Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Monday, December 26, 2005

2006 BTF Hall of Fame Ballot Discussion

We’ll have one week of discussion and then the ballot thread will be posted next Monday (the election will end on Jan. 9).

The eligible candiates are: Rick Aguilera*, Albert Belle*, Bert Blyleven, Will Clark*, Dave Concepcion, Andre Dawson, Gary DiSarcina*, Alex Fernandez*, Gary Gaetti*, Steve Garvey, Dwight Gooden*, Rich Gossage, Ozzie Guillen*, Orel Hershiser*, Gregg Jeffries*, Tommy John, Doug Jones*, Don Mattingly, Willie McGee, Hal Morris*, Jack Morris, Dale Murphy, Dave Parker, Jim Rice, Lee Smith, Bruce Sutter, Alan Trammell, Walt Weiss*, and John Wetteland*.

Just to make sure everyone knows the rules, as we did last year, each ballot should follow BBWAA rules. That means you can have up to 10 players on your ballot in no particular order. Write-in’s are acceptable to add to your ballot, but as in reality, they wont count.

* 1st-year candidates.

John (You Can Call Me Grandma) Murphy Posted: December 26, 2005 at 05:46 PM | 147 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 2 of 2 pages  < 1 2
   101. Daryn Posted: January 05, 2006 at 05:13 PM (#1807638)
The contrast between Rice's '78, whcih at the time and since was seen as a historic season and Lynn's '79, which wasn't noticed much even at the time (Yaz's 300th hit and 400th HR were the big stories) is extraordinary. By OPS+ Lynn's was much the better season. Anyone know why nobody noticed it?

I would guess that it got a bit lost because it was virtually identical (in terms of conventional statistics) to his teammate Rice's season. That, of course, does not explain why the public was swooning all over Don Baylor when his teammate Bobby Grich was having a statistically similar season with little notice.

That was my third season watching baseball closely and I remember think Rice was the best hitter in the game and that Lynn was fragile even though Lynn didn't really show his consistent fragility until much later.
   102. Chris Cobb Posted: January 05, 2006 at 05:15 PM (#1807645)
By OPS+ Lynn's was much the better season. Anyone know why nobody noticed it?

Before the days of OPS+ and even OBP, Lynn's 1979 season was probably overshadowed by Rice's 1979 season, which was nearly as good as his 1978 season.

In 1978, Rice passed a bunch of counting benchmarks: 200 hits, 40 home runs, 400 total bases, 120 runs scored, 130 rbis. He hit great, he didn't lose at bats to walks, and he played every game..

In 1979, he wasn't as good and he missed a few games, but he still had 200 hits, 39 home runs, 130 rbis, and he hit .325 / .596 (and a 154 OPS+)

In 1979, Fred Lynn had a 176 OPS+, but he played only 147 games, had only 177 hits, and 122 rbis. He matched Rice in home runs with 39, and he hit for higher batting and slugging averages, .333 / .637.

He probably just didn't look to most people like he was any better than Rice in 1979, and he missed the big counting benchmarks that set Rice's 1978 season apart. The fact that Lynn never had another season nearly as good probably gives his 1979 season something of fluke status, though what was a fluke was not that Lynn was that good, but that he stayed healthy . . .
   103. Chris Cobb Posted: January 05, 2006 at 05:34 PM (#1807681)
And then not a single one of these ten guys did it more than once (in the period Paul defines). Not Yaz, not Reggie, not Brett, whose careers are fully or at least for all intents and purposes encompassed by the period in question.

In the period _and_ the league Paul defines. Allen slugged .632 in the NL in 1966.

In many ways, what this list shows is that the AL was the weaker league during this period. Hitting conditions may have been somewhat different in the NL, but here's the comparable list:

NL players slugging .600+, 1963-1992

1964 .607 Mays
1965 .645 Mays
1966 .632 Allen
1969 .656 McCovey
1969 .607 Aaron
1970 .612 McCovey
1971 .669 Aaron
1971 .628 Stargell
1972 .606 B. Williams
1973 .646 Stargell
1977 .631 Foster
1979 .613 Kingman
1980 .624 Schmidt
1981 .644 Schmidt
1989 .635 Mitchell
1992 .624 Ba. Bonds

16 seasons by 11 players, with Mays, Allen, Aaron, McCovey, Stargell, and Schmidt doubling up within the period.
   104. DavidFoss Posted: January 05, 2006 at 05:49 PM (#1807699)
He probably just didn't look to most people like he was any better than Rice in 1979

FWIW, Lynn was 4th in MVP voting in 1979. Rice was 5th. (Baylor, Singleton and Brett were 1-2-3). The batting title used to be a bigger deal back then. (Of course, today, the Sabermetric Triple Crown (AVG/OBP/SLG) sticks out.)
   105. DavidFoss Posted: January 05, 2006 at 05:55 PM (#1807707)
So I don't see #98 as much of an argument against Rice, though I recognize that there are such arguments.

