Here's Who the BBWAA Should Elect to the HOF

[Freakshow]

The Ultimate Quest for Candidates (http://www.baseball-fever.com/showthread.php?t=68815)has now revealed to all the next players that shold be elected to the Hall. For the next BBWAA election, this is the pecking order (after Rickey Henderson, of course):

1 - Bert Blyleven
2 - Tim Raines
3 - Alan Trammell
4 - Mark McGwire
5 - Dale Murphy
6 - Andre Dawson

These six ranked among the top 20 candidates for the Hall among all eligible MLB players. Just outside the top 20 we named Jim Rice as a strong candidate. Dave Parker is another to whom we gave good support that is eligible for the 2009 BBWAA election. Others we named among the top 100 candidates are Lee Smith, Tommy John, David Cone, Don Mattingly and Jack Morris.

The Hall of Merit ran a survey earlier this year that answered the same question: Who are the best candidates on the 2009 BBWAA ballot? Here are their results:

1 - Bert Blyleven
2 - Tim Raines
3 - Mark McGwire
4 - Alan Trammell
5 - Andre Dawson

Those are the only five they see as worthy of election. David Cone is the only other strong candidate for them currently. Dale Murphy, Tommy John and Lee Smith also have noticeable support in their elections.

The Hall of Merit (http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/election_results_blyleven_and_raines_are_deemed_th e_cream_of_this_crop_by_u/) also see a number of players as deserving of reinstatement to the BBWAA ballot. These players are members of the HoM that the BBWAA cut from the ballot due to less than 5% support:

Lou Whitaker
Will Clark
Keith Hernandez
Dwight Evans
Dave Stieb
Bret Saberhagen
Willie Randolph

To these seven, the BBFHOF would add Albert Belle and Dan Quisenberry.

[jalbright]

I've become intrigued by the Willie Randolph case since Bill James put Randolph and some other SS and 2B guys' win and loss shares on his (for pay) site. I've been messing around with a three part Fibonacci share score, one for career, one for top three seasons, and one for best five consecutive. Fibonacci scores are win shares * (win shares/(win shares + loss shares)) + win shares - loss shares. In other words, (win shares * winning percentage) + (win shares - loss shares). Guys who have as many win shares as loss shares aren't going to do well, and guys with less win shares than loss shares are going to be quite low. Guys with really high winning percentages will get projected forward. Anyway, for those 2B and SS, here's what I have when I total up the three Fibonacci scores for the SS and 2B I have:


SS/2B........ score
Robin Yount 704.1
Barry Larkin 648.3
Derek Jeter* 592.9
Lou Whitaker 561.0
Alan Trammell 557.7
Willie Randolph 528.3
Ozzie Smith 524.7
Nellie Fox.. 510.9
Bill Mazeroski 327.0
Omar Vizquel* 211.2
Frank White 210.0
Bobby Richardson 76.2


*--still active

Randolph falls in above two guys I like for the Hall in Ozzie and Fox, and overall I can buy this ranking (though like any other ranking, it's not perfect). I'll probably need to do some adjusting for position, because what little I have on OF/1B suggests we'd have way too many if we went down to Fox's level, just to point out one. Mariano Rivera does much better, but his score suggests we may have to have a separate standard for closers. It's preliminary because James hasn't put the results for more than about 30 players (if that) out there yet, but I'm intrigued by what I see.

[JDD]

I just don't see Alan Trammell as a Hall of Famer.

Someone make a case...

[Freakshow]

I just don't see Alan Trammell as a Hall of Famer.

Someone make a case...
Compare him to many shortstop in the Hall of Fame.

I hope you wouldn't rank any of these ahead of Trammell:

Maranville
Tinker
Bancroft
Jackson T
Aparicio

How about these?

WALLACE
SEWELL J
WARD J
Rizzuto
JENNINGS

This group is about where Tram ranks:

SMITH O
DAVIS G
CRONIN
BOUDREAU
REESE

The Hall of Merit recently ranked the HOF-eligible shortstops like this:

1. Honus Wagner
2. John Henry Lloyd
3. Cal Ripken Jr
4. Arky Vaughan
5. George Davis
6. Bill Dahlen
7. Robin Yount
8. Luke Appling
9. George Wright
10. Joe Cronin
11. Ernie Banks
12. Willie Wells
13. Home Run Johnson
14. Ozzie Smith
15. Alan Trammell
16. Pee Wee Reese
17. Lou Boudreau
18. Jack Glasscock
19. John Ward
20. Hughie Jennings
21. Dickie Pearce
22. Bobby Wallace
23. Dobie Moore
24. Dick Lundy
25. Joe Sewell

[dgarza]

This group is about where Tram ranks:

SMITH O
DAVIS G
CRONIN
BOUDREAU
REESE


I'd say he's at the bottom of that group.

