THE BOOK cover
The Unwritten Book is Finally Written!
An in-depth analysis of: The sacrifice bunt, batter/pitcher matchups, the intentional base on balls, optimizing a batting lineup, hot and cold streaks, clutch performance, platooning strategies, and much more.
Read Excerpts & Customer Reviews

Buy The Book from Amazon


SABR101 required reading if you enter this site. Check out the Sabermetric Wiki. And interesting baseball books.
MOST RECENT ARTICLES
MAIL : You ask | We say

Advanced


THE BOOK--Playing The Percentages In Baseball

<< Back to main

Sunday, January 01, 2012

Rebooted Hall Of Fame

By Tangotiger, 12:42 AM

If you follow sci-fi or comic book movies, you know what I’m talking about here.  Sometimes, the weight of history is too much for the mythology to survive into the next generation.  Bad decisions don’t need to be propogated or respected. Do we need to follow Batman’s Clooney, Schwarzenengger, and Schumacher, or can we just start over with the Frank Miller inspired version, with Christian Bale and Christopher Nolan?

You get a reboot. 

My rules for the Hall of Fame is that it’s based on birth year, when the player turns 45 (and if he’s still active, then so be it… worked fine for Gordie Howe, Mario Lemieux and Guy Lafleur… an active-playing Hall of Famer), and he remains on the ballot for at most ten years.  Each year, you have ten new players nominated (that is, the ten best players of the new birth year), and the top ten players of the previous ballot carrying over (so we always have twenty players on the ballot, except for the inaugural year).  Every ballot is an ordered ballot, and must have ten names.  The top two each ballot gets elected the Hall Of Fame.  It’s an elect-2 system.

There’s going to be two groups, one that starts with birth year 1931, and then we’ll have another one for pre-1931.  Why 1931?  Well, it’s the first year where you have an over-abundance of qualified players, following a dearth of under-qualified players.  (It also gives us some breathing room with the Negro League stars.) I’ve studied it, and it’s going to work.  I’ll show it next time.

Anyway, we start with imagining this is the end of 1976, and players born in 1931, and here are the ten players on the 1977 ballot:


Willie Mays
Mickey Mantle
Eddie Mathews
Ernie Banks
Ken Boyer *
Ed Bailey * +
Joe Cunningham *
Bill Virdon

Jim Bunning *
Larry Jackson * + - /

This becomes an easy selection. I don’t think we need to do any kind of mock balloting, or trying to use past ballots to try to simulate this ballot.  (It might be fun in other years, and maybe we’ll do that, but not here.)

So, our first two Hall of Famers are Willie and Mickey.

That’s how you reboot something.

(Don’t pay attention to the symbols after their names.  It’ll be explained in a few minutes.)

Now we go to players born in 1932.  It’s a pretty weak year:
Maury Wills * + - /
Woodie Held * +
Eddie Broussard *
Don Blasimgame
Jim King
Eddie Kasko

Johnny Podres * +
Ron Kline *
Billy Hoeft
Hal Woodeshick

The next two members are carry-overs: Eddie Mathews and Ernie Banks.

Only ten can carry-over onto the next ballot, and they’ve been marked by *.

Now it’s on to 1933:
Rocky Colavito + - /
Norm Sieburn +
John Roseboro +
Ed Charles
Jim Davenport
Jerry Lumpe
Tito Francona

Billy O’Dell +
Ken Johnson +
Bob Shaw

The next two members in the rebooted Hall of Fame: Jim Bunning, Ken Boyer.

The ten carryovers have been marked with a +.

Now it’s 1934 birth year players, and we’ve got a great set of players:
Hank Aaron
Al Kaline
Roberto Clemente - /
Norm Cash - /
Luis Aparicio - /
Roger Maris - /
Bob Allison - /
Bill White -

Camilio Pascual -
Turk Farrell

Our two new members are Hank Aaron and Al Kaline. 

The carryovers are marked with -.

It’s the 1935 birth year players, and we’ve got a whole mess of pitchers to contend with:
Frank Robinson
Felipe Alou /
Earl Battey

Bob Gibson
Sandy Koufax /
Jim Perry
Bob Veale
Lindy McDaniel
Pedro Ramos
Don Cardwell

Our two new members are Frank Robinson and Bob Gibson.  Hard to believe, but Roberto Clemente has to carryover yet another year.

The carryovers are marked with /.

We’ll do one more, then we’ll stop for now.  The 1936 birth year players on the ballot:
Harmon Killebrew
Frank Howard
Bill Mazeroski
Tony Gonzalez
Floyd Robinson
Clay Dalrymple

Don Drysdale
Stan Williams
Bill Monbouquette
Ron Perranonski

Finally, we get to clear out some carryovers: Roberto Clemente, Sandy Koufax.

After six elections, we have these as our twelve players in the rebooted Hall of Fame:
1977: Willie Mays and Mickey Mantle.
1978: Eddie Mathews and Ernie Banks.
1979: Jim Bunning, Ken Boyer.
1980: Hank Aaron and Al Kaline. 
1981: Frank Robinson and Bob Gibson.
1982: Roberto Clemente, Sandy Koufax.

And our ten players that carryover to the next ballot:
Harmon Killebrew (b. 1936)
Norm Cash (b. 1934)
Luis Aparicis (b. 1934)
Rocky Colavito (b. 1933)
Roger Maris (b. 1934)
Felipe Alou (b. 1935)
Frank Howard (b. 1936)
Maury Wills (b. 1932)

Don Drysdale (b. 1936)
Larry Jackson (b. 1931)

Next time, I’ll do players born 1937-1942.

#1    Mike G.      (see all posts) 2012/01/01 (Sun) @ 22:33

I played with a HOF reboot last year, so this prompted me to formalize it with a program.

Methodology:
Using Lahman database (through 2010) plus rWAR from BBRef. Calculated a seasonal weighted WAR based on:

0.002x^3 + 0.13x^2 + 0.695x
(fitted onto double for 3-5.9 WAR, triple for 6-8.9 WAR, etc.)

Eligible at age 45, for 10 years. Selected by sum of seasonal weighted WARs. All players carryover.

Shortcomings: Dependent solely on this one number. Doesn’t account for missing time (WWII, pre-integration). Likely sells relievers short, minimally b/c they can’t get high season WAR levels.

Now, the results:
1977: Willie Mays, Mickey Mantle
1978: Eddie Mathews, Ernie Banks
1979: Jim Bunning, Ken Boyer
1980: Hank Aaron, Al Kaline
1981: Frank Robinson, Bob Gibson
1982: Roberto Clemente, Don Drysdale
1983: Juan Marichal, Brooks Robinson
1984: Gaylord Perry, Willie McCovey
1985: Phil Niekro, Carl Yastrzemski
1986: Ron Santo, Sandy Koufax
1987: Pete Rose, Harmon Killebrew
1988: Fergie Jenkins, Dick Allen
1989: Joe Morgan, Jimmy Wynn
1990: Tom Seaver, Steve Carlton
1991: Rod Carew, Jim Palmer
1992: Reggie Jackson, Sal Bando
1993: Nolan Ryan, Johnny Bench
1994: Carlton Fisk, Reggie Smith
1995: Mike Schmidt, Bobby Grich
1996: Rick Reuschel, Don Sutton
1997: Bert Blyleven, Graig Nettles
1998: Buddy Bell, Bobby Bonds
1999: George Brett, Keith Hernandez
2000: Gary Carter, Ozzie Smith
2001: Robin Yount, Dave Winfield
2002: Paul Molitor, Eddie Murray
2003: Lou Whitaker, Dwight Evans
2004: Rickey Henderson, Wade Boggs
2005: Alan Trammell, Ryne Sandberg
2006: Cal Ripken, Tony Gwynn
2007: Tim Raines, Andre Dawson
2008: Roger Clemens, Dave Stieb
2009: Randy Johnson, Edgar Martinez
2010: Barry Bonds, Barry Larkin
2011: Craig Biggio, Kevin Brown
2012: Greg Maddux, Curt Schilling
2013: Larry Walker, Kenny Lofton
2014: Jeff Bagwell, Frank Thomas
2015: Ken Griffey, Mike Mussina
2016: Jim Edmonds, Jim Thome
2017: Pedro Martinez, Sammy Sosa
2018: Chipper Jones, Manny Ramirez
2019: Roberto Alomar, Todd Helton
2020: Derek Jeter, Mike Piazza
[These last few years still have a few active players competing, but seemed pretty likely.]

The only difference in the first 6 years was picking Drysdale over Koufax. Regular rWAR has Drysdale 69.8 to 48.7. Weighted rWAR was 102.6 to 96.1.

