You figure that a 1B ten runs below average is around -20, but that 1B suffers a bit more physical wear over the season due to fielding and allows the team a little more flexibility to put someone else at DH. I think that's worth something in considering a player's value.

Dan (#53) has " Defensive value is chances times difficulty of chances times run impact per chance."

This is exactly where I always end up; I agree with this sentence. But, for a DH, defensive value is, according to the sentence, zero times zero times zero (= zero). DH has no chances. No chances of any difficulty. No run impact per chance, because no chances. This is what I mean by "absolute zero." No chances accepted. No defensive value. I'm pretty sure I have the concept right. It's quantifying how much this version of zero is, compared to other systems' versions of zero, that causes me the problems. Pete Palmer has zero as the league average. WAR systems have zero at the replacement rate, which depends on the WAR system at hand. Bill James has zero at the margin. But all of those zero points have at least the theoretical possibility of playing worse than that - having a negative defensive number. I think, very strongly, that no one can have less defensive value than the DH. Zero times zero times zero may equal zero, but that particular zero has to be lower than anyone else's defensive number, and all other systems have some people with negative numbers. I don't think that should be possible. No one can field worse than a DH. No one else has the chance to have that little defensive value. So I have to set my DH "zero" somewhere below what other systems generate as a negative number. It's trying to figure out how far below other systems' zeroes that is causing me the problem.

Another way I phrase this to myself is this: We start out with absolute zero as a player who has never played or at least has never accepted a chance. When he does play, and accepts chances successfully, then we can attribute the value of those accepted chances to the player. Anyone who actually plays a position will accumulate some defensive value over that starting point zero, because, no matter how bad he is, he will accept some chances successfully. You could put me out there at shortstop, and eventually someone would hit an easy grounder right at me and I'd successfully accept the chance. But a DH won't. He just stays there, at absolute zero. That's my starting point in the analysis. No one has been able to talk me out of that starting point, although I'm very willing to listen to people try. I'm aware that my starting point and zero point are, if not unique, very unusual in sabermetrics. That doesn't mean I think I'm wrong, but it does imply that I ought to listen to opposing opinions. So, I do appreciate your taking the time to put up your opinion. And I agree with your sentence. But when all the numbers in an equation are zero, and all you're doing is multiplying, then all you're going to get is zero. You can't get a positive number and you can't get a negative number, either. That's where I'm hung up. - Brock

Brock,

Do you deal with DH-era AL pitchers' hitting in the same way? If you don't, then your reasoning for dealing with pitchers who never get an at bat differently that you deal with batters who never get a fielding change might be the starting point for convincing yourself that you need to handle DHing differently.

It seems to me that the key issue for DHs, from a Hall-of-Merit perspective, is that players' merits be evaluated fairly. Quantifying players' value accurately is probably the single most beneficial approach to evaluating merit fairly, but it's not the sole consideration. In this case, the question as I see it is whether your approach to the DH is fair to the DHs and fair to the first basemen. It looks to me like your quest to identify absolute zero, while an interesting intellectual problem, is simply unfair to HoM-discussable DHs, most of whom could and would have been playing first base if there had been no DH, because, even with their defensive liabilities, they still would have been more valuable to their teams than any alternatives at first base. A fair evaluation (which, to my mind, is what DanR and Crosbybird are aiming for), would use overall replacement level assessments to adjust DHs in relation to first basemen.

Since the DH exists, and since teams decided these players' value to the team would be maximized by having them assume the designated hitter role, why try to find some abstract penalty that distorts their actual value to their teams?

Chris - Yes, I do treat AL pitcher hitting in the same way as I do DH fielding. In some ways, I actually regard the two players - the pitcher and the DH - as occupying one roster slot. One guy gets all the run prevention credit and the other gets all the runs created benefit. This is, of course, a value evaluation rather than a rate evaluation. And first basemen proved to be such a big pain to deal with that I started looking for other ways to define absolute zero instead of "what would happen if you had no 1B out there at all." I realized that the result would be so catastrophic that teams would have to short another position to move someone to 1B. 1B doesn't have much defensive value because value is related to how close to you all your competitors are. Since 1B get 90-95% of their accepted chances on things that you or I could handle - easy, on-target throws from infielders, all 1B look pretty much alike until you get rid of those pesky catches. Bill James does some of this in the Win Shares system.