The HOM SS ranking is a bit generous to Trammell.
Since they include NLers, he should be closer to #20 than #15.

[Brad Harris]

I just don't see Alan Trammell as a Hall of Famer.

Someone make a case...

If you do just a tiny bit of poking around this website you'd find a number of pretty good arguments on Trammell's behalf. To my way of thinking, it's people who base their opinions because they "just don't see" it a certain way on whom the burden of proof rests to make their case. So no, you make a case as to why he's not a Hall of Famer.

[Brad Harris]

I've become intrigued by the Willie Randolph case since Bill James put Randolph and some other SS and 2B guys' win and loss shares on his (for pay) site. I've been messing around with a three part Fibonacci share score, one for career, one for top three seasons, and one for best five consecutive. Fibonacci scores are win shares * (win shares/(win shares + loss shares)) + win shares - loss shares. In other words, (win shares * winning percentage) + (win shares - loss shares). Guys who have as many win shares as loss shares aren't going to do well, and guys with less win shares than loss shares are going to be quite low. Guys with really high winning percentages will get projected forward. Anyway, for those 2B and SS, here's what I have when I total up the three Fibonacci scores for the SS and 2B I have:


SS/2B........ score
Robin Yount 704.1
Barry Larkin 648.3
Derek Jeter* 592.9
Lou Whitaker 561.0
Alan Trammell 557.7
Willie Randolph 528.3
Ozzie Smith 524.7
Nellie Fox.. 510.9
Bill Mazeroski 327.0
Omar Vizquel* 211.2
Frank White 210.0
Bobby Richardson 76.2


*--still active

Randolph falls in above two guys I like for the Hall in Ozzie and Fox, and overall I can buy this ranking (though like any other ranking, it's not perfect). I'll probably need to do some adjusting for position, because what little I have on OF/1B suggests we'd have way too many if we went down to Fox's level, just to point out one. Mariano Rivera does much better, but his score suggests we may have to have a separate standard for closers. It's preliminary because James hasn't put the results for more than about 30 players (if that) out there yet, but I'm intrigued by what I see.

Seeing Omar Vizquel's name in an appropriate place on that list makes my day. When does the book come out? :D

[Freakshow]

I'd say he's at the bottom of that group.

The HOM SS ranking is a bit generous to Trammell.
Since they include NLers, he should be closer to #20 than #15.

Which still puts him ahead of about ten HOF shortstops:

Maranville
Tinker
Bancroft
Jackson T
Aparicio
Wallace
Sewell
Ward
Rizzuto
Jennings

Which indicates Tram would be a mid-level HOFer, not one of the dregs of the hall.

[jalbright]

Seeing Omar Vizquel's name in an appropriate place on that list makes my day. When does the book come out? :D

No word yet. I'm guessing around the beginning of next season, knowing how Bill James is with meeting deadlines (I'd guess the publisher would prefer before opening day, but my bet is Bill won't quite make that.).

[bambambaseball]

I can't believe Tim Raines is still not in the Hall of Fame!:eek:

[Greg Maddux's Biggest Fan]

The Ultimate Quest for Candidates (http://www.baseball-fever.com/showthread.php?t=68815)has now revealed to all the next players that shold be elected to the Hall. For the next BBWAA election, this is the pecking order (after Rickey Henderson, of course):

1 - Bert Blyleven
2 - Tim Raines
3 - Alan Trammell
4 - Mark McGwire
5 - Dale Murphy
6 - Andre Dawson
- -- - - -- - - -- - --- - -- -- - -- -
Lou Whitaker
Will Clark
Keith Hernandez
Dwight Evans
Dave Stieb
Bret Saberhagen
Willie Randolph

To these seven, the BBFHOF would add Albert Belle and Dan Quisenberry.

Of all these guys, I only support Blyleven and Raines for the Hall (and McGwire if they let the Roids Boys in).

[BlueBlood]

I'm only big on the Top 3. No to McGwire for PEDs. Murphys falls way short upon closer inspection, I'd now rank him 19th of our Top 20 in this project, sitting with Torre & McGwire below the line. Dawson's right at the borderline for me and I'm starting to lean "no" rather than yes.

[Los Bravos]

To my way of thinking, it's people who base their opinions because they "just don't see" it a certain way on whom the burden of proof rests to make their case. So no, you make a case as to why he's not a Hall of Famer.Well put :applaud:

[Brad Harris]

Of all these guys, I only support Blyleven and Raines for the Hall (and McGwire if they let the Roids Boys in).

I don't understand your comment here. You're only going to support a particular candidate if he's elected? Or by "letting them in" do you mean allowing them to be eligible (which they are)?