I ran it starting with pre-1856 birthdays, which is when enough HOF-level players were around for 2 per year and has the happy coincidence of the first induction year of 1901.

“Worst” players to make it (rWAR/wrWAR):
- Roy Campanella (36.2/52.2)
- Al Rosen (33.0/54.2)
- Vern Stephens (43.5/56.3)
- Jack Quinn (48.2/58.1)
- Gil Hodges (44.6/59.3)
- Nellie Fox (44.4/60.6)

“Best” players dropped (rWAR/wrWAR):
- Mark McGwire (63.1/91.0)
- Tom Glavine (71.6/89.6)
- Rafael Palmeiro (66.0/88.5)
- Luis Tiant (59.4/83.7)
- John Smoltz (65.3/81.7)
- John Olerud (56.8/80.9)
[Yes, Glavine, Smoltz, and Olerud are dropped after 2020, but there are already enough players ahead of them to keep them out.]

Best players that have to wait till next year:
- 1934 Pete Alexander(106.8/179.8) - 1933: W Johnson and E Collins (plus Speaker in 1934).
- 1978 Eddie Mathews(98.3/165.8) - 1977: Mays and Mantle
- 1982 Roberto Clemente(83.8/129.1) - 1980: Aaron and Kaline; 1981: F Robinson and Gibson
- 1978 Ernie Banks(64.4/110.5) - 1977 (what a year!)
- 1950 Paul Waner(73.8/107.3) - 1949: Gehrig and Charlie Gehringer
- 2015 Mike Mussina(74.6/104.0) - 2014: Bagwell and Thomas; 2015: Griffey, Jr.

Players elected in their final year (or two):
- Will White (1909)
- Jimmy Collins (1924 - year #9)
- Tony Lazzeri (1958)
- Dizzy Trout (1970)
- Vern Stephens (1975)
- Roy Campanella (1976)


#2    Tangotiger      (see all posts) 2012/01/01 (Sun) @ 23:00

Mike: fantastic, thanks for doing all my work for me!

One thing I was thinking is that every 3 years (after SOME point, not sure which), there’d be an elect-3.

Looking at your list, there’s a clumping of the 4 best guys (born between 1963-1967).  So, this would have been anticipated.

Perhaps around 2012 or so, we’d start with an elect-3 every 3 years.

Because we see here that it would be absurd to have Kevin Brown in, but not Glavine and Smoltz.

I’ll tell you my rules for pre-1931 later on, and maybe you can program that in as well.


#3    Tangotiger      (see all posts) 2012/01/01 (Sun) @ 23:04

By the way, can you also show for those guys not elected on the 1st ballot, how long they were on the ballot.

Interested to see how long a player was on the ballot before he got in, and whether this wait time has been more in the more recent years (hence a reason for the extra elect-3 seasons, because too much wait time).

Again, great job!


#4    Mike G.      (see all posts) 2012/01/02 (Mon) @ 09:35

I noticed the same thing about needing to expand the election, and was trying to come up with a rule for determining the size. My first thoughts were either based on number of teams existing, or number of players accumulating 10+ seasons in the majors.

Here’s the full list, complete with elect-3 every 3 years starting in 2012, and the wait time for each.