Pitchers are actually easier to evaluate than hitters, because their absolute zero can be at least estimated. You start with Win Share' margin for hitters (half the league average runs created) as your starting point, and then try to figure out how much value is contained in that "half the league runs created" in terms of final contribution. You take the margin, subtract the value of half the league runs scored, and you've got a zero point that is very close to absolute zero. The trouble with trying to do this to defense is that Bill's run prevention margin is 1.5 x the league ERA, while the upper bound is undefined, and there are actual pitchers who have had ERAs of infinity. The open-endedness is the problem. Neither Linear Weights or WAR systems solve the problem. They have different zero points from Bill's margin, but they are still open-ended in terms of defining absolute zero on defense. Unable to get a decent definition out of any of the systems I could find, I resorted to "at least a little worse than the worst season anyone actually played in the field for a whole year." I've done enough checking to find out that this doesn't result in DHs with negative values because their defensive lacks overcame their bats' gains. A guy playing out the string at DH will still have positive value. But he will have a lot less than he would have if he had been able to play the field (so to speak).

The main reason that I started thinking about DHs this way is that I noted that hitter careers are getting longer, as the better hitters are able to play for several years after they have become so bad on defense that no one will play them in the field any more. This is queering the numbers of Hall candidates coming up. It got started over in the Hall of Merit, when someone wanted to say that Dave Winfield had a long career, and Harry Heilmann did not. This is only true if you count the years where Winfield was a DH. Otherwise, Harry's career is just as long as Dave's. I realized that the DH rule was gong to produce more of these guys, and we were in serious danger of having even the Hall of Merit overrun with DHs, because they can accrue so much value in years when, in earlier times, they would not have been able to play at all. Buzz Arlett, for example, had only one MLB season, although everyone knew he could hit at MLB level, because he just could not play right field (which was, at the time, a less demanding defensive position than 1B). To treat Frank Thomas' DH years as if he were playing average defense would lead me to give MLE credit for Arlett for about 18 seasons of PCL play, because the lack of a DH rule was hardly a circumstance over which he had any control. I'd end up adding about 5 year's to Heilmann, to represent how long he could have played had there been a DH rule. And Arlett and Heilmann are only the tip of that iceberg. Babe Herman. Gavy Cravath. Dick Stuart. There are a LOT of big hitters whose careers were shorted or nonexistant because they could not DH.

In my mind, what I am trying to do is give fair credit to players from all eras and leagues. To do that, I either have to give Frank Thomas deductions for inability to play the field or I have to do a HUGE amount of minor league research to determining who could have been a career DH if he had only been playing now. Option #1 is the one I'm willing to deal with. Option #2 is a nightmare.

I hope that helps to explain not just what I'm doing, but WHY I'm doing it. Trying to be fair to all seasons and leagues. I'd much rather take a DH deduction for those guys who play DH than try to figure out how long Babe Herman's career would have been if he had not had to go out and risk his head in the outfield. - Brock Hanke

Oh! I have another question to ask, just for timing. When does the ballot thread start? I haven't seen it, and the discussion thread here started in December. My life will be much less complicated if I know when the ballot thread begins and when the final deadline is. Thanks in advance to whoever knows! -Brock

Yearly HoM ballots start around Thanksgiving. We have all year to discuss this.

From Mike Gimbel:

I just became aware of this discussion, so excuse my late entry.
I've been rating all players from 1900-2012 using a modified form of the RPA rating system. These ratings, after year and stadium adjustments, are based upon the value, in runs, plus or minus, over average pitcher or player that season and each year added together for the total career run value.

Here is the RPA list of the top pitchers in MLB since 1900. Note that the total value for Cy young includes only his years that started with the 1900 season. His prior seasons, therefore, are not added into his rating.