[Greg Maddux's Biggest Fan]

I don't understand your comment here. You're only going to support a particular candidate if he's elected? Or by "letting them in" do you mean allowing them to be eligible (which they are)?

I mean I only believe that of all the candidates listed, those two should actually make the HOF. There's no way Dawson, Murphy et al should get in.

Were the rest good enough to be at least considered by the BBWAA? Absolutely.

[Freakshow]

I mean I only believe that of all the candidates listed, those two should actually make the HOF. There's no way Dawson, Murphy et al should get in.

Were the rest good enough to be at least considered by the BBWAA? Absolutely.
Really? How many MLB players are currently in your version of the Hall of Fame?

[Greg Maddux's Biggest Fan]

Maybe I have a more strict criteria for my HOF than most, but I don't know the exact total. Sure Dawson and Murphy are better than some of the wankers put in by the BBWAA in the 30's and 40's, but I judge their worthiness based on today's criteria. They don't make it in my mind.

I want to see the HOF reserved for great players and not go down the road of allowing sub-great player like hockey.

This is why I don't support Curt Schilling although most people here do. I don't support Barry Larkin either although I could go either way on that one.

[Captain Cold Nose]

Maybe I have a more strict criteria for my HOF than most, but I don't know the exact total. Sure Dawson and Murphy are better than some of the wankers put in by the BBWAA in the 30's and 40's, but I judge their worthiness based on today's criteria. They don't make it in my mind.

I want to see the HOF reserved for great players and not go down the road of allowing sub-great player like hockey.

This is why I don't support Curt Schilling although most people here do. I don't support Barry Larkin either although I could go either way on that one.

The 30's? What player elected in the 30's was a bad choice? George Sisler might be the worst, and he wasn't really a wanker, just had a relatively short peak and didn't walk that often.

Or do you mean from that era?

[BlueBlood]

Probably means from that era.

On a darker note, this thread makes me sad. To think we have all of these great candidates as the fruition of our labor yet the BBWAA is rubbing its hands together in anticipation of electing Jim Rice. :noidea

[Captain Cold Nose]

Probably means from that era.

On a darker note, this thread makes me sad. To think we have all of these great candidates as the fruition of our labor yet the BBWAA is rubbing its hands together in anticipation of electing Jim Rice. :noidea

It took them 15 years to get to that point. He's at least borderline, and that's as much a nod to his reputation for when he played. So it's not a horrible choice IMO.

But, yes, there are better choices out there.

[KCGHOST]

If you do just a tiny bit of poking around this website you'd find a number of pretty good arguments on Trammell's behalf. To my way of thinking, it's people who base their opinions because they "just don't see" it a certain way on whom the burden of proof rests to make their case. So no, you make a case as to why he's not a Hall of Famer.

A nice bit of word play, but erroneous. The "don't see it" people don't need to convince you to vote against Trammell. If you want him in you have to change their minds And they are under no obligation to make it easy for you.

You see it and I see, but we need a more compelling argument to win these people over.

[Freakshow]

A nice bit of word play, but erroneous. The "don't see it" people don't need to convince you to vote against Trammell. If you want him in you have to change their minds And they are under no obligation to make it easy for you.

You see it and I see, but we need a more compelling argument to win these people over.
You make an important point: We can all do more to publicize and promote the arguements for players like Tram to the Mainstream media and fans.

At the same time, if a poster here who "just don't see it" can give us a few clues where the guy comes up short, we can better show that person what they're missing.

[Paul Wendt]

The Ultimate Quest for Candidates (http://www.baseball-fever.com/showthread.php?t=68815)has now revealed to all the next players that shold be elected to the Hall. For the next BBWAA election, this is the pecking order (after Rickey Henderson, of course):

1 - Bert Blyleven
2 - Tim Raines
3 - Alan Trammell
4 - Mark McGwire
5 - Dale Murphy
6 - Andre Dawson

These six ranked among the top 20 candidates for the Hall among all eligible MLB players. Just outside the top 20 we named Jim Rice as a strong candidate. Dave Parker is another to whom we gave good support that is eligible for the 2009 BBWAA election. Others we named among the top 100 candidates are Lee Smith, Tommy John, David Cone, Don Mattingly and Jack Morris.

There is a pattern that interests me, not big enough to be sure it is more than fortune. The disagreement may say more about the BBWAA than about us but it takes two to disagree.

The BBWAA is close to electing Jim Rice, Andre Dawson, and Bert Blyleven. Three pitchers are next in their standings, with strong showings that fall short of making them likely winners.