1901: [2]Charley Radbourn (73.6/178.2), [4]Cap Anson (88.9/124.7)
1902: [1]Pud Galvin (58.8/147.3), [1]Jim McCormick (63.7/129.0)
1903: [1]Tim Keefe (79.0/170.3), [1]Roger Connor (87.2/136.8)
1904: [1]Dan Brouthers (81.7/134.9), [3]Tommy Bond (40.3/96.1)
1905: [1]Tony Mullane (69.6/115.4), [1]Mickey Welch (55.4/105.1)
1906: [5]Guy Hecker (39.3/86.1), [1]John Ward (64.9/83.4)
1907: [1]John Clarkson (81.3/191.3), [1]Charlie Buffinton (57.4/128.6)
1908: [1]Ed Morris (37.6/98.9), [6]Jack Glasscock (58.2/82.2)
1909: [8]Dave Foutz (48.1/76.5), [10]Will White (32.5/73.4)
1910: [1]Bob Caruthers (71.4/124.6), [8]Jim Whitney (45.0/70.9)
1911: [7]Bid McPhee (57.9/70.7), [7]Buck Ewing (51.4/65.6)
1912: [1]Billy Hamilton (69.6/107.7), [1]Gus Weyhing (34.5/72.6)
1913: [1]Cy Young (143.2/279.7), [1]Ed Delahanty (74.7/128.5)
1914: [1]Silver King (53.1/136.6), [1]Jesse Burkett (64.1/92.2)
1915: [1]Kid Nichols (103.2/201.2), [1]Hughie Jennings (48.0/87.0)
1916: [1]George Davis (90.3/135.9), [1]Bill Dahlen (75.9/104.9)
1917: [1]Amos Rusie (62.1/124.0), [1]Joe McGinnity (49.5/91.6)
1918: [1]Fred Clarke (73.4/101.5), [1]Willie Keeler (60.8/86.0)
1919: [1]Bobby Wallace (65.9/89.7), [3]Joe Kelley (55.5/80.1)
1920: [1]Honus Wagner (134.5/254.9), [1]Nap Lajoie (104.2/177.3)
1921: [1]Eddie Plank (76.8/107.4), [8]Jack Stivetts (53.0/77.7)
1922: [1]Elmer Flick (56.7/86.7), [1]Vic Willis (50.4/85.2)
1923: [2]Mordecai Brown (55.6/83.3), [2]Rube Waddell (46.4/77.3)
1924: [6]John McGraw (49.3/77.1), [9]Jimmy Collins (53.0/75.8)
1925: [4]Frank Chance (49.5/70.1), [2]Jimmy Sheckard (51.8/67.7)
1926: [1]Christy Mathewson (90.7/163.5), [1]Sam Crawford (76.6/107.4)
1927: [1]Ed Walsh (56.8/102.3), [5]Tommy Leach (50.9/65.0)
1928: [1]Babe Adams (46.3/65.0), [4]Noodles Hahn (38.3/64.4)
1929: [3]Johnny Evers (48.4/62.6), [4]Joe Tinker (49.2/62.6)
1930: [1]Sherry Magee (59.1/81.2), [1]Eddie Cicotte (49.6/76.0)
1931: [2]Nap Rucker (39.9/63.2), [3]Jack Quinn (48.2/58.1)
1932: [1]Ty Cobb (159.5/306.1), [1]Frank Baker (63.7/105.9)
1933: [1]Walter Johnson (139.8/250.0), [1]Eddie Collins (126.7/230.8)
1934: [1]Tris Speaker (132.9/240.0), [2]Pete Alexander (106.8/179.8)
1935: [1]Joe Jackson (62.9/109.4), [1]Stan Coveleski (48.5/78.9)
1936: [3]Red Faber (51.3/73.9), [3]Zack Wheat (57.8/73.3)
1937: [1]Dazzy Vance (53.8/88.6), [2]Urban Shocker (50.2/66.2)
1938: [2]Dave Bancroft (46.4/65.5), [6]Harry Hooper (52.6/64.2)
1939: [1]George Sisler (52.5/84.0), [5]George Burns (45.1/64.0)
1940: [1]Harry Heilmann (69.4/105.4), [5]Max Carey (50.6/61.9)
1941: [1]Babe Ruth (190.0/425.2), [7]Heinie Groh (46.4/61.8)
1942: [1]Rogers Hornsby (127.8/266.1), [6]Eppa Rixey (48.4/61.6)
1943: [6]Wilbur Cooper (47.5/61.5), [8]Sam Rice (51.2/61.1)
1944: [1]Frankie Frisch (74.8/114.2), [1]Bill Terry (55.4/83.2)
1945: [2]Kiki Cuyler (49.6/68.4), [2]Joe Sewell (48.4/63.1)
1946: [1]Lefty Grove (93.3/168.9), [1]Goose Goslin (63.0/88.0)
1947: [2]Ted Lyons (62.9/78.5), [3]Earle Combs (44.7/61.0)
1948: [1]Al Simmons (63.6/93.5), [1]Earl Averill (45.0/62.3)
1949: [1]Lou Gehrig (118.4/242.6), [1]Charlie Gehringer (80.9/131.8)
1950: [2]Paul Waner (73.8/107.3), [2]Carl Hubbell (63.6/97.5)
1951: [1]Red Ruffing (67.3/79.9), [1]Bob Johnson (53.2/70.0)
1952: [1]Joe Cronin (62.5/92.0), [4]Mickey Cochrane (51.2/69.3)
1953: [1]Jimmie Foxx (95.2/164.9), [1]Luke Appling (69.3/99.1)
1954: [1]Wes Ferrell (53.3/73.0), [2]Bill Dickey (54.4/69.8)
1955: [1]Mel Ott (109.3/180.2), [1]Billy Herman (55.6/75.2)
1956: [2]Stan Hack (54.8/72.7), [5]Tommy Bridges (49.6/65.8)
1957: [1]Hank Greenberg (56.8/91.1), [1]Joe Medwick (55.8/81.6)
1958: [1]Arky Vaughan (75.6/123.5), [10]Tony Lazzeri (48.3/64.1)
1959: [1]Johnny Mize (70.2/110.5), [7]Dolph Camilli (43.0/63.1)
1960: [1]Joe DiMaggio (83.6/142.4), [7]Lefty Gomez (38.2/62.7)
1961: [1]Joe Gordon (54.9/84.3), [6]Dizzy Dean (41.8/62.3)
1962: [1]Enos Slaughter (54.1/69.2), [1]Bob Elliott (52.3/68.5)
1963: [1]Lou Boudreau (56.0/88.3), [2]Charlie Keller (43.4/63.9)
1964: [1]Ted Williams (125.3/244.8), [1]Bob Feller (63.3/107.3)
1965: [1]Jackie Robinson (63.2/112.6), [2]Pee Wee Reese (66.7/94.7)
1966: [1]Stan Musial (127.8/229.6), [1]Early Wynn (58.3/74.9)
1967: [1]Warren Spahn (89.7/142.5), [1]Hal Newhouser (57.5/90.4)
1968: [1]Ralph Kiner (45.9/72.2), [3]Bob Lemon (51.0/64.5)
1969: [1]Larry Doby (47.4/67.7), [6]Bobby Doerr (47.7/62.7)
1970: [10]Dizzy Trout (47.1/62.0), [1]Gil Hodges (44.6/59.3)
1971: [1]Yogi Berra (61.9/83.1), [1]Minnie Minoso (52.8/79.3)
1972: [1]Robin Roberts (82.7/127.8), [1]Duke Snider (67.5/108.5)
1973: [1]Richie Ashburn (58.0/83.7), [1]Billy Pierce (51.8/72.5)
1974: [1]Whitey Ford (56.8/71.6), [2]Nellie Fox (44.4/60.6)
1975: [10]Vern Stephens (43.5/56.3), [6]Al Rosen (33.0/54.2)
1976: [1]Bob Friend (42.7/61.4), [10]Roy Campanella (36.2/52.2)
1977: [1]Willie Mays (154.7/304.9), [1]Mickey Mantle (120.2/234.3)
1978: [2]Eddie Mathews (98.3/165.8), [2]Ernie Banks (64.4/110.5)
1979: [3]Jim Bunning (57.1/95.1), [3]Ken Boyer (58.4/84.7)
1980: [1]Hank Aaron (141.6/251.8), [1]Al Kaline (91.0/136.5)
1981: [1]Frank Robinson (107.4/167.6), [1]Bob Gibson (91.8/157.4)
1982: [3]Roberto Clemente (83.8/129.1), [1]Don Drysdale (69.8/102.6)
1983: [1]Juan Marichal (62.7/105.9), [1]Brooks Robinson (69.1/99.7)
1984: [1]Gaylord Perry (90.5/152.2), [1]Willie McCovey (65.1/97.0)
1985: [1]Phil Niekro (93.0/148.5), [1]Carl Yastrzemski (88.7/144.3)
1986: [1]Ron Santo (66.4/111.7), [6]Sandy Koufax (48.7/96.1)
1987: [1]Pete Rose (75.3/110.7), [6]Harmon Killebrew (61.1/85.5)
1988: [1]Fergie Jenkins (81.5/125.1), [1]Dick Allen (61.2/99.7)
1989: [1]Joe Morgan (103.5/182.7), [2]Jimmy Wynn (59.8/96.0)
1990: [1]Tom Seaver (106.1/172.4), [1]Steve Carlton (86.9/137.7)
1991: [1]Rod Carew (79.1/122.2), [1]Jim Palmer (63.1/94.9)
1992: [1]Reggie Jackson (74.6/116.5), [3]Sal Bando (60.6/93.8)
1993: [1]Nolan Ryan (80.5/112.7), [1]Johnny Bench (71.3/108.9)
1994: [2]Carlton Fisk (67.3/88.7), [4]Reggie Smith (63.4/85.5)
1995: [1]Mike Schmidt (108.3/196.1), [1]Bobby Grich (67.6/97.2)
1996: [2]Rick Reuschel (65.3/91.9), [6]Don Sutton (65.6/84.5)
1997: [1]Bert Blyleven (87.6/135.0), [8]Graig Nettles (61.6/83.5)
1998: [2]Buddy Bell (60.8/82.6), [7]Bobby Bonds (57.0/80.9)
1999: [1]George Brett (85.0/136.8), [1]Keith Hernandez (61.0/87.9)
2000: [1]Gary Carter (66.3/98.3), [1]Ozzie Smith (64.6/87.1)
2001: [1]Robin Yount (76.9/117.2), [5]Dave Winfield (59.7/80.1)
2002: [1]Paul Molitor (74.8/102.8), [1]Eddie Murray (66.7/91.1)
2003: [1]Lou Whitaker (69.7/91.4), [7]Dwight Evans (61.8/79.6)
2004: [1]Rickey Henderson (113.1/189.2), [1]Wade Boggs (89.0/147.5)
2005: [2]Alan Trammell (66.9/98.0), [1]Ryne Sandberg (62.0/91.9)
2006: [1]Cal Ripken (89.9/144.8), [1]Tony Gwynn (68.4/92.2)
2007: [3]Tim Raines (64.6/88.4), [8]Andre Dawson (57.0/79.4)
2008: [1]Roger Clemens (128.8/214.0), [6]Dave Stieb (53.0/78.6)
2009: [1]Randy Johnson (89.6/148.1), [1]Edgar Martinez (67.2/99.5)
2010: [1]Barry Bonds (171.8/355.2), [1]Barry Larkin (68.9/96.6)
2011: [1]Craig Biggio (66.2/95.1), [1]Kevin Brown (64.0/95.0)
2012: [1]Greg Maddux (96.9/148.1), [1]Curt Schilling (67.7/98.7), [1]Larry Walker (67.3/96.0)
2013: [1]Kenny Lofton (65.3/92.5), [5]Mark McGwire (63.1/91.0)
2014: [1]Jeff Bagwell (79.9/128.2), [1]Frank Thomas (75.9/112.6)
2015: [1]Ken Griffey (78.5/130.5), [2]Mike Mussina (74.6/104.0), [2]Sammy Sosa (59.7/94.2)
2016: [1]Jim Edmonds (68.3/105.6), [1]Jim Thome (70.3/99.8)
2017: [1]Pedro Martinez (73.5/122.1), [4]Roberto Alomar (63.5/92.1)
2018: [1]Chipper Jones (80.1/119.1), [1]Manny Ramirez (67.5/92.3), [5]Mike Piazza (59.1/90.5)
2019: [1]Todd Helton (57.9/90.6), [3]Ivan Rodriguez (67.7/89.7)
2020: [1]Derek Jeter (70.1/103.3), [9]Tom Glavine (71.6/89.6)


#5    Tangotiger      (see all posts) 2012/01/02 (Mon) @ 11:02

How is Smoltz not getting in there?

Great job again…


#6    Tangotiger      (see all posts) 2012/01/02 (Mon) @ 12:08

Since you are doing such a fantastic job here, let me tell you how I was handling the pre-1931 born players.

The worst player (among the HOFers) born in the 1930s is probably Ken Boyer.  So, for players born in the 1910s-1920s, I don’t want anyone worse than Boyer in.  AND, I don’t want anyone BETTER than Boyer NOT in (such that we’d want to carry him over to the Mays/Mantle ballot).

That is, I want to be able to draw that line at Mays/Mantle, such that no one before them would carryover, and be elected in a weak year (Bunning/Boyer).

Therefore, here is what you can play with, if you have the time:

1. How late do you start the balloting process, such that the above is true?  Basically, you want it so that Minnie Minosa and/or Whitey Ford is elected in the 1976 ballot.  Anybody worse than them (born in the 1910s-20s) may carryover to Mays/Mantle, but since they’d be worse than Boyer/Bunning, wouldn’t be elected.

(We of course have to contend with Negro League players, and we’ll discuss them shortly).

2. How late do you start if it was an elect-1 (up to Mays/Mantle, which goes to elect-2), while still holding to the above?