Here's the list with the career total run value above the average pitcher:

1. Roger Clemens 537.30 runs
2. Walter Johnson 468.58 runs
3. Greg Maddux 467.04 runs
4. Pete Alexander 395.01 runs
5. Randy Johnson 360.74 runs
6. Lefty Grove 356.10 runs
7. Christy Mathewson 335.27 runs
8. Pedro Martinez 334.42 runs
9. Tom Seaver 309.86 runs
10. John Smoltz 284.00 runs
11. Mike Mussina 275.97 runs
12. Curt Schilling 271.66 runs
13. Bert Blyleven 265.22 runs
14. Bob Gibsn 263.43 runs
15. Gaylord Perry 262.27 runs
16. Fregie Jenkins 251.73 Runs
17. Cy Young 246.97 Runs

From Mike Gimbel:

This is a follow-up to the previous comment, this time in regards to the career run values of the top hitters in MLB fom 1900-2012. It must be noted that defensive ability is not included, even though I have very good defensive ratings for recent MLB performers. By not including defensive performance this penalizes many excellent defensive performers such as Rickey Henderson, but this was unavoidable since there is no way to rate the defensive abilities of historic performers. The only fair way to compare is on the same scale. Therefore I had to ignore defense in these ratings:

1. Babe Ruth 1,326.16 runs
2. Barry Bonds 1,046.19 runs
3. Ted Williams 988.32 runs
4. Mickey Mantle 911.91 runs
5. Lou Gehrig 891.47 runs
6. Ty Cobb 818.17 runs
7. Mel Ott 813.14 runs
8. Rogers Hornsby 734.21 runs
9. Stan Musial 733.89 runs
10. Willie Mays 712.02 runs
11. Hank Aaron 688.03 runs
12. Frank Robinson 676.76 runs
13. Jimmie Foxx 670.49 runs
14. Tris Speaker 656.42 runs
15. Joe Morgan 622.04 runs
16. Eddie Mathews 621.33 runs
17. Frank Thomas 618.24 runs
18. Eddie Collins 578.53 runs
19. Manny Ramirez 561.38 runs
20. Rickey Henderson 555.06 runs
21. Mark McGwire 535.89 runs

DL - Thanks for the timeline note. If you look at my comments in the 2013 ballot thread (which I had to look up, you'll find that I was sick. That illness lasted three weeks, and cost me most of December. In the process, I actually forgot that I did vote in the 2013 election. When you said that there are months ahead of us on the 2014, I realized something was wrong. I thought that this was the thread that should have had its ballot up a couple of months ago. I'm straightened out,now. I just had forgotten that I voted in the 2013. I was sicker than I thought. - Brock

Mike (caiman) - That's certainly a plausible and defensible list. There's only one third baseman and no catchers, which sounds right, given what you're listing. Just to check, there's nothing like WWII credit or minor league credit or negro league credit or anything, is there? You say MLB 1900-2012, and I think you mean just that. I don't think that it would make a lot of difference anyway (Williams would pass Bonds, but not Ruth; Musial would pass Hornsby, some negro leaguers would show up on the list, Honus Wagner would probably make the list if you started with his rookie year instead of 1900), but I thought I'd ask, just in case. - Brock

Hi Brock,

How would I add these credits and remain accurate and defensible? The credits, such as WWII, would be nothing more than educated guesses. While I'd love to include the Negro League players, I can't, simply because there is no statistically defensible way to make an accurate comparison, as far as I'm aware.
Note: My excel files are massive and comprehensive. I'd love to get them up on a website. Any suggestions? --Mike

You can upload to the yahoo groups hall of merit site if you wish.

I can upload to the Yahoo site my large excel files? Please explain in more details. - Mike

Mike - I agree with you completely. I just wanted to make sure that your list really limits itself to MLB players after 1899. I was hoping you'd respond that there's nothing besides What you said was there. The lack of anything that requires MLs or war credit or whatever are good to have for evaluation, but just the raw list of hitters is a very good tool, and I intend to use it. What I didn't want was to get a response saying that the list had this bias or that bias. Now, I feel safe quoting from it. - Brock

Join the yahoo Hall of Merit group. Then it should be pretty easy to figure out how to upload.

Not sure where to put this, so I'll put it here. For those who missed it, FanGraphs posted a copy of Dan R's Sloan Analytics presentation: http://www.fangraphs.com/blogs/index.php/hitting-em-where-they-are/

I've been lurking since the early 1900s and I never had the urge to vote because you all were doing such a great job and I didn't think I would add any value. I've decided to give it a shot now, not because you're all doing any worse, but partly because the number of voters seems to be falling a bit and because I my curiosity about how I would do this got the better of me.