That's a pretty good spin. I think #98 was a response to #76, though. Rice did indeed have a nice peak, but it wasn't anywhere near as historic as you had earlier claimed.
   106. Dr. Chaleeko Posted: January 05, 2006 at 06:27 PM (#1807762)
A quick check on the influence of park among .600 sluggers. Park factor bb-ref, and aSLG is the SLG adjusted by the PF. DIFF is the difference between the two SLGs.
AL .600 sluggers
YEAR  SLG NAME         PF  aSLG  DIFF
-----------------------------------------
1963 .605 Mantle       99 .611  +0.06
1964 .606 Powell      100 .606   0.00
1966 .637 Robinson     98 .650  +0.13    
1967 .622 Yaz         108 .572  -0.50
1969 .608 Jackson      96 .620  +0.12
1972 .603 Allen       102 .591  -0.12
1978 .600 Rice        111 .534  -0.66
1979 .637 Lynn        106 .599  -0.38
1980 .664 Brett       101 .657  -0.07
1987 .618 McGwire      93 .661  +0.43
1987 .605 Bell        102 .593  -0.12

NL .600 Sluggers
YEAR  SLG NAME         PF aSLG   DIFF
-------------------------------------
1964 .607 Mays        102 .595  -0.12
1965 .645 Mays        103 .626  -0.19
1966 .632 Allen       100 .632   0.00
1969 .656 McCovey      98 .669  +0.13
1969 .607 Aaron       100 .607   0.00
1970 .612 McCovey      99 .618  +0.06
1971 .669 Aaron       106 .629  -0.40
1971 .628 Stargell    100 .628   0.00
1972 .606 B. Williams 111 .539  -0.67
1973 .646 Stargell     97 .665  +0.09
1977 .631 Foster      102 .618  -0.13
1979 .613 Kingman     110 .552  -0.61
1980 .624 Schmidt     106 .587  -0.37
1981 .644 Schmidt     104 .618  -0.36
1989 .635 Mitchell     97 .654  +0.19
1992 .624 Ba. Bonds    99 .630  +0.06




I may be oversimplifying the application of the PF, but still, I don't care so much about the 400 TB if a good chunk of them are park puffs.
   107. Dr. Chaleeko Posted: January 05, 2006 at 06:28 PM (#1807766)
Trying again after fogetting the <pre> problem

AL .600 sluggers
YEAR  SLG NAME         PF  aSLG  DIFF
-----------------------------------------
1963 .605 Mantle       99 .611  +0.06
1964 .606 Powell      100 .606   0.00
1966 .637 Robinson     98 .650  
+0.13    
1967 .622 Yaz         108 .572  
-0.50
1969 .608 Jackson      96 .620  
+0.12
1972 .603 Allen       102 .591  
-0.12
1978 .600 Rice        111 .534  
-0.66
1979 .637 Lynn        106 .599  
-0.38
1980 .664 Brett       101 .657  
-0.07
1987 .618 McGwire      93 .661  
+0.43
1987 .605 Bell        102 .593  
-0.12

NL .600 Sluggers
YEAR  SLG NAME         PF aSLG   DIFF
-------------------------------------
1964 .607 Mays        102 .595  -0.12
1965 .645 Mays        103 .626  
-0.19
1966 .632 Allen       100 .632   0.00
1969 .656 McCovey      98 .669  
+0.13
1969 .607 Aaron       100 .607   0.00
1970 .612 McCovey      99 .618  
+0.06
1971 .669 Aaron       106 .629  
-0.40
1971 .628 Stargell    100 .628   0.00
1972 .606 B
Williams 111 .539  -0.67
1973 .646 Stargell     97 .665  
+0.09
1977 .631 Foster      102 .618  
-0.13
1979 .613 Kingman     110 .552  
-0.61
1980 .624 Schmidt     106 .587  
-0.37
1981 .644 Schmidt     104 .618  
-0.36
1989 .635 Mitchell     97 .654  
+0.19
1992 .624 Ba
Bonds    99 .630  +0.06 
   108. Brent Posted: January 06, 2006 at 05:18 AM (#1808741)
Dr. C,

I think your adjustments are too severe. The park factor is designed for adjusting runs and statistics that are proportional to runs (such as OPS). Because of the nonlinear relationships of other statistics to runs, they can't use the same park factor.