#7          (see all posts) 2012/01/02 (Mon) @ 12:08

My original thought was that if you were doing pick 2 from all the years of 16 teams in MLB, (and earlier), then you would want pick 3 by 1990, maybe earlier, and we would be phasing into pick 4 now.  But looking at the many weak selections in the earlier years, I think that alternating pick 1 and pick 2 for the first 70 years would better fit my preference for minimum quality.

Thanks for all the work.  Could you run lists of the best players left out for various scenarios? 

(One problem I had with the HoM pick 2, pick 3 arrangement was that no one ever fell off the back log, so they were selecting players with very low vote % and no consensus.)


#8    Tangotiger      (see all posts) 2012/01/02 (Mon) @ 12:15

Counting real quick:

From the Mays/Mantle election onwards:
8 of the first 20 players elected are non-whites
7 of the next 20 are non-whites
7 of the next 20 are non-whites
8 of the next 22 are non-whites

That’s a pretty good pattern of saying around 35-40% of superstar players are non-whites.

So, for the pre-Mays/Mantle elections, we want to ensure that that remains true.  That means that we want that many non-whites elected.  And if you don’t know enough, then leave some spots open.  So, you can do an elect-1, elect-1, elect-2 system, which will give you 2 of 6 non-white players.

That gives breathing room to elect Negro League players.

And since baseball likely didn’t attract the talent it does now in terms of vocation, we can probably stick to elect-1 up to Mays/Mantle, and we’ll get close to what we want.


#9    Tangotiger      (see all posts) 2012/01/02 (Mon) @ 12:20

I disagree strongly that you want the elect-N to be based on the number of teams.

Whether you have 8 teams in MLB or you have 80, you STILL have the same amount of top talent.  The top talent gets more opportunities in an 80-team league to shine (a bad year means they still are playing, as opposed to an 8-team league where they might get sent down to the lower-leagues).

The ONLY reason you want more players these days is if the pool of incoming SUPERSTAR talent is greater these days, because the money today draws more of the great players to choose baseball, to develop as baseball players early on in their lives, etc.


#10    Mike G.      (see all posts) 2012/01/02 (Mon) @ 21:24

Good point about the top talent count being unrelated to the number of teams/qualifying players. If anything, I’d think the number of teams inflate the WAR (i.e. lower replacement level). Still, I’d love to come up with a heuristic for choosing the years to change elect-N. But I can see how the best we can do is feel/historical knowledge to estimate when to raise it up.

Given that there are 71 players better than Boyer (according to wWAR) before 1977 Mays/Mantle, it seems like elect-1 is the best fit. I thought about making it smaller at the beginning, but there’s quite a bit of top talent right up front that it just didn’t work. In fact, I had to start in 1899 just to get Cap Anson in this Hall.

As I mentioned when I first posted it, wWAR seems to underweight relievers. NO reliever made it into this cut. No Eck, Goose, Smoltz, or even Mo. As an experiment, I tried multiplying the seasonal WAR by aLI. It still feels wrong, but it did pull in the top relievers albeit probably too many (first number in parens is how much was added to WAR thanks to the aLI factor):

[1]Mariano Rivera(48.1/101/175.9)
[1]Rich Gossage(34.1/73.6/141.6)
[10]Trevor Hoffman(30.2/60.5/86.9)
[6]Lee Smith(28/57.7/81.9)
[1]Bruce Sutter(24.1/48.4/97.5)
[10]John Hiller(18.3/45.8/83.4)
[1]Hoyt Wilhelm(14.4/52.3/69.2)
[1]Dennis Eckersley(9.7/68/89.5)
[1]John Smoltz(6.2/71.5/91.3)

As to the overall results, here’s the latest run. When I’m out of time on this, I’ll import this into a Google spreadsheet and share the link here. (Note: number next to the year is total number of candidates on the ballot… those who played 10+ years)