I'm a modified career voter, and use B-R WAA as my base metric. But I have a principle to make each candidate look as good as he can look, so I give credit for peak, postseason heroics, wars, unfair placement in minor leagues, Negro, Cuban, Japanese and other foreign leagues, and I zero out negative values in a given year. I don't penalize pitchers for having bad peripherals but I do give bonus points for excellent peripherals. I depend almost entirely on the work of Chris Cobb, Dr. Chaleeko et al for Negro League performance evaluation, but I do look at the real stats from seamheads.com and the player thread as a reality check. I do not penalize Negro Leaguers or anyone else for higher error bars regarding their career value. Just as you guys did, I started in 1898 and went year by year with all the candidates in the discussion threads, except that I did it over the past three months and used the latest B-R WAA figures. So I've run about 1400 players through my "system". I'm pretty happy with the balance I've gotten with Jake Beckley and Hughie Jennings scoring about equally well. I have a PHOM. I welcome any questions or critiques -- I may not change my ballot, but I promise to try to understand anything that anyone takes the time to say.

1) Greg Maddux -- #5 Pitcher, Top 25 player among All-time HOM Eligibles thru 2014 ("HOM Candidate List")
2) Curt Schilling -- Most postseason credit of any player on HOM Candidate List, and great peripherals (PHOM 2013)
3) Mike Mussina -- #15 Pitcher on HOM List
4) Tom Glavine -- Mediocre peripherals but great at stranding runners; #19 pitcher on HOM Candidate List between Steve Carlton and Phil Niekro; big gap between him and…..
5) Luke Easter -- Career obliterated by WWII, demise of Negro Leagues, post-Jackie racial discrimination; #9 1B on HOM Candidate List (PHOM 1960)
6) Orel Hershiser -- #2 postseason credit of any player on HOM Candidate List (PHOM 2006); as an A's fan….ugh
7) Frank Thomas -- Great hitter, terrible fielder; #13 1B on HOM Candidate List between Buck Leonard and Dick Allen. I fully penalized him for his terrible fielding in part because he didn't like DHing and hit worse when he did
8) Luis Tiant -- #42 Pitcher on HOM Candidate List between Ed Walsh and Rube Waddell; 119 ERA+ in 3000 IP 1964-1978 (PHOM 1988)
9) Artie Wilson -- Career value ruined Easter-style, somewhere in the range of Bobby Wallace and Joe Cronin with credit given for WWII and minor leagues (PHOM 1964)
10) Leroy Matlock -- Somewhere between David Cone and Dwight Gooden in career value (PHOM 1965)
11) Dwight Gooden -- Bonus points for dominant peripherals get him on the ballot (PHOM 2006)
12) Kenny Lofton -- 19th CF on HOM Candidate List; between Cool Papa Bell and Richie Ashburn among comparable players
13) Sammy Sosa -- 14th RF on HOM Candidate List between Tony Gwynn and Enos Slaughter; career shape sort of like Dwight Evans but higher peak
14) Kevin Appier -- 140 ERA+ in 1600 IP 1990-1997; overall value similar to Stan Coveleski (PHOM 2009)
15) Bus Clarkson -- Another 1940s Negro whose numbers were beat up by WWII and discriminatory desegregation; bit worse career value than Dick Lundy (PHOM 1968)

Required Comments:

21) Ben Taylor -- Similar to Jake Beckley who is also in PHOM, but the competition is much tougher now (PHOM 1936)
44) Dick Redding -- WWI credit; excellent but not a high enough peak 1915-1919 and not enough outside of that; last made PHOM "ballot" in 2012
51) Vic Willis -- Behind Young, Mathewson, Walsh, Brown, and Plank among contemporaries; doesn't get peripheral credit like Waddell, and below Rube Foster
79) Phil Rizzuto -- Even with WWII credit behind Art Fletcher and Joe Tinker among non-HOM SS, just above Johnny Pesky

deleted

deleted

Glad you decided to take the plunge, Ivan. I like the internal consistency within your ballot (for example, it's hard to like one of Hershiser or Gooden without liking the other). And I'm glad to see that you gave due consideration to Negro League players. One question though: He's not a required disclosure yet but newcomer Jeff Kent figures to place on quite a number of ballots. How do you rank him?

#73 -- Chris, I have Kent #90...he seems pretty similar in overall value to Chuck Knoblauch, Tony Perez and Robin Ventura who are in that same vicinity. They're all a bit above Jim Fregosi among good hitting infielders.

Ivan, glad you're in. Looking at your picks, I was wondering, where do you rate Hilton Smith and Marvin Williams? I would think they would be comparable to Wilson and Matlock and Clarkson.