43% Lee Smith
42% Jack Morris
29% Tommy John

(Raines and McGwire are the others with 20% support --eight in all.)
It's true that the three pitchers are in our Top 100, but all were merely also-rans in round two, not honorable mention, not close to reaching round three. By my count there were 50 players in our round three.

In round two Lee Smith and Jack Morris finished in a 4-way tie for twelfth from the 1980s (with Don Mattingly, also on the BBWAA ballot, and Dan Quisenberry).

Tommy John finished in a 3-way tie for eleventh form the 1970s --behind Bobby Bonds, Thurman Munson, Reggie Smith, and Sparky Lyle as well as the six players who are in the Hall of Merit or in our round three.

[Paul Wendt]

Compare [Trammell] to many shortstop in the Hall of Fame.
This group is about where Tram ranks:

SMITH O
DAVIS G
CRONIN
BOUDREAU
REESE

The Hall of Merit recently ranked the HOF-eligible shortstops like this:

1. Honus Wagner
2. John Henry Lloyd
3. Cal Ripken Jr
4. Arky Vaughan
5. George Davis
[That quintet is in the Hall of Fame.]
Joe Cronin, ok, but George Davis is a stretch beyond the bounds of your license for interpreting Hall of Merit results!

--The HOM Police

[Paul Wendt]

Willie Randolph rates just above Ozzie Smith and Nellie Fox in some draft ratings by Bill James. We don't yet know how high that is because James has published ratings for only selected players.
I've become intrigued by the Willie Randolph case since Bill James put Randolph and some other SS and 2B guys' win and loss shares on his (for pay) site. I've been messing around with a three part Fibonacci share score, one for career, one for top three seasons, and one for best five consecutive. Fibonacci scores are win shares * (win shares/(win shares + loss shares)) + win shares - loss shares. In other words, (win shares * winning percentage) + (win shares - loss shares). Guys who have as many win shares as loss shares aren't going to do well, and guys with less win shares than loss shares are going to be quite low.

Beside the new loss shares ratings, there is the "Fibonacci" combination of win shares and loss shares into one number. I suppose that the combination is arbitrary, rather than grounded in reason or data, so beside reading the foundations of loss shares (do Loss Shares make sense?) I will need to see how the one "Fibonacci score" varies with alternative versions.

--
Note on mathematics and natural history.
The Bill James formula does not provide a close match for the Fibonacci series (Leonardo of Pisa, 1202) or the golden ratio between numbers that the Fibonacci series defines.

Let w be a pitcher's or player's or team's "winning percentage" between 0 and 1, defined by win shares and loss shares (w = W/(W+L)).

"guys with less win shares than loss shares are going to be quite low." --J.A.

The "zero point" where a player has Fibonacci score zero is about w=.414. For Fibonacci the zero point would be about w=.382. For example, W-L approximately 25-35 for Bill James's Fibonacci score; roughly 23-37 for Fibonacci. Forty percent wins or 24-36 is a little below zero for Bill James, a little above zero for Fibonacci.
Those winning percentages correspond to roughly 67-95 and 62-100 in 162 games.

The Fibonacci ratios --roughly .382 and .618 or 23/60 and 37/60, and 1.618 or 37/23-- recur in nature, "for example, in the arms of spiral galaxies and sunflower heads."
"Golden spiral" at wikipedia (http://en.wikipedia.org/wiki/Golden_spiral)

[Freakshow]

Joe Cronin, ok, but George Davis is a stretch beyond the bounds of your license for interpreting Hall of Merit results!

--The HOM Police
The HOM doesn't timeline; the Hall of Fame does. The first group of five shortstops is based on [my perception of] how the HOF sees them, where Tram is comparable to Davis. The next group of five is based on how the HOM sees them, where Davis leaves Tram in the dust.

Call me consistently inconsistent. Or vice versa.

[jalbright]

Willie Randolph rates just above Ozzie Smith and Nellie Fox in some draft ratings by Bill James. We don't yet know how high that is because James has published ratings for only selected players.


Beside the new loss shares ratings, there is the "Fibonacci" combination of win shares and loss shares into one number. I suppose that the combination is arbitrary, rather than grounded in reason or data, so beside reading the foundations of loss shares (do Loss Shares make sense?) I will need to see how the one "Fibonacci score" varies with alternative versions.

--
Note on mathematics and natural history.
The Bill James formula does not provide a close match for the Fibonacci series (Leonardo of Pisa, 1202) or the golden ratio between numbers that the Fibonacci series defines.

Let w be a pitcher's or player's or team's "winning percentage" between 0 and 1, defined by win shares and loss shares (w = W/(W+L)).

"guys with less win shares than loss shares are going to be quite low." --J.A.