1899[12]: [2]Cap Anson (88.9/124.7),
1900[16]: [1]Charley Radbourn (73.6/178.2),
1901[22]: [2]Will White (32.5/73.4),
1902[30]: [1]Pud Galvin (58.8/147.3),
1903[44]: [1]Tim Keefe (79.0/170.3),
1904[55]: [2]Roger Connor (87.2/136.8),
1905[64]: [2]Dan Brouthers (81.7/134.9),
1906[68]: [5]Jim McCormick (63.7/129.0),
1907[70]: [1]John Clarkson (81.3/191.3),
1908[72]: [2]Charlie Buffinton (57.4/128.6),
1909[81]: [5]Tony Mullane (69.6/115.4),
1910[88]: [1]Bob Caruthers (71.4/124.6),
1911[89]: [7]Mickey Welch (55.4/105.1),
1912[95]: [1]Billy Hamilton (69.6/107.7),
1913[92]: [1]Cy Young (143.2/279.7),
1914[90]: [1]Silver King (53.1/136.6),
1915[97]: [1]Kid Nichols (103.2/201.2),
1916[97]: [1]George Davis (90.3/135.9),
1917[101]: [5]Ed Delahanty (74.7/128.5),
1918[105]: [2]Amos Rusie (62.1/124.0),
1919[106]: [4]Bill Dahlen (75.9/104.9),
1920[109]: [1]Honus Wagner (134.5/254.9),
1921[109]: [2]Nap Lajoie (104.2/177.3),
1922[116]: [2]Eddie Plank (76.8/107.4),
1923[109]: [6]Fred Clarke (73.4/101.5),
1924[108]: [8]Joe McGinnity (49.5/91.6),
1925[102]: [7]Bobby Wallace (65.9/89.7),
1926[108]: [1]Christy Mathewson (90.7/163.5),
1927[110]: [2]Sam Crawford (76.6/107.4),
1928[107]: [2]Ed Walsh (56.8/102.3),
1929[104]: [8]Elmer Flick (56.7/86.7),
1930[106]: [9]Vic Willis (50.4/85.2),
1931[108]: [10]Mordecai Brown (55.6/83.3),
1932[98]: [1]Ty Cobb (159.5/306.1),
1933[111]: [1]Walter Johnson (139.8/250.0),
1934[118]: [1]Tris Speaker (132.9/240.0),
1935[125]: [3]Eddie Collins (126.7/230.8),
1936[134]: [4]Pete Alexander (106.8/179.8),
1937[143]: [3]Joe Jackson (62.9/109.4),
1938[158]: [7]Frank Baker (63.7/105.9),
1939[165]: [3]Dazzy Vance (53.8/88.6),
1940[165]: [1]Harry Heilmann (69.4/105.4),
1941[166]: [1]Babe Ruth (190.0/425.2),
1942[181]: [1]Rogers Hornsby (127.8/266.1),
1943[182]: [5]George Sisler (52.5/84.0),
1944[180]: [1]Frankie Frisch (74.8/114.2),
1945[178]: [2]Bill Terry (55.4/83.2),
1946[176]: [1]Lefty Grove (93.3/168.9),
1947[172]: [2]Goose Goslin (63.0/88.0),
1948[161]: [1]Al Simmons (63.6/93.5),
1949[163]: [1]Lou Gehrig (118.4/242.6),
1950[163]: [2]Charlie Gehringer (80.9/131.8),
1951[167]: [3]Paul Waner (73.8/107.3),
1952[162]: [4]Carl Hubbell (63.6/97.5),
1953[171]: [1]Jimmie Foxx (95.2/164.9),
1954[177]: [2]Luke Appling (69.3/99.1),
1955[176]: [1]Mel Ott (109.3/180.2),
1956[177]: [5]Joe Cronin (62.5/92.0),
1957[174]: [1]Hank Greenberg (56.8/91.1),
1958[175]: [1]Arky Vaughan (75.6/123.5),
1959[170]: [1]Johnny Mize (70.2/110.5),
1960[165]: [1]Joe DiMaggio (83.6/142.4),
1961[162]: [1]Joe Gordon (54.9/84.3),
1962[171]: [6]Joe Medwick (55.8/81.6),
1963[167]: [1]Lou Boudreau (56.0/88.3),
1964[160]: [1]Ted Williams (125.3/244.8),
1965[159]: [1]Jackie Robinson (63.2/112.6),
1966[157]: [1]Stan Musial (127.8/229.6),
1967[162]: [1]Warren Spahn (95.4/158.9),
1968[169]: [5]Bob Feller (63.2/107.3),
1969[170]: [6]Pee Wee Reese (66.7/94.7),
1970[172]: [4]Hal Newhouser (57.8/90.8),
1971[171]: [1]Yogi Berra (61.9/83.1),
1972[167]: [1]Robin Roberts (83.5/129.4),
1973[173]: [2]Duke Snider (67.5/108.5),
1974[182]: [2]Richie Ashburn (58.0/83.7),
1975[191]: [5]Minnie Minoso (52.8/79.3),
1976[197]: [4]Billy Pierce (53.5/77.3),
1977[199]: [1]Willie Mays (154.7/304.9), [1]Mickey Mantle (120.2/234.3)
1978[192]: [2]Eddie Mathews (98.3/165.8), [2]Ernie Banks (64.4/110.5)
1979[193]: [3]Jim Bunning (59.0/101.3), [3]Ken Boyer (58.4/84.7)
1980[201]: [1]Hank Aaron (141.6/251.8), [1]Al Kaline (91.0/136.5)
1981[214]: [1]Bob Gibson (98.9/183.3), [1]Frank Robinson (107.4/167.6)
1982[216]: [3]Roberto Clemente (83.8/129.1), [1]Don Drysdale (72.9/111.6)
1983[211]: [1]Juan Marichal (62.3/106.1), [3]Sandy Koufax (51.2/104.6)
1984[216]: [1]Gaylord Perry (98.5/177.6), [2]Brooks Robinson (69.1/99.7)
1985[220]: [1]Phil Niekro (99.1/165.5), [1]Carl Yastrzemski (88.7/144.3)
1986[221]: [1]Ron Santo (66.4/111.7), [3]Willie McCovey (65.1/97.0)
1987[243]: [1]Pete Rose (75.3/110.7), [2]Luis Tiant (62.4/90.9)
1988[261]: [1]Fergie Jenkins (82.4/129.7), [1]Dick Allen (61.2/99.7)
1989[273]: [1]Joe Morgan (103.5/182.7), [2]Jimmy Wynn (59.8/96.0)
1990[288]: [1]Tom Seaver (108.4/178.1), [1]Steve Carlton (91.4/152.6)
1991[288]: [1]Rod Carew (79.1/122.2), [2]Sal Bando (60.6/93.8)
1992[297]: [1]Reggie Jackson (74.6/116.5), [2]Jim Palmer (61.6/92.6)
1993[315]: [1]Nolan Ryan (84.7/123.9), [1]Johnny Bench (71.3/108.9)
1994[327]: [2]Carlton Fisk (67.3/88.7), [4]Don Sutton (66.3/87.2)
1995[338]: [1]Mike Schmidt (108.3/196.1), [1]Rick Reuschel (69.5/102.0)
1996[350]: [2]Bobby Grich (67.6/97.2), [6]Reggie Smith (63.4/85.5)
1997[359]: [1]Rich Gossage (73.6/141.6), [1]Bert Blyleven (88.8/138.6)
1998[365]: [9]Graig Nettles (61.6/83.5), [10]John Hiller (45.8/83.4)
1999[373]: [1]George Brett (85.0/136.8), [1]Bruce Sutter (48.4/97.5)
2000[378]: [1]Gary Carter (66.3/98.3), [1]Dennis Eckersley (68.0/89.5)
2001[387]: [1]Robin Yount (76.9/117.2), [3]Keith Hernandez (61.0/87.9)
2002[389]: [1]Paul Molitor (74.8/102.8), [1]Eddie Murray (66.7/91.1)
2003[394]: [1]Lou Whitaker (69.7/91.4), [1]Dave Stieb (55.8/87.5)
2004[392]: [1]Rickey Henderson (113.1/189.2), [1]Wade Boggs (89.0/147.5)
2005[393]: [2]Alan Trammell (66.9/98.0), [1]Ryne Sandberg (62.0/91.9)
2006[404]: [1]Cal Ripken (89.9/144.8), [1]Tony Gwynn (68.4/92.2)
2007[393]: [3]Tim Raines (64.6/88.4), [8]Ozzie Smith (64.6/87.1)
2008[394]: [1]Roger Clemens (130.0/217.6), [6]Lee Smith (57.7/81.9)
2009[406]: [1]Randy Johnson (89.5/148.9), [1]Edgar Martinez (67.2/99.5)
2010[416]: [1]Barry Bonds (171.8/355.2), [1]Barry Larkin (68.9/96.6)
2011[428]: [1]Kevin Brown (65.7/99.5), [1]Craig Biggio (66.2/95.1)
2012[441]: [1]Greg Maddux (98.3/154.8), [1]Larry Walker (67.3/96.0), [1]Curt Schilling (65.4/93.9)
2013[454]: [1]Kenny Lofton (65.3/92.5), [1]John Smoltz (71.5/91.3)
2014[467]: [1]Jeff Bagwell (79.9/128.2), [1]Frank Thomas (75.9/112.6)
2015[475]: [1]Mariano Rivera (101.0/175.9), [1]Ken Griffey (78.5/130.5), [2]Mike Mussina (70.5/95.5)
2016[462]: [1]Jim Edmonds (68.3/105.6), [1]Jim Thome (70.3/99.8)
2017[472]: [1]Pedro Martinez (73.6/123.2), [4]Sammy Sosa (59.7/94.2)
2018[474]: [1]Chipper Jones (80.1/119.1), [1]Manny Ramirez (67.5/92.3), [5]Roberto Alomar (63.5/92.1)
2019[471]: [1]Todd Helton (57.9/90.6), [6]Mike Piazza (59.1/90.5)
2020[464]: [1]Derek Jeter (70.1/103.3), [9]Tom Glavine (72.3/90.5)

Top wWAR who didn’t play 10 years:
1908 Ed Morris 37.6 98.9
1902 Tommy Bond 40.3 96.1
1902 Guy Hecker 39.3 86.1
1913 Sadie McMahon 36.8 68.4

Top wWAR who were dropped from the ballot:
1923 Jesse Burkett 64.1 92.2
2018 Mark McGwire 63.1 91.0
2019 Rafael Palmeiro 66.0 88.5
1924 Hughie Jennings 48.0 87.0
1927 Willie Keeler 60.8 86.0
1991 Harmon Killebrew 61.1 85.5
1997 Sam McDowell 46.3 83.9
1915 John Ward 64.9 83.4
2018 David Cone 57.1 83.0
2006 Buddy Bell 60.8 82.6

Players elected before Ken Boyer worse than Ken Boyer:
YOB Year Cand Yr Name WAR wWAR
1855 1901 22 2 Will White 32.5 73.4
1930 1976 197 4 Billy Pierce 53.5 77.3
1929 1975 191 5 Minnie Minoso 52.8 79.3
1916 1962 171 6 Joe Medwick 55.8 81.6
1925 1971 171 1 Yogi Berra 61.9 83.1
1899 1945 178 2 Bill Terry 55.4 83.2
1885 1931 108 10 Mordecai Brown 55.6 83.3
1928 1974 182 2 Richie Ashburn 58.0 83.7
1897 1943 182 5 George Sisler 52.5 84.0
1915 1961 162 1 Joe Gordon 54.9 84.3


#11    Matthew Cornwell      (see all posts) 2012/01/02 (Mon) @ 21:52

Let’s throw in a WPA “Clutch” component. Maybe 1/2 career “Clutch”? And a postseason component while we are at it?  Maybe post-season WPA?


#12    Tangotiger      (see all posts) 2012/01/02 (Mon) @ 22:54

Mike: again, fantastic job.

I think we’ve shown here how with very little effort, we can come up with a pretty solid system.

I’ll have to look at the reliever issue.


#13    Tangotiger      (see all posts) 2012/01/04 (Wed) @ 12:23

Going back to his first list, we see this:

- Mark McGwire (63.1/91.0)
- Tom Glavine (71.6/89.6)
- Rafael Palmeiro (66.0/88.5)
- Luis Tiant (59.4/83.7)
- John Smoltz (65.3/81.7)
- John Olerud (56.8/80.9)

That’s under the strict elect-2 (and no post-season allowance).

A commenter at Primer said he doesn’t think the system works because Glavine didn’t make it.

Now, granted that I don’t like that Glavine is out and Ke Brown is in.  However, given the CURRENT hall of fame standards, the one that won’t elect McGwire, Palmeiro, didn’t elect Tiant and Olerud (i.e., we have agreement here), the only worthy players not inducted under my revised system is Glavine and Smoltz.

But, I do get Trammell, and I get Dw Evans, and Keith Hernandez, and Lou Whitaker.  I don’t make players wait many years, with the majority elected in their first year.

So, I don’t get why I don’t get Glavine is necessarily a fail of the system. 

And in any case, as I said, we can introduce elect-3 every 3 years, when the voters think they need it.

As well, if you use a straight WAR, as opposed to the WAR-squared, Glavine DOES make it. 