#75 - theorioleway, I have Smith at #41, with a dominant peak but only 4 really good years in his career; I think I need to look at Williams again. I seem to have been underrating him. He is likely to be on my ballot someplace, maybe in the 5) thru 10) range which would displace Clarkson. Thanks for pointing him out to me.

Welcome, Ivan, with your prelim...

Whatever your opinions of Frank Thomas he was still considered a better fielder than Luke Easter. That's what helps to keep Easter off my ballot, dreadful defense. He was a born DH.

It also takes some pretty generous MLEs to turn Artie Wilson into Joe Cronin.

#77 -- Thanks Howie

#78 -- DL, can you provide some numbers on Easter's defense other than what's in his thread? Also which MLEs are you using for Artie Wilson? I'm happy to reconsider both of these guys if additional information is available that I haven't yet seen.

DL -

As Luke Easter's now second best friend, I don't know where this opinion of Easter's defense is coming from. Most of the WAR metrics for his ML years have him as at worst, an average defensive 1b. And this is when he is in his mid to late 30's with two bad knees.

I know the new BP metrics are not a big fan of his D, but B-R, FG and old Clay Davenport numbers all have him as above average for his ML years.

It's anectdotal but Luke Easter was described as a "statue" for his AAA career. I think he may have been fine when younger but the known injuries caught up with him. I like Easter but even with good offensive numbers and proper credit I'd have a very hard time ranking Frank Thomas below him just based on defensive ratings.

I don't have recent MLE's for Artie Wilson but you'd have to basically give him a 100% translation from the PCL to the majors to make him Joe Cronin.

3) Conversion rate: .90 for 1944-1946, 1947-1948 I went with .88, AAA level. The PCL was an open-classified league at this time, with lots of NgLers of MLB caliber, so I used .90 for it. If the group thinks the PCL of that era was as strong as Japan today, then I'll ratchet up to .92.

This is cut and pasted from the Artie Wilson thread. I re-checked the numbers and it looks like Cronin had negative WAA in 1927, 1928, 1936, and 1944. His WS for the remaining years were 317 in 7990 PA. Dr. Chaleeko puts Wilson at 311 WS in 8135 PA. Yes, Cronin's better and had he been eligible in 2014 I would have had him higher than Wilson on my ballot, but the difference isn't very big. I guess one could argue that Cronin had a 119 OPS+ for his career and in Dr. Chaleeko's MLEs Wilson has a 94, and yet their BWS aren't that far apart so there might be something strange there, but I didn't re-calculate BWS partly because I don't think I could do it any better than he did, and partly because it looks like Wilson's OPS+ numbers were dragged down by three years where the OBP+ was fairly OK -- 1950 (OBP+ 90, OPS+ 69), 1951 (OBP+ 90, OPS+ 68), 1956 (OPB+ 99, OPS+ 81). It also looks like Cronin's OPS+ was somewhat SLG-driven while Wilson's was very OBP-driven. Still, it could well be that Dr. Chaleeko's WS estimates are not directly comparable with seamheads's WS estimates. I'll look into it again. Thanks for pointing it out.

Regarding Thomas, I didn't mean that I demoted him because his fielding was bad. I just meant that I used BR-WAA's numbers as is and didn't cap his negative fielding runs at a DH level in part because he didn't like DHing and he hit worse as a DH.

Ivan - This is a perfectly defensible ballot. I don't know what exactly you did where, but I see the possibility of comparing Luke Easter as an old man (which includes AAA) to Frank Thomas' entire career. Overall, I think Easter was probably the better defender (Thomas is the worst glove I've seen since Dick Stuart, and sigh, yes, I'm old enough to have seen Dick Stuart). But avoiding comparing them at different ages would involve comparing Easter's early defense to Thomas' early defense, and that means Negro League translations. I, personally, tend to be a little conservative about Negro League rankings, and have no shortstops except Pop Lloyd and Devil Wells. I know nothing about Artie Wilson. My current hobbyhorse is comparing Sammy Sosa to Bobby Bonds. They end up, by my methods, as almost exactly the same in value. Bonds is a little ahead in career, Sosa in peak and prime. The only systematic complaint that I might have is that there are no really early guys on your ballot. Granted, we've had years to pick them clean, but I think there are a few gems left to discover.