The "zero point" where a player has Fibonacci score zero is about w=.414. For Fibonacci the zero point would be about w=.382. For example, W-L approximately 25-35 for Bill James's Fibonacci score; roughly 23-37 for Fibonacci. Forty percent wins or 24-36 is a little below zero for Bill James, a little above zero for Fibonacci.
Those winning percentages correspond to roughly 67-95 and 62-100 in 162 games.

The Fibonacci ratios --roughly .382 and .618 or 23/60 and 37/60, and 1.618 or 37/23-- recur in nature, "for example, in the arms of spiral galaxies and sunflower heads."
"Golden spiral" at wikipedia (http://en.wikipedia.org/wiki/Golden_spiral)

I've got to see more on Loss Shares myself, but what I've seen looks good, though what I've seen is obviously a tiny piece of the puzzle. The Fibonacci idea is borrowed from what Bill James did for win/loss records of pitchers in his Whatever Happened to the Hall of Fame a/k/a Politics of Glory book. The basic ideas are: .413 means nothing, .500 isn't good enough to get anywhere near the Hall, and the better you are, the more the measure pushes a guy forward. It rewards those .900 winning percentage seasons, and means even a .600 career winning percentage guy either better play a very long time or have some significantly bigger seasons than that to make it. Right now, I also have only a vague idea of where the lines need to be, and that can't be refined until I have a lot more to play with.

[jjpm74]

If he's using winning Pct as a benchmark, I don't even know where to begin on how useless this new system will be. For starters a pitcher like Bill Dinneen (http://www.baseball-reference.com/d/dinnebi01.shtml) had an amazing 5 year peak despite having a losing career record. Also, a pitcher like Terry leach who went 11 and 1 in 1987 would have then been more valuable to the Mets than Doc Gooden and Sid Fernandez of the same season if emphasizing winning pct. That's not to mention the fact that it is very common for a relief pitcher to have a losing record but be valuable to their team. Mike Marshall was 97 and 112 in his career.

[jalbright]

That's not what he's doing, at least not what I've read. What he's doing is saying every team has 486 (162 times 3) win shares plus loss shares. He determines how many of those 486 each player is responsible for, and the loss shares is determined by that number minus win shares. Guys like Mariano Rivera are 10-1, 10-0, 12-0, stuff like that; a guy at .500 might be 18-18 if he's full time, a top player might be 33-3 or 40-2. The hard part is the fielding aspect, just as with win shares. For pitchers, it's going to be largely determined by IP, for batters, the number of outs expended. There's some adjustments to get to a baseline and all, but the most likely issues will come in the fielding aspect.

My suggestion is to wait until you get a chance to read more, maybe even withholding judgment until you either buy a copy of whatever book it will be introduced in or get it out of a nearby library. Once you've done that, you can pass judgment. I'm not about to pass judgment yet, and I've at least gotten to see some of it by paying $3 a month for the right to use James' site (which now has win shares for current players, thus saving me from having to buy the annual Handbooks, which is half of what I'm paying right away).

[jalbright]

I'm not going to repeat the whole article he did on 2B, but here's some highlights to give you a taste of what Bill James is doing with Loss Shares:

Nellie Fox
Career Won-Lost 296-219 .575
Offensively 201-191 .512
Defensively 96- 28 .774
Best season: 1957, 29-4.


Bill Mazeroski
Career Won-Lost 245-211 .538
Offensively 135-203 .401
Defensively 110-8 .932
Best season: 1960, 21-10


Bobby Richardson
Career Won-Lost 129-176 .422
Offensively 90-144 .385
Defensively 39-33 .541
Best season: 1962, 21-17

Willie Randolph
Career Won-Lost 283-164 .634
Offensively: 213-136 .610
Defensively: 69-27 .720
Best season: 1980, 25-4


Lou Whitaker
Career Won-Lost 301-164 .647
Offensively: 241-127 .655
Defensively 60-37 .627
Best season: 1983, 26-8


Frank White
Career Won-Lost 222-244 .476
Offensively: 137-213 .391
Defensively 85-31 .731
Best season: 1986, 19-13

[BlueBlood]

Boy, does this ever raise a host of questions.

1) Do we get a new career number out of this? Am I supposed to subtract the losses from the wins which would give quality career guys more time to distance themselves from the spread?

Lou Whitaker: +137 Win Shares
Willie Randolph: +119 Win Shares
Nellie Fox: +77 Win Shares
Bill Mazeroski: +34 Win Shares
Frank White: -22 Win Shares
Bobby Richardson -47 Win Shares

That seems like it could be right. Only jump is that Randolph is far ahead of Fox whereas last time he was ranked two spots below the guy. Then again, this sort of addition/subtraction could just give us the career and the new peak values would be like (+11, +9, +6) and (+45) or something like that.

2) Will the new book have a Top 100 per position rank?