So, my system was based strictly on using the numbers, but clearly, a voter will use his own system, and Glavine is one that will get more votes than a WAR-squared system would presume.


#14    Mike G.      (see all posts) 2012/01/04 (Wed) @ 21:36

Here is a Google Docs link to a bunch of the data, as last processed.

https://docs.google.com/spreadsheet/ccc?key=0AsxDiA48eX0ldGlEb0daVGdvU3d4b2hWdEhRamJsNmc

Sheet 1: “By Year”
What happened in each year:
- How many candidates
- Who was elected
- Who was the top player eliminated
- Who were the top 5 players carried over to next year
Each player shows the year on the ballot, the raw WAR from BBRef, and the weighted WAR (wWAR).

Sheet 2: “Elected”
All those elected into the rebooted HOF, with:
- YOB: Year of birth
- Elect: Year elected
- Yrs: The ballot year of election
- WAR: BBRef’s WAR
- liWAR: how much WAR was added by factoring in leverage
- PF: “Peak Factor” - how much the weighting inflates the flat WAR.
Note: wWAR = (WAR + liWAR) * PF

Sheet 3: “Leveraged”
The players most helped by factoring in aLI, with:
- WAR/PF/wWAR - those values without using aLI
- wLI - ratio of leveraged WAR to raw WAR
- liWAR/liPF/liwWAR - the values using aLI
Note that liPF and liwWAR correspond to the PF and wWAR from other sheets. Sorry for the confusion.

Sheet 4: “Ineligible”
The highest wWAR players who did not play for 10 or more seasons (and thus are ineligible).

Sheet 5: “Missed”
The highest wWAR players who did not make it in the HOF reboot, with:
- Their first and last year on the ballot
- The lowest wWAR player during those 10 years who was elected (with their election year and wWAR). Effectively, the player who kept them out.

My hope is that you can sort/filter the sheets to your heart’s delight. If that’s not the case, could someone let me know how to do that? I’m new at sharing Google Docs.

Enjoy!


#15    Matthew Cornwell      (see all posts) 2012/01/04 (Wed) @ 22:42

Mike-

Trying to figure out how PF works:

According to WAR, Glavine, for example had 31 WAR in his top 5 seasons, but Sutton had closer to 25. Glavine leads over their next 5-10 years too, but Sutton has a PF of over 1.3 to Glavine’s 1.25.

I know pitcher offense is considered in their career WAR totals, but are you not considering it in their PF or wWAR? 

Their liWAR are the same, so that isn’t it.


#16    Matthew Cornwell      (see all posts) 2012/01/04 (Wed) @ 23:22

Given Ruffing’s lower-than-expected score (under 80) despite a tWAR of close to 70 and Walter Johnson’s finishing below Cy Young, I am guessing that my hunch about pitcher offense not being included in the squaring/wWAR is right.

If that is true, the Primer commenter’s question would be void, as Glavine would pick up more wWAR points and move ahead of quite a few players that he is currently behind.


#17    Mike G.      (see all posts) 2012/01/05 (Thu) @ 07:00

Matthew,

The PF is simply a ratio of the weighted WAR (sum of 0.002x^3 + 0.13x^2 + 0.695x, per season) to the unweighted (but leveraged) WAR. I thought it would be more interesting to view it that way instead of an additive amount.

Pitcher offense is considered in wWAR, though not as part of the leverage.

Regarding Glavine versus Sutton, I calculated it by hand (well, MS Excel), and I’m getting different PF values (Glavine=1.34, Sutton=1.23). It’s entirely possible I’ve got a bug in my calculation. I’ve only “smell-tested” it so far. Thanks for bringing this to my attention; I’ll try to get my hands dirty and figure out what’s going on. If I discover a bug, I’ll update the Google Docs accordingly (and let everyone know).

I don’t know what the Primer commenter’s question is, so I can’t speak to that. Could you post the link to that?

Thanks.


#18    Matthew Cornwell      (see all posts) 2012/01/05 (Thu) @ 10:10

#17

I was referencing Tom’s comment about how the Primer commenter thought the system was flawed since Glavine did not make it.  I haven’t seen the comment myself.


#19          (see all posts) 2012/01/05 (Thu) @ 10:21

Is there some way we can adjust for voter opinions, like including MVP or Cy Young Award shares? That might change the relievers some, and will account for the fact that the voters won’t vote for Jimmy Wynn come hell or high water.


#20    pierre      (see all posts) 2012/01/05 (Thu) @ 11:14

#19- I guess you’d adjust for team success (somehow) and dampen the park adjustment factors.  Question is how much to reward Jimmy Wynn for the HR he would have hit if he hadn’t played in the Astrodome circa 1970 for a team nobody cared about.  imho, Wynn is a really tough sell.  To me, the moral is that life isn’t always fair, not that Wynn should be in the HOF.


#21    JEH      (see all posts) 2012/01/05 (Thu) @ 13:32

Tango/#9

“disagree strongly that you want the elect-N to be based on the number of teams.

Whether you have 8 teams in MLB or you have 80, you STILL have the same amount of top talent.  “

That’s a strong statement.

If we went from 30 to 80 teams tomorrow that would make sense, but the increase in teams would seem to be tied to an increase in population and an increase in population would seem to be lead to an increase in top talent.  So baseball would seem to have a larger talent pool and be more effectively developing it.

The population of the US (and the world) has quadrupled since 1900 (of more relevance is the increase in births per year which was up about 50% in the US from 1900 to 1965).  The percentage of the world playing baseball has increased.  I will assume the percentage of the population playing organized baseball has also increased. 

In short, I think a system that increases (or varies, relative to the number of teams, population or some other proxy for the talent pool) the number elected over time makes sense.


#22          (see all posts) 2012/01/05 (Thu) @ 14:44

A few things went through my head when seeing Rick Rueschel’s placement…

1.  In post #4, Rueschel had a 91.2 wWAR.  After leveraging, Rueschel was at 102 wWAR in post #10.  The Google Docs spread sheet gives him a LI of +4. I am pulling out from the descriptions above that LI does not = 1 wWAR point, but is calculated and weighed at the seasonal level first and figured into the overall wWAR?  So Rueschel’s positive leveraged seasons must have come during some of his “peakiest” seasons, I guess. Just making sure I understand correctly.

2. Looking at Rueschel’s WPA “Clutch” score, it looks like he performed a little worse during high-leveraged situations.  How would that change what we do with his LI?

3.  Looking at Rueschel’s leveraging (and those of Carlton, Perry, Niekro, etc.) it is obvious that most HOF pitchers from that era all have significantly better LI than modern pitchers.  I am assuming that is no coincidence.  Is it because pitchers of that era went into the 8th and 9th innings so much more than modern pitchers?  Then I am trying to figure out how Mike Mussina had a LI -4? 

4.  I have heard people discuss that park factors are not adjusted pitcher-to-pitcher, and that an extreme GB pitcher like Rueschel was not negatively effected by Wrigly nearly as much as pitcher with normal GB/FB distribution. I have heard this case made for Niekro too and Nolan Ryan on the flip-side as well. Is there any truth to this, and if so, how could we adjust his WAR accordingly? 

Thanks!


#23    Tangotiger      (see all posts) 2012/01/05 (Thu) @ 16:26

I just want to make clear that I am not necessarily advocating for a one-size-fits-all system, in terms of using WAR or weighted-WAR or whatever.

I am ONLY showing what happens if you DO use that, within the overall framework I’m proposing (elect-2, based on birth year, staying eligible for 10 years, eligible as of age 45).

Using WAR is ONLY to show what it would look like as a final result.  Think of it as a “mock draft”.  That’s the only reason we are using WAR here.

We have something tangible to discuss.


#24    Tangotiger      (see all posts) 2012/01/05 (Thu) @ 16:30

It’s possible for example that if Glavine is on his last ballot, that there would be a push for him, rather than some first timer.

Looking at Mike’s fantastic google chart, we see Killebrew eliminated on the last ballot, in favor of Sal Bando, who was 2nd year on the ballot.  I think there’s no doubt here that the electorate, seeing how close they may be, will simply rank Killebrew ahead of Bando.

So, a more realistic Mock Ballot process would give “bonus points” for guys who are lasting a long time on the ballot.


#25    Mike G.      (see all posts) 2012/01/05 (Thu) @ 21:19

Matthew/#15,

I figured out where the Glavine vs. Sutton unusual PF is coming in. It’s because I’m computing the wWAR for batting and pitching separately for each season. At the low seasonal batting WAR (-1 to 1) each of them has, the WAR squared algorithm actually moves the number closer to 0 (e.g., 0.5^2 = 0.25). So, Sutton’s -5.2 offensive WAR goes to -4.0, and Glavine’s 4.6 goes to 3.5. That’s how Sutton is getting the edge here. Unfortunately, I see now that the batting and pitching should be summed before squaring them. If I have time, I’ll update the spreadsheet accordingly.