BBRef has a discussion of how it park adjusts OPS+ on its glossary page. It's pretty complicated, but my suggestion for park adjusting SLG is to take the ratio of bbref's *lgSLG (.403 for Boston in 1978) to league slugging (.385 for the 1978 AL) as your park factor. That is, for Jim Rice 1978 I'd use

.600 / (.403/.385) = .600 / 1.047 = .573.
   109. Brent Posted: January 06, 2006 at 06:06 AM (#1808784)
In # 95 John DiFool2 asked about Will Clark:

He's clearly a better hitter than just about any position player on the ballot, even if you
penalize him for a short peak. I want someone to convince me he belongs-demonstrate exactly
how clearly his can be separated from the Rice/Mattingly/Parker/Bell/Andre pack.


On my ballot I said that Clark meets the "greatest common denominator criterion"--that is, just about everyone similar to him who is eligible has been inducted in the Hall of Fame. When I wrote that, I was thinking of his top 3 and top 7 seasons as measured in win shares, which are

44-37-34-28-27*-25-25
(* adjusting WS for 1994 to 162 games).

This relatively high peak, combined with a moderate-length career (1976 games) puts him in a category of players that have generally not been passed over by Cooperstown.

The only post-1900 HOF-eligible player that I'm aware of whose 3-year and 7-year peaks are comparable (actually better) and hasn't been enshrined is Dick Allen:

42*-41-35-33-32-29-29
(* adjusting WS for 1972 to 162 games).

Rice/Mattingly/Parker/Dawson/Murphy all have fairly good peaks according to win shares, but still significantly below Clark's (especially for top 3 seasons).

However, looking at the batting statistics on bbref, or looking at the WARP statistics on bp, Clark doesn't look nearly as impressive. I mean he still looks very good (OPS+ of 175-160-153-152-150-145-140 in his top seasons), but his stats don't tower above those of the other guys the way they do with win shares. Is this one of those win shares anomalies? Or is WS actually capturing something real that doesn't show up in OPS+, EQA, and similar statistics?

Bill James was aware of this issue and actually devoted a page of his Win Shares book to a discussion of Clark's (best) 1989 season. He wrote "When I developed the Win Shares system I was surprised to see Clark's 1989 season with the bat actually ranks as not only better than Mitchell's [who won the MVP], but actually better than any other major league season of the 1980s. Why? Basically Clark starts out with MVP-type numbers, and everything you look at tends to make him look a little bit better."

James cites the following:
- He grounded into only 6 double plays. (Throughout his career, Clark's GIDP rates are very low; this is important and is not captured by OPS or by WARP/EQA.)
- In 1989 he was 8 for 11 as a base stealer. (On the other hand, Clark's stolen base record was not a net plus over his full career)
- 1989 NL was a pitcher's league, with one of the lowest R/G of the last 25 years.
- San Francisco was a pitchers' park with a very low park factor.
- For recent years, win shares incorporates information on batting performance in "run sensitive situations." In 1989 Clark hit .389 with men in scoring position, and 13 of his 23 home runs were hit with runners on base.

Checking for win-shares anomalies, I see that San Francisco did not exceed its Pythagorean expectation in 1989 (differences from Pythagorean expecations are frequently cited as a source of anomalies with win shares). IMO, I think the reasons that Bill James cites are all examples where win shares is providing us with additional, useful information about Clark's performance. Unfortunately, the data on batting with runners in scoring position and on home runs with runners on base are apparently proprietary and not freely available, so we can't do a similar examination of the other seasons in Clark's career.

After looking some more at the data, I'm a little more cautious about Clark than I was when I posted my ballot, but I still think that Clark's high win-shares peak is legitimate. Compared to someone like Bill Terry (a borderline Hall-of-Famer, rather than an outright mistake) Will Clark looks really good. I would still vote for him.
   110. Joey Numbaz (Scruff) Posted: January 06, 2006 at 07:00 AM (#1808838)
The Clark/Terry comparison is a great one. I'd never thought of that. Clark had a slightly longer career (basically 2 seasons) and is quite similar rate wise - compared to park adjusted league averages:

Clark +.040/+.052/+.091
Terry +.052/+.046/+.094

From the 4 years of numbers on BB-ref, it doesn't look like Terry hit into too many DPs either.

Terry hit for a higher average, but Clark drew more walks. Both had very good power. Both are 6' 1", 190-200 lb., Giant 1B, with 5-letter last names from the Southeast too :-) Clark is 41.9 on the HoF standards, and Terry 42. Of course on the monitor, Terry hits 169 and Clark just 83.5, entirely because Terry played in an offensive explosion while Clark played in pitchers parks and his prime was in a pitcher's era.