3) Is he going to re-evaluate 19th Century Players and actually make Win Shares for those back in 1971? Considering that he didn't do Win Shares for that period, it was a major copout when he didn't include Barnes in the Historical Abstract ranking. Seemed lazy.

[jalbright]

I have no clue what will be in what I presume will be a new book, nor do I have anything more than an educated guess as to when the book will come out. I do think I like ("winning percentage" times win shares) plus (win shares minus loss shares) as a ranking method, as it pushes high win percentage guys forward and really slows down the "compiler" type.

[Brad Harris]

I have no clue what will be in what I presume will be a new book, nor do I have anything more than an educated guess as to when the book will come out. I do think I like ("winning percentage" times win shares) plus (win shares minus loss shares) as a ranking method, as it pushes high win percentage guys forward and really slows down the "compiler" type.

I love the Fibonacci Point system! Is Bill using that on the site at all or was that your brilliant idea? I've been musing that FP might actually be the best way to "sum" the new system if it's going to be expressed as a WS-LS "record" and I have to say I'm very excited about the possibilities that this new system is going to produce.

Win Shares wasn't perfect, but it was still good - quite good - and the revised version is only going to be an improvement on that. Any statistic/method that gives us a more accurate picture of how people performed and gives tangible reasons for additional confidence in "ranking" such performances is a good thing.

Bring it on, Bill!

[jalbright]

I don't think Bill's using it this time in this way, so in that sense it's my idea. However, I'm borrowing the method from his Whatever Happened to the Hall of Fame? book's use of the method in relation to pitcher won-lost records. I just think it's got real potential to sort out the top candidates using win shares and loss shares.

[Brad Harris]

I don't think Bill's using it this time in this way, so in that sense it's my idea. However, I'm borrowing the method from his Whatever Happened to the Hall of Fame? book's use of the method in relation to pitcher won-lost records. I just think it's got real potential to sort out the top candidates using win shares and loss shares.

Completely agree. Great concept!

[Fuzzy Bear]

Maybe I have a more strict criteria for my HOF than most, but I don't know the exact total. Sure Dawson and Murphy are better than some of the wankers put in by the BBWAA in the 30's and 40's, but I judge their worthiness based on today's criteria. They don't make it in my mind.

I want to see the HOF reserved for great players and not go down the road of allowing sub-great player like hockey.

This is why I don't support Curt Schilling although most people here do. I don't support Barry Larkin either although I could go either way on that one.

This argument as to who SHOULD be in the HOF is moot, given that the standards for who gets in the REAL HOF have been pretty constant since the early or mid-forties. It was only for a brief moment that the HOF was reserved for only "the greatest of the great".

Roger Bresnahan, Jimmy Collins, Joe Tinker, Rube Waddell, Tommy McCarthy, Jack Chesbro, Johnny Evers, Frank Chance, Herb Pennock, and Pie Traynor were all inducted in the 1940s. I'm saying this to debunk the idea (which hasn't been said yet, but someone will) that it is a recent occurrance that the HOF standards have been "dumbed down". The HOF has been inducting players of this caliber, rightly or wrongly, since the 1940s. Yes, there is a gray area; there are a number of players similar to the men listed above who are not, and never will be, in the HOF. But the HOF is what it is, and what it has been.

The realistic standard for the HOF has been this: A position player who has demonstrated that he has been the best in the league at his position, or an All-Star performer during an era where his position is stocked with talent (Snider in the era of Mays) has generally been considered a viable HOF candidate, and most players who have met that standard do get inducted to the HOF eventually. There are exceptions, of course (Trammell and Whitaker being glaring exceptions).

I am of the belief that throwing out over sixty years of this standard and jacking up the standard to where only the "inner circle" greats are inducted is flat out unfair, and just plain wrong. It is unfair to any number of players, and it is also unfair to the fans of those players; players who would be outside the HOF while seeing a number (perhaps scores) of equal or lesser players inducted. This is not to say that we should induct Omar Vizquel because Phil Rizzuto is in. But it is saying that we should induct Bert Blyleven and Alan Trammell because every other player that has done what they have done are in the HOF, as well as scores of players who posted lesser accomplishments.

[jalbright]

I don't think Bill's using it this time in this way, so in that sense it's my idea. However, I'm borrowing the method from his Whatever Happened to the Hall of Fame? book's use of the method in relation to pitcher won-lost records. I just think it's got real potential to sort out the top candidates using win shares and loss shares.

Classic responded
Completely agree. Great concept!