Matthew/#22,

Looking at BBRef’s splits, Reuschel pitched in the 9th in 25% of his starts (133 of 529). Mussina pitched in 12% (64 of 536). They had nearly the same amount of Batters Faced, but Reuschel has a ton more high leverage PAs (3003 to 2161).

liWAR will be positive for any year with an aLI > 1.0, and negative for any year < 1.0. Mussina had 10 negative years and no positive ones. Reuschel had 2 negative and 11 positive.

Tango/#24,

I like the idea of the “bonus points” for long-time ballot-riders.  If I do update the charts, I’ll throw that in, too.  Looking at Bando vs. Killebrew, I had no idea that Bando had an incredible ‘69, and one more 5+ season than Killebrew.


#26    Matthew Cornwell      (see all posts) 2012/01/05 (Thu) @ 22:24

I have thought about this recently:  WAR gives more credit to pitchers from eras who were allowed to pitch more innings, of course. This is why Phil Niekro’s best seasons look similar to Greg Maddux’s best seasons and why Jim Bunning’s best WAR seasons look like Randy Johnson’s best seasons.  I have always wondered how much we should consider the fact that the large IP difference between 60’s-70’s guys and 90’s-00’s guys is not the “fault” of the modern pitcher, but a manager issue or just a result of the era changing.  Is it really Greg Maddux’s fault that he wasn’t allowed to pitch more than 260 innings?  Maddux finished 1st in IP 5-6 times, but was acquiring the typical replacement value those seasons that a guy routinely finishing 8th or 9th would in the 70’s.

I don’t know what can be done about it, other than using WAA instead of WAR, which post-80’s pitchers do much better compared with earlier pitchers relative to how they compare using WAR. I don’t think an adjusted IP (or BF) would work.

I feel the same thing may apply with LI...I hate to see modern pitchers “punished” since they were not allowed to pitch deeper.  Unless we feel that they have become physically unable to do so. 

Just thinking out-loud here.


#27          (see all posts) 2012/01/06 (Fri) @ 17:38

I have always wondered how much we should consider the fact that the large IP difference between 60’s-70’s guys and 90’s-00’s guys is not the “fault” of the modern pitcher, but a manager issue or just a result of the era changing.

There are a lot of little reasons that have led to that. Better offenses, switching from linear hitting to rotational hitting, offense preferred to defense in many positions, shorter fences, smaller strike zone, etc.

In other words, pitchers don;t breeze through the second half of the lineup anymore, and put more effort into each pitch, etc.

But is that reflected in WAR? There’s more teams now, so wouldn’t the replacement level be lower than when there were 14 teams?

There’s a lot of aspects to consider.

Personally, I don’t care to compare players from different eras. I am primarily interested in how a player dominated his era. So, I would prefer to look at things like WAA to see how a player stood out among his peers. I suppose WAR would be just fine since it’s WAA - 20% or something like that.

I would also choose to ignore any seasons that were league average or below to eliminate the “accumulation” of WAR and/or counting statistics in seasons that are not “better than average seasons”.

Wouldn’t the WAR of a replacement level pitcher in the 250+IP era be higher than the WAR of the SP in the 150 IP era?

It’s still measuring the same thing, right? How mcuh more vlaue did the pitcher provide than a replacement level player (and it may even be determined using the league average pitcher as a base).


#28    Matthew Cornwell      (see all posts) 2012/01/06 (Fri) @ 18:57

I know what you are saying. The thing that gets me is that Seaver, for example, who never finished higher than 3rd in his league in IP was receiving more replacement value during his prime than did Gred Maddux who dominated his league in IP like few ever have. 

Anyway, I like to look at WAA too, but am not fond of the idea of ignoring seasons below average all-together, as accumulating any value over replacement is still better than being injured or retired or on the bench or having a AAA player pitching instead.  That’s why weighted WAR methods work well. The focus-in on peak, but don’t ignore the value a player accumulates over time due to longevity and being better than a replacement player would too.


#29    pierre      (see all posts) 2012/01/07 (Sat) @ 10:12

WAM.  Wins Above Mediocre.  The justification for using it would be that given time, teams are able to find suitable replacements.  i.e. the guy who spot starts comes off the scrap heap, but the guys) who would have pitched instead of the HOFer probably wouldn’t have been half bad.  WAM or WAA would “like” Sandy Koufax-types more than WAR. 

regarding the 200 IP v 300 IP SP, I’m not sure there’s much you can do other than bear it in mind and/or rank order pitchers within each era.  The big, bad #1 starter was in fact a more powerful, valuable player back in the day.


#30    Matthew Cornwell      (see all posts) 2012/01/07 (Sat) @ 11:00

I have been thinking abut that a lot too…

Of course if you do look at an Above Average stat like Base-Out Wins Saved (REW), those “big, bad, more valubale #1 starters from back in the day like Perry, Niekro, Ryan, Jenkins, Marichal and Carlton ALL finish below Brown, Mussina, Glavine, Smoltz, Haliday, and Schilling.  Of course REW doesn’t include defensive support or pitcher offense, so those rankings may not be 100% true after those adjustments were made.  And yet, those rankings doesn’t seem right either - I automatically want to reward those 60’s 70’s guys for pitching so much more.  Oh well...wasn’t trying to hijack the thread, just thinking if I agree with using LI for non-relievers.  You are right, all we can do is adjust subjectively or only compare pitchers within eras.

Just out of curiosity, where would you personally put the baseline for “mediocre”?  About halfway between Rally’s (or FG if you prefer) and average?


#31    Mike G.      (see all posts) 2012/01/07 (Sat) @ 11:07

I agree with Matthew… the seasons that a player plays above replacement but below average do count for something.

I think I’ve come up with a better weighting algorithm to address four problems I saw in the original one:
1 - Relievers wouldn’t get much weighting, because their seasonal WARs never got very high (due to limited innings). Correcting for this by using aLI had problems as well.
2 - Conversely, the early pitchers were overweighted due to their extreme amount of innings pitched.
3 - Shortened seasons (1981, 1994) were undervalued.
4 - Batting performances by pitchers were skewed.

The new idea is to calculate the seasonal Peak Factor (PF) on a rate stat (WAR/650PA or WAR/300IP). Here are a few examples:

In 2004, Mariano Rivera earned 4.8 WAR over 78.2 innings. That’s a WAR/300IP of 18.4! Running that (well, absolute value, in case it’s negative) through the PF algorithm (0.002x^2 + 0.13*x + 0.695) produces a PF of 3.8. Multiply that by the original WAR of 4.8 yields a weighted WAR of 18.1. The old wWAR for this season (not adjusted by aLI) was 6.6.

In 1884, Mickey Welch earned 11.4 WAR over 557.1 innings. This season used to register 27.8 wWAR, but now is 17.9.

          Year      IP   WAR  WAR/300IP    PF    wWAR    Old
Rivera    2004    78.2   4.8      18.41   3.8    18.1    6.6
Welch     1884   557.1  11.4       6.14   1.6    17.9   27.8

The same applies for batters (e.g., 1994 Jeff Bagwell and even 1996 Tom Glavine).

          Year     PA    WAR  WAR/650PA    PF    wWAR    Old
Bagwell   1994    479    8.4      11.40   2.4    20.5   16.2
Glavine   1996     96    0.9       6.09   1.6     1.4    0.7

I’ve created another Google spreadsheet using this updated formula, plus a couple other adjustments:

https://docs.google.com/spreadsheet/ccc?key=0AsxDiA48eX0ldGZpV2dzX2c5dVpKYmRub0x2blhCbUE#gid=0

- As mentioned above, this algorithm works pretty well without needing to pull in aLI. So it doesn’t.

- This algorithm seemed to work better starting at the aesthetically-pleasing 1901.  It also doesn’t need to bump up to elect-3 every third year to get (almost) all the best modern players.

- Per Tango/#24, I tried to address the player that’s on his final ballot. I didn’t like rating him “better” just because he was in his final year, but instead decided that if he’s the top player not elected in the year, he gets to stay on the ballot.

For instance, in 1923, I have Amos Rusie (92.0 wWAR) elected, and Silver King (91.6) is in his final year. However, since he’s the best remaining, he gets to stay on for 1924. And he’s elected.

Technically, that means a single player could remain on the ballot indefinitely. Practically, with this rating system, no one ends up getting more than one chance. (Just look at the players who have a “10” in column “O” on the first page to find them.)