Clark had a longer career and a higher peak, with similar overall rate stats. I don't see how one could rate Terry higher.

None of that is using Win Shares either, which make him look even better than the raw numbers. He's a fairly easy pick once you start digging, unless you think Bill Terry was an awful mistake.
   111. Paul Wendt Posted: January 06, 2006 at 07:12 AM (#1808840)
677 at bats
677 at bats

So I don't see #98 as much of an argument against Rice, though I recognize that there are such arguments.

#98 is literally an argument against the so-called historic peak achieved by Jim Rice.

Its force concerning his HOF candidacy depends on context: slugging is his calling card, etc.

Not Yaz, not Reggie, not Brett, whose careers are fully or at least for all intents and purposes encompassed by the period in question.

But Rice is known for hits and home runs. The typical Jim Rice fan would be shocked that Yaz, Reggie, and Brett achieved greater career highs in slugging average.

677 at bats
677 at bats
   112. Dr. Chaleeko Posted: January 06, 2006 at 03:15 PM (#1809011)
Re W. Clark/Terry

Completely agreed, and in fact, I'll take it a step further. Clark is also as good or better than Sisler. In fact his career parrallel's Sisler's in the sense that they both came out with bats a-blazing and then had long declines where they were just in the pack at their position.

Actually that undersells Clark who had a much better rest-of-his-career than did Sisler.

Anyway, I personall see Clark as just a little better than Sisler and somewhat better than Terry, but it's only my opinion.
   113. sunnyday2 Posted: January 06, 2006 at 04:03 PM (#1809057)
The problem with this discussion is that Sisler and Terry were arguably among the top 3-5 1B of their time. Clark, well, actually you could make the same argument but others might simply note that he is one of about 10 outstanding 1B of his era.

IOW comparing the numbers directly misses the point that Clark has significant advantages over Sisler and Terry in terms of nutrition, conditioning, coaching from an early age, technique, etc. etc. And many other players had the same advantages. If all I did was look at the raw numbers and if I didn't know when these guys played, I could easily elect 10 1B from 1980-2005 and 3 from 1915-1945.

That would not be the HoM of course. That would be Bill James' rankings or just simply the best 220 players of all-time. Our goal is not to elect the best 220 players overall but to recognize all eras "equally" or shall I say appropriately.

When we get into the backlog, this argument says to just elect modern players--Richie Ashburn over Pete Browning or Fielder Jones, Will Clark and...and...and...and... over George Sisler, and so on. It is not unreasonable to do any one of those things, but we are in danger of dismissing the first 100 years of MLB out of hand. I hope that doesn't happen.
   114. Dr. Chaleeko Posted: January 06, 2006 at 05:01 PM (#1809122)
Two quick notes on Andre Dawson.

1) Walked 589 times. 143 were intentional. Unintentional walk rate is about 1 for every 23 PAs.

2) GIDPed 217 times, 52nd all time.
   115. DanG Posted: January 06, 2006 at 05:16 PM (#1809142)
But Rice is known for hits and home runs. The typical Jim Rice fan would be shocked that Yaz, Reggie, and Brett achieved greater career highs in slugging average.

This is probably true. But if you look at more than a single year to define peak, Rice's slugging average looks a lot more "historic".

In the period defined, 1963-1992, Rice was the only AL player with three qualifying seasons slugging over .565, topping .590 in 1977-78-79. Making this more impressive was doing this for 2144 plate appearances.

Here are the six AL players who slugged over .550 in a three year period 1963-92, minimum 1500 PA:

.596 77-79 Rice
.575 79-81 Brett
.567 85-87 Mattingly
.562 66-68 F.Robinson
.558 68-70 F.Howard
.555 88-90 Canseco

I'm not a big fan of Rice, but there really are clear reasons why the average fan saw his peak as historic.
   116. Chris Cobb Posted: January 06, 2006 at 05:25 PM (#1809156)
On Terry vs. Clark

Saying that Terry is among the 3-5 best 1B of his time and that Will Clark is among the top 10 of his time and that therefore Terry is better is off target in two ways. First, there were, on average, 28 major league teams in Clark’s era to 16 in Terry’s, with a correspondingly larger player pool, so we should expect the 4th best 1B of the 30s to be equivalent to the 7th best 1B of the 90s, acknowledging that both eras were boom times for slugging 1Bmen.

Second, to say that Clark is among the 10 best first baseman of his era and leave it at that simply underrates him. Here’s a way of finding equivalent comparative cohorts for each player. Find the midpoint of each’s career, and compare them to all the first basemen who were active at that time.