There's still going to be issues. If you want to place Japanese and Negro League players in the mix, there's obvious difficulties. Whether you think you can overcome them is up to you. I'm really curious to see how the concept holds up for pitchers. Relief pitchers are probably still going to have to be handled separately, at least in terms of setting the bar. We will still have to deal with wartime service, even if it's only by ignoring it (not my choice). Catcher playing time may still require adjustments. 2B, 3B and SS probably will require a different standard from OF/1B/DH types--and if you go that far, you've got to deal with the mixed position guys (Pete Rose, Harmon Killebrew, Dick Allen, etc.) The 19th century guys may well still need some help due to shorter schedules. There's a lot to work out, and I can't begin to even see if it can reasonably be done until the data for the top 500-600 players of all time are out there. You need at least that many to really test the system. But I think the Fibonacci method used on Win Shares-Loss Shares may provide a good platform to work from.

[jalbright]

This argument as to who SHOULD be in the HOF is moot, given that the standards for who gets in the REAL HOF have been pretty constant since the early or mid-forties. It was only for a brief moment that the HOF was reserved for only "the greatest of the great".

Roger Bresnahan, Jimmy Collins, Joe Tinker, Rube Waddell, Tommy McCarthy, Jack Chesbro, Johnny Evers, Frank Chance, Herb Pennock, and Pie Traynor were all inducted in the 1940s. I'm saying this to debunk the idea (which hasn't been said yet, but someone will) that it is a recent occurrance that the HOF standards have been "dumbed down". The HOF has been inducting players of this caliber, rightly or wrongly, since the 1940s. Yes, there is a gray area; there are a number of players similar to the men listed above who are not, and never will be, in the HOF. But the HOF is what it is, and what it has been.

The realistic standard for the HOF has been this: A position player who has demonstrated that he has been the best in the league at his position, or an All-Star performer during an era where his position is stocked with talent (Snider in the era of Mays) has generally been considered a viable HOF candidate, and most players who have met that standard do get inducted to the HOF eventually. There are exceptions, of course (Trammell and Whitaker being glaring exceptions).

I am of the belief that throwing out over sixty years of this standard and jacking up the standard to where only the "inner circle" greats are inducted is flat out unfair, and just plain wrong. It is unfair to any number of players, and it is also unfair to the fans of those players; players who would be outside the HOF while seeing a number (perhaps scores) of equal or lesser players inducted. This is not to say that we should induct Omar Vizquel because Phil Rizzuto is in. But it is saying that we should induct Bert Blyleven and Alan Trammell because every other player that has done what they have done are in the HOF, as well as scores of players who posted lesser accomplishments.

All good points. Other ways of dealing with what HOF standards are can be to look at how many are in the HOF--and take that many from history. Those above that line should be in, and those below out. Or, separate out pitchers as one category, and then hitters as another--and you can divide the hitters by eight and thus figure out how many guys should make it at each position. Then induct guys who meet those standards forward. Any of the methods Fuzzy or I have presented deal with what the HOFs standards truly are, not some idealized notion of what they should be. The methods we propose all are also aimed at preventing the kind of historical unfairness Fuzzy so eloquently points out would result from changing the game. We don't have to use the clear mistakes as our guide, because that's disastrous, too--but we can't ignore what has happened and substitute our own notions of what the Hall "should have" been from the get-go as our guidepost for the future.

[Freakshow]

I don't think Bill's using it this time in this way, so in that sense it's my idea. However, I'm borrowing the method from his Whatever Happened to the Hall of Fame? book's use of the method in relation to pitcher won-lost records. I just think it's got real potential to sort out the top candidates using win shares and loss shares.
The Fibonacci number system is an amusing mathematical niche. Remember, though, that as James used it, it turnes out to be nothing more than Wins Above .414. That's a bit higher than actual Replacement Level (or the level of Freely Available Talent) is usually estimated to be.

Once we get W-L Shares for everyone it will be easy to set the replacement level wherever one prefers. This is the nub of most of our disputes over value on this site: Where does value begin? Should Replacement Level be set differently for different positions/at different times? What value counts towards the Hall of Fame? Is value multiplied if it is accumulated in larger amounts per season (the old Koufax-Sutton conundrum; or Beckley-Chance, if you prefer)? Each of these questions deserves its own thread.

[Paul Wendt]

The Fibonacci number system is an amusing mathematical niche. Remember, though, that as James used it, it turnes out to be nothing more than Wins Above .414. That's a bit higher than actual Replacement Level (or the level of Freely Available Talent) is usually estimated to be.
I think you mean two times wins above .414, as W-L is two times wins above .500

No, it's quadratic, not linear. The zero point is .414 but "Fib" gives parabolically more credit as w increases from .414 to 1.000. In contrast W-L is linear. It's zero point is .500 and the extra credit is equal for every extra win --every conversion of one loss to one win.