Brock - My highest "early" players were:

18) Babe Adams (PHOM 1933) -- Postseason credit and peripherals help
25) Chief Bender (PHOM 1927) -- Postseason credit helps
26) Wally Schang (PHOM 1976) -- Catcher bonus
29) Dizzy Dean (PHOM 1985) -- Postseason credit and peripherals
30) Ned Williamson (PHOM 1930) -- New BR WAA helps, and I have him above Ezra Sutton but both are PHOM
31) Fred Dunlap (PHOM 1931) -- Also helped by new BR WAA and I have him above Hardy Richardson who just misses PHOM

Regarding Luke Easter, I basically used Dr. Chaleeko's WS numbers shown in his thread, and looked at James Newburg's and Alex King's numbers as a check. I made no other adjustments for defense since I have no outside knowledge of Easter's defense.

As for Bobby Bonds:
20) Bobby Bonds (PHOM 1987) -- Not that different from Sosa, but Sosa has some postseason credit and peak bonus points

Ivan - Hah! So there are TWO of us who think highly of Babe Adams, although you don't have him on your ballot. I do. Please feel free to look at my ballot this year, or in prior years if you want, and make comments. There is, all things considered, a reasonable amount of support for the six guys who passed through Pittsburgh in the early part of the 20th century - Phillippe, Leever, Tannehill, Willis, Cooper, and Adams. What there is NOT is any agreement on which one of these six to focus on. My favorites are Adams and Phillippe, because I give credit for being control outliers. There are Willis supporters (he IS in the HoF), and, I think, one guy who likes Cooper and one who likes Tannehill and one who used to like Leever. If we could all decide on one to attract our focus, he'd probably get elected. Just to check, you do massively adjust Dunlap's superdominant season in the Union Association, don't you? - Brock

85 - Brock, I adjusted the Union League 40%, which is the biggest adjustment I made to any league/year.

Regarding the other Pirate hurlers, none was particularly close to the ballot nor in my PHOM:

52) Vic Willis
64) Wilbur Cooper
119) Jesse Tannehill
322) Deacon Phillippe
342) Sam Leever

I confess I didn't look that closely at any of them other than running their numbers through the system.

Ivan - Interesting. I would adjust the Union Association down TO no more than 40% of actual MLB, and I'm not sure whether you were saying that or that you adjusted the UA to 60% of MLB. And I will agree that there is no case for Deacon Phillippe that doesn't involve giving credit for being the best control pitcher of the 20th century. But I do like my outliers. I vote for Dizzy Dean, partially because of the postseason, but also partially because it was very clear, as with Koufax, that contemporary observers, including players, really did think that Dizzy was something special.

If you really want to hurt your head, look up the story of the 1903 World Series, and try to figure out whether to give Phillippe credit for it. I don't, because I'm already giving him credit for control, but he does have a story. Essentially Boston showed up with Cy Young and Bill Dineen, Cy pitching games 1,3,5,7,9, and Bill the even-numbered ones. It LOOKS, if you just scan the season rosters, like the Pirates have a large advantage in pitching depth, but it's an illusion. Ed Doheny had, literally, gone insane and was in a sanitarium. Everybody else except the Deacon was hurt or his arm had gone dead from overwork (Sam Leever, the #2 starter on the staff, was a curve ball specialist, and his arm was useless in the WS). Well, manager Fred Clarke started Deacon against Cy - and Deacon beat him. After Leever proved that he was gong to be useless in Game 2, the Deacon came up against Cy again, and beat him again. At this point, Clarke blinked. Instead of just hoping that either 1) Deacon really could win 5 in a row from Cy, or 2) that the Pirates would have an offensive explosion and win a barnburner against Dineen, he decided that it was 1883, not 1903, and that Deacon could just pitch every game, and beat both Cy and Bill. Deacon did beat Dineen in Game 4, but that was it. He ended up 3-2, because the start against Dineen had done for his arm. Do you give credit for that? This is the kind of question that I find most fascinating in the HoM. The hard ones. Where there are points to be scored on both sides, and not a whole lot of hard evidence. Anyway, thanks for the very useful comment! - Brock

1903 WS:

There are a lot of people who believe that the Pilgrims threw Game 1. Pittsburgh scored four runs in the first, aided by three Boston errors, at least two of which were more than a little suspicious. There are also people who believe that the Pirates may have thrown Game 8, because by that time interest in the series had waned and it was pretty clear that a Game 9, with Phillippe likely unavailable against Young, wouldn't draw much of anything - Game 8 attendance, FWIW, was nearly 10,000 less than Game 7. Tommy Leach was picked off third in the fourth inning, when the Pirates were still in a scoreless tie, and Phillippe was picked off first in the sixth when the score was just 2-0.