#32    pierre      (see all posts) 2012/01/07 (Sat) @ 11:14

not sure where you’d put it, exactly.  Make it the typical back-end starter or 4th outfielder?  It’s arbitrary, but so is replacement level, really. 

not familiar with REW.  I’d think that even if the baseline was set at average, the Perrys and Carltons would still dominate by virtue of having some so much more of it.  What am I missing?


#33    Mike G.      (see all posts) 2012/01/07 (Sat) @ 12:01

New algorithm analysis:

A total of four relievers (by my count) were elected: two full-time (Gossage and Rivera) and two part-time (Eckersley and Smoltz). This feels more right to me.

Only 11 players changed between the old and new algorithm. (Imagine that! Ruth, Mays, and Bonds were still elected.) 4 more players dropped due to changing the start date and keeping an elect-2 for the current era.  Here are the differences:

Players dropped - year elected: (WAR/old wWAR)
1911: Mickey Welch (55.4/105.1)
1924: Joe McGinnity (49.5/91.6)
1930: Vic Willis (50.4/85.2)
1947: Goose Goslin (63.0/88.0)
1962: Joe Medwick (55.8/81.6)
1974: Richie Ashburn (58.0/83.7)
1975: Minnie Minoso (52.8/79.3)
1998: John Hiller (27.5/83.4)
1999: Bruce Sutter (24.3/97.5)
2008: Lee Smith (29.7/81.9)
2011: Craig Biggio (66.2/95.1)
2017: Sammy Sosa (59.7/94.2)
2018: Roberto Alomar (63.5/92.1)
2018: Manny Ramirez (67.5/92.3)
2019: Todd Helton (57.9/90.6)

Players added: - year elected: (WAR/new wWAR)
1925: Hughie Jennings (48.0/91.5)
1931: Sherry Magee (59.1/84.9)
1947: Ted Lyons (62.9/91.7)
1956: Red Ruffing (67.3/100.0)
1974: Early Wynn (58.3/89.1)
1975: Whitey Ford (56.8/85.7)
1998: Buddy Bell (60.8/87.9)
2007: Frank Tanana (55.0/84.0)
2008: Chuck Finley (54.3/87.7)
2009: David Cone (56.9/103.1)
2019: Mark McGwire (63.1/101.7)

Manny & Sosa are currently front-runners for being elected post-2020, while Alomar and Helton get squeezed out in the drop to elect-2.

Hiller, Sutter and Smith are no longer getting the advantage from aLI.

Welch, McGinnity, and Willis are no longer getting the advantage from the huge number of innings they pitched each season.

McGwire gets in on year #11, pushing Lofton to the next year, who delays Pudge Rodriguez’s selection.

Tanana/Finley/Cone: I’ll have to delve into the details to see what made these three a surprising selection instead of players such as Dawson and Randolph. Count Saberhagen in this as well, as he’s the highest-rated player not to get elected.

Biggio: Somehow he missed getting elected, and it really wasn’t close.  The best he ever finished was 5th, behind Brown, Edgar, McGwire, and Saberhagen.  To get him elected, I had to bump it up to elect-3 every TWO years.

Future thoughts (as long as we’re entertained by tweaks to the “mock draft"):

I’d love to factor in postseason play (preferably an “above replacement” measure).  Unfortunately, I’m not aware of any easy way to get this information.

I’m also thinking that ideally the WAR rates used for calculating PF should be somehow regressed (though I’m statistically-challenged in that area). My unsophisticated idea would be that seasons below the 650PA/300IP marks should have their WAR rates based on nearby rates as well as the current season. That might make the numbers line up with my intuition on Saberhagen vs. Biggio (among others).


#34    Matthew Cornwell      (see all posts) 2012/01/07 (Sat) @ 13:04

#32

It is a WPA stat…

http://www.fangraphs.com/blogs/index.php/glossary/


#35          (see all posts) 2012/01/07 (Sat) @ 15:07

No Matthew, it is not a WPA stat, it never considers inning or score.  REW = RE24/(runs to wins conversion factor).  To get RE24 take each of the 24 base/out situations and figure the average change in run expectancy (RE) with each event,(single, walk, HR, CS etc.), in that one of the 24.  Apply these values to each event in a player/season and you have the RE24 estimate of runs above/below average for that player/season.  Do a weighted average for each event across the 24 base/out states and you have Linear Weights.

For certain uses I prefer RE24 to Linear Weights.  RE24 gives a more accurate measure of retrospective value, while LW is more of an estimate of true talent taken out of the specific context.  Take Alfonzo Soriano, (please....), he has been well above average at hitting HR’s and well below average at taking walks.  RE24 would see that as a lead off hitter he is more likely to come up with the bases empty, making his HR’s worth less than average, and his walks on average each worth more.  I think for MVP and HoF RE24 and REW are better than LW and LW converted to wins.  (Batting Wins?)

REW should be very close to WPA/LI, both ignore inning/score and both give more credit for a 3 run HR than a solo shot.  It is convenient for Fangraphs and B-Ref to file REW and RE24 with WPA, but they are really in the same family as LW.


#36    Tangotiger      (see all posts) 2012/01/07 (Sat) @ 21:44

First, I like this:

“if he’s the top player not elected in the year, he gets to stay on the ballot. “

Secondly, I guess we don’t really need to limit it to a maximum of 10 years on the ballot.

The top 10 guys who gets votes, stay on the ballot indefinitely. 

My idea was that we should only compare players to their peers, but realistically, a guy like, I dunno, Richie Ashburn, won’t keep getting high votes 15-20 years later.  So, you could simply keep him on indefinitely, as long as he finishes in the top 10 not elected.


#37    Matthew Cornwell      (see all posts) 2012/01/07 (Sat) @ 23:05

#35 - Thank you.  So it is a run expectancy measure similar to using Lin. weights, not a win probability measure.  Got it.  Either way, stripping away replacement value, Schilling, Mussina, Glavine, etc. all have higher totals than Niekro, Perry, Ryan, etc.


#38          (see all posts) 2012/01/08 (Sun) @ 00:56

Since we’re only discussing HoF, we can come up with whatever formula we like.

I wonder how something like HFP (Hall of Fame Points) would work where you take a player’s season WAR and add it to the difference between it and league average it and double it.

So, DrK’s 11 WAR season would 11+ 2(11-2)= 29 points.

Something like this really accentuates dominant seasons while marginalizing those seasons (but still counting them) where players are getting 100 hits and just slowly moving up the list until they get 3K hits and retire.

I was thinking of the Yount vs. Trammel discussion at FG. Guys like Molitor, Winfield, Yount, etc that just hung around for Milestones. Accumulators as some call them. A truly dominant season might be worth 5 times an average season. So a player’s 10 WAR season would be more HoF valuable than 5 2 WAR seasons at the end of a career, where a player just adds 800 more career hits or something.

This probably isn’t a new idea or even a well thought out one. But when some draw lines such as 65 WAR, some players will be rewarded for having 5 WAR combined over their last 5 seasons as opposed to retiring when their performance driops significantly.

I haven’t really looked at wWAR lately so maybe it does something similar. It just seems that we’ve got a whole host of players in the 60-68 WAR range and not really much to separate them.

I would guess that this would heavily help players like Andruw Jones that were awesome for a decade but experienced rapid decline in their performance.


#39    Matthew Cornwell      (see all posts) 2012/01/08 (Sun) @ 01:04

#38

Yeah, in effect, that is what Mike G’s system and Adam Darowski’s wWAR do.  Slight differences of course with different weights.


#40          (see all posts) 2012/01/08 (Sun) @ 19:02

Thanks, Matthew.

Wow, Jeff Bagwell and Kevin Brown.

I may need to rethink my opinion on Kevin Brown.


#41          (see all posts) 2012/01/15 (Sun) @ 14:01

Mike G. :

Are you the same Mike G. who frequents the ABPRmetric forum?

If so, I have some questions if don’t mind.


Page 1 of 1 pages


Name (required)
E-Mail (optional; WILL be published)
Website (optional)

<< Back to main


Latest...

COMMENTS

May 25 11:02
Do pitcher’s reach back for velocity when needed?

May 25 10:58
Rooting for laundry

May 25 10:14
Largest demonstration in Canadian history?

May 25 09:39
What sabermetrics is NOT

May 25 06:39
Lack of hustle during a game

May 25 02:38
NFLPA lawsuit against collusion

May 25 01:43
Neal Huntington’s best moves

May 24 17:04
Firefox, IE, or Chrome?

May 24 12:07
How to beat the shift

May 24 11:11
Incredible story