Using W-L, 25-35 one year and 45-15 the next year is worth the same total score as 35-25 and 35-25.
Using "Fib", 25-35 and 45-15 is better than 35-25 and 35-25. There is a built-in bonus for peak seasons even if scattered and a built-in discount for consistency --if Fib is calculated at the season level before summing the career.

I agree that most people who think about replacement level players as the zero point will find .414 too high for the zero point. Also anyone who thinks consistency from year to year is worth extra credit should "hate Fibbing" at the season level. ;)

But some people will find their own ideas about the proper zero point and their own ideas about extra credit for peak seasons built into the Fib score. Others may find that some alternative to the particular formula captures their own ideas of replacement (zero point) and peak bonus. ALmost anyone who finds that the win shares and loss shares ratings separately make sense will be able
(with some tinkering) to come up with a personally satisfactory way of combining them.

[jalbright]

A 12-0 season is 24 points under a Fibonacci approach,
A 9-3 season is 12.75 points under a Fibonacci approach,
A 6-6 season is 3 points under a Fibonacci approach
A 3-9 season is below zero under a Fibonacci approach.

So it's not linear. It's meant to push high winning percentages up.

[Brad Harris]

Classic responded


There's still going to be issues. If you want to place Japanese and Negro League players in the mix, there's obvious difficulties. Whether you think you can overcome them is up to you. I'm really curious to see how the concept holds up for pitchers. Relief pitchers are probably still going to have to be handled separately, at least in terms of setting the bar. We will still have to deal with wartime service, even if it's only by ignoring it (not my choice). Catcher playing time may still require adjustments. 2B, 3B and SS probably will require a different standard from OF/1B/DH types--and if you go that far, you've got to deal with the mixed position guys (Pete Rose, Harmon Killebrew, Dick Allen, etc.) The 19th century guys may well still need some help due to shorter schedules. There's a lot to work out, and I can't begin to even see if it can reasonably be done until the data for the top 500-600 players of all time are out there. You need at least that many to really test the system. But I think the Fibonacci method used on Win Shares-Loss Shares may provide a good platform to work from.

All of the issues you mentioned are important in making player comparisons, but few (IMHO) really ought to be addressed by a uniform statistical method. Taking wartime service into account is something best judged by each individual doing the rankings. In the past, I've simply felt that years lost in that way should be counted as some percentage of the difference between the seasonal values of the first full season immediately preceeding and succeeding the missing years. Whatever that percentage is will differ depending on the individual doing the assessment. I've known people who give 0% credit up to people who give 100% credit of whatever that average is. Win shares was a great tool for expressing that missing average, but there's no way that win shares itself should have accounted for that kind of omission. The fewer subjective elements in an equation, the more accurate/useful it is. Adding the subjective elements is the prerogative of the individual using the statistical tool and really ought not be part of the composition of the tool itself, if this is making any sense?

As for negro league and japanese players, there are two ways to do that, as I see it. Firstly, we use the most complete statistical data available about the context in which those players played and we can attempt to derive win/loss shares using known quantities for those leagues alone. That would, at least, give us a fair estimate with which to compare negro/japanese players with each other, which is helpful. A direct comparison of such numbers would be pointless ("Josh Gibson had 450 win shares. Johnny Bench had 350 win shares. Therefore, Gibson was better.") You're not comparing like leagues. So here's where the processes diverge. One could, if one had the data with which to do it, "discount" the "other" league statistics based on league quality assumptions/data (which is, of course, probably including a subject element once again) in order to say something akin to "the 1932 Negro National League was 94% as good as the 1932 National League". Along those lines anyway...I think you get the picture.

The other method is, of course, to make assumptions about the percentage of negro leaguers who would have cracked the major leagues in their respective years, somehow derive estimates of who would see what kind of playing time (and make correlated assumptions about the playing time of actual major leaguers whose jobs would have been effected). This "integrated league" approach, however, is enormously unstable and, in my mind, a gigantic undertaking whose results won't carry enough gravitas to make it worthwhile in a serious analysis. More practical would be to look at the negro/japanese players and their contemporary major league counterparts and attempt to guesstimate how many "top" players would have existed at one particular time and in what order those players might have fallen. For example, Josh Gibson played at the same time as Mickey Cochrane, Gabby Hartnett and Bill Dickey. Would a major league version of Josh Gibson simply been an addition to the golden age of catchers or would Gibson's performance in the majors have affected that of the other stars. If Gibson led the AL in home runs in 1935, for example, it means Jimmie Foxx would not have. We can't simply superimpose a plethora of "other" league stars over an era of major leaguers and assume the quality of the league as a whole wouldn't rise (and thus fewer people would stand out from the pack) as a result. That's counter-intuitive.

I hope at least some of the above is clear enough to be debated. :laugh