Leever went the distance in Game 6 and didn't pitch all that badly. Young and Dineen had thrown more innings during the season than had Leever and Phillipe (indeed, Clarke was one of the first managers not to ride the front end of his staff all that hard), so I'm not sure that the "tired arm" excuse is entirely valid - Phillippe pitched poorly only in Game 7, and part of that was poor defense.

-- MWE

Mike - I'd never heard anything about shenanigans in the 1903 WS, so thanks for the new info. My read on Clarke and pitcher workloads is that Fred, realizing the kind of infield he had, specialized in acquiring and training curve ball pitchers. Deacon had a good fastball as well as a good curve, but Leever, and I think the rest of the staff, were curve ball specialists. This,of course, meant that they could handle lesser loads than the fastball guys like Young. I talk about the "Pittsburgh Six": Phillippe, Leever, Tannehill, Willis, Adams, and Cooper. One of the reasons that I group them is that they all had superior curve balls, and relied on them more heavily than other pitchers. In general, I give Clarke manager points for figuring this out, but it did, I think, cost him at the end of the season and in the postseason, when all the curves had worn the arms down. It makes sense that Phillippe, who was the least-curve-dependent pitcher on the 1903 staff, would have the best arm remaining for the WS. It also makes sense that Leever, taken out quickly in Game 2 and then rested for several days, would have a good game later in the Series. But I still think that Clarke's best option, at the end of Game Three, was to just pitch the Deacon in rotation and hope that he could either win another three from Cy or that the offense would win a blowout against Dineen. According to Honus Wagner's biography, Honus blamed himself for the WS loss, since he had not hit well. When Deacon won Game 4 against Dineen, the Pitttsburgh fans apparently had a big parade in town, feeling that their new hero would certainly clean up the BoSox in short order. I also think that Clarke might have made a different decision about Deacon if the WS had been 7 games instead of 9, so Deacon would only have to win 2 more against Young, not 3. - Brock.

Bump - we're getting closer.

I'm working on converting my rankings of major-league players to be based on the second- and third-generation WAR stats now available. Once I finish that conversion, I plan to start re-doing MLEs for Negro League players based on the improved statistical record. I'll start with Ben Taylor and Dick Redding

Chris - how far did you get? Those two players are required disclosures this year.

Yes we are getting closer, I guess it is finally time to start thinking about scheduling the election. Any thoughts? It would be good to not have it throw off the MMP elections, i.e. we can try to squeeze the HoM during an obvious break point for the MMP, unless you want to run those in parallel, I can't remember how we did it in the past.

We can run in parallel and offset by a couple weeks. 1988 MMP will be November. 2013 MMP will be December. If we close out the HoM ballots before Christmas we'll be fine.

Another way I phrase this to myself is this: We start out with absolute zero as a player who has never played or at least has never accepted a chance. When he does play, and accepts chances successfully, then we can attribute the value of those accepted chances to the player. Anyone who actually plays a position will accumulate some defensive value over that starting point zero, because, no matter how bad he is, he will accept some chances successfully. You could put me out there at shortstop, and eventually someone would hit an easy grounder right at me and I'd successfully accept the chance. But a DH won't. He just stays there, at absolute zero.

This is more or less my view exactly.

Hypothetical case: a Tony La Russa type wants to save his bullpen by using Position Player John Smith to toss meatballs on the mound every time the other team gets an 8 run lead or more. He ends up pitching around 60 innings a season with an ERA of 28.30, for 20 years.

Such a player is going to come out looking awful when compared to his position player peers. And yet that doesn't make any sense. Nor does it make sense to me to 'correct' for this fact by subtracting WAR from every other position player 'as if' they pitched every season. Especially since we have no idea how good or bad a pitcher they would be. Rather, it makes more sense to give John Smith some (albeit very small) credit for contributing to his team in a way other players didn't.

So: the acid test. Two players, John Smith and Bob Roberts. Both with identical (adjusted) hitting, fielding, and baserunning stats, except Smith pitches in blowouts as above. Who do you put first on your ballot? I would put Smith, for the reasons above.

I would say I can't tell the difference. John Smith had a stupid manager.

Brock/90:

There were rumors of thrown games in just about every Series between 1903 and 1919.

Phillippe/Leever: By 1906, when Willis came over from Boston, Clarke was starting to spot both of them. 1905 was Phillippe's last big season, Leever's was 1906. Even when they were at the top of their games from 1903-1906, neither was pitching anywhere near the innings that other top pitchers of the era were throwing. Clarke more or less routinely gave them three-four days off between starts, running guys like Patsy Flaherty, Mike Lynch, and Charlie Case out. I don't know whether it's because they couldn't handle the load or because Clarke thought they were more effective when pitching less often; he certainly didn't do that with Willis, who was the ace of the staff from the moment he arrived in Pittsburgh to the moment he departed.

-- MWE

Mike - Thanks for the info. I've paid almost no attention to the throwing of games, because I don't know anything beyond the Black Sox. I'll file it in my head that some of those series may not have been completely clean. My take on the Pirate pitching thing would, if I tried to explain it in detail, require me to discuss the history of the pitching rotation starting in 1879. That would be boring, so here are the biggest highlights to me: 1) By 1883 or so, it was becoming clear that even the best pitchers could not pitch 500 innings in a season without it compromising their next season. 2) It looks like Charlie Comiskey was the first manager to completely convert to the two-man rotation, which explains a lot of Bob Caruthers', Dave Foutz's and Silver King's careers. 3) When 1893 brought in the 60-foot pitching distance, IP plummeted and rotations expanded because it was suddenly harder to pitch. 4) By the turn of the century, you can argue that some managers were overcompensating, going with four and even five-man rotations when they could have gotten away with three. 5) With less strain on arms due to low workloads, curves and spitballs, which both apparently hurt arms more than fastballs, came into vogue, because no one was going to pitch huge workloads except for a few super-arm fastball strikeout pitchers like Johnson and Young. 6) Two teams - the Pirates and the Cubs - got on the low-workload / lots of curve balls train a bit before the others, with the result that they dominated the NL in the first decade of the 1900s.

The thing that started me on this line of thinking was when I found out that Bob Gibson does NOT hold the 20th-century record for lowest ERA in a full season. Three-Finger Brown holds that record, if stated exactly as above. The record that Gibson holds is the lowest ERA in a season of OVER 300 INNINGS. Brown, 60 years earlier, was pitching fewer innings, while the general progress of baseball is for the IP of starters to go down. I started looking at pitchers of the 1900-1910 period, and that's what led me to all this analysis.

On the other hand, just because you wrote the above, I'm going to take another look at Vic Willis. It would make a difference if he was pitching large workloads. I know he pitched a lot of seasons. - Brock

It's hard to characterize Clarke's pitching usage. I suppose I should go upstairs and grab Chris Jaffe's book.

In 1904, no Pirates starter threw as many as 20% of the team's innings, which was virtually unheard of in that era. In 1905, Phillippe was the only starter to crack 20%. When Willis came over in 1906, he had percentages of 23.7%, 21.5%, 21.7% and 20.7% - still fairly low for the era but quite a bit more than the 1904-1905 teams. More importantly, the allocation of innings behind Willis shuffled regularly. In 1906 the #2 guy was Leever, in 1907 Leifield, in 1908 Maddox, in 1909 Camnitz. (Maddox would never be consistently effective again after that 1908 season.)

-- MWE

The Cubs did some of the same things, with pitching usage spread around fairly well and some pitchers being very effective but just for one or two seasons. Chance did have Mordecai Brown carry somewhat more of the load, but I still think of the Cubs and Pirates as being on the same or at least related plans.

I have a tough time rating Schilling vs. Mussina vs. Glavine. It's not that important for a HOF vote, binary as that is, where I don't have to hesitate before saying yes to all 3.

Looking at career WAR, it's 80-83-81. BBref also has WAR7, which is not quite the same as peak as the 7 best years do not have to be consecutive. By that one, it's 49-44-44. There just isn't enough difference in those numbers to clearly place one ahead of the others. Schilling has the best rate stats and pitched the least. Glavine has the worst rate stats and pitched the most. Mussina's the guy in the middle.