For Terry, that’s 1929/30. Active first basemen as good as or better than Terry include

Gehrig, Foxx, Suttles, Sisler

With input from the HoM, I’d place Terry at #4 in this group: ahead of Sisler but behind Gehrig, Foxx, and Suttles, though I have Sisler ahead in my own rankings. (Greenberg had a cup of coffee in 1930 but didn't really break in to the majors until 1932, so I won't include him in Terry's cohort.)

For Clark, the midpoint is 1993. He overlaps, therefore, with the following group:

Murray, Thomas, Bagwell, Palmeiro, McGwire, McGriff, Grace, Olerud, Mattingly, Delgado, Vaughn, Tino Martinez

Where does he place within this group? By WARP3 (necessary because of DH adjustment), here’s a ranking of the players by adding up their rank order positions in career value and top 5 consecutive seasons in the group:

125.3 (3) 52.9 (1) Bagwell
114.7 (4) 52.5 (2) Thomas
131.1 (2) 49.5 (5) Murray
137.7 (1) 46.8 (7) Palmeiro
101.6 (7) 50.2 (4) Clark
101.8 (6) 47.2 (6) McGwire
111.4 (5) 45.7 (8) Olerud
92.4 (10) 51.3 (3) Mattingly
96.0 (8) 42.8 (9) McGriff
93.5 (9) 39.4 (11) Grace
73.7 (12) 41.8 (10) Delgado
76.7 (11) 38.6 (12) Martinez
56.3 (13) 36.7 (13) Vaughn

Will Clark, by this measure, is at the top of a tightly bunched 5-8 group that also includes Mark McGwire, John Olerud, and Don Mattingly, which could easily shift around depending on how one weights peak vs. career.

In context, it looks to me like Will Clark has an excellent argument to be considered as good as or better than Bill Terry. Terry and Sisler for the 1920s-1930s correspond to Clark, McGwire, Olerud, and Mattingly for the 1980s-1990s. They are not all time greats (despite McGwire’s apparent stature) but they have good chances of becoming HoMers.
   117. sunnyday2 Posted: January 06, 2006 at 05:44 PM (#1809182)
>First, there were, on average, 28 major league teams in Clark’s era to 16 in Terry’s, with a correspondingly larger player pool, so we should expect the 4th best 1B of the 30s to be equivalent to the 7th best 1B of the 90s, acknowledging that both eras were boom times for slugging 1Bmen.

This to me is exactly wrong. The caliber of players has nothing directly to do with how many major league teams there are. This implies that if there had still been 16 teams in 1985-1995 that Will Clark would somehow have been a different player than who he was. The pool is the pool, opportunities don't change that.

Secondly, I acknowledged that you could make an argument that Clark ranks just as highly in his time as Terry or Sisler did in his. The real point is whether you want to elect 10 1B from 1980-2005 and 3 from 1915-1935. If all you look at is raw numbers--and I mean to include stuff like WS and WARP and EQA etc. as raw numbers, not just HR and BA and RBI.

Insert the names of the 10th versus the 3rd as you see fit. On your list we are talking McGriff or Delgado, perhaps, and you and I both know that there are plenty of folks out there who would vote for McGriff ahead of Sisler or Terry or Beckley or probably even Hank Greenberg (if you look at career HR and RBI, e.g.), and in fact some of those will probably be joining the HoM voters in time to do that.
   118. Dr. Chaleeko Posted: January 06, 2006 at 05:55 PM (#1809196)
On Will Clark

Win Shares has him as
-the NL’s best player in 1988 and 1989,
-the league’s best first baseman in 1991,
-the best player on his team in 1987, 1988, 1989, 1991, 1992, 1994, and 1995

Over the period covering his career (1986-2000) his total WS are second among 1Bs to Mark McGwire's (334 to 331).

Over the five year period 1986-1990, Clark is MLBs' best 1B by WS, ditto 1987-1991, 1988-1992, 1989-1993.

Over the eight year period 1986-1993, he was the best 1B in baseball, ditto the eight year periods 1987-1994, and 1988-1995.

Over the ten year periods 1986-1995, 1987-1996, and 1988-1997 he is the best 1B in baseball.

Etc, etc etc...

Of course it's slice-'n-dice, but on the other hand, it's not like there's not plenty of competition. McGwire, McGriff, and Palmeiro are exact contemporaries, and Murray wasn't quite half way through is career.

In his time Clark exhibited dominance over his positional competition, as you'd expect a HOFer to.
   119. Chris Cobb Posted: January 06, 2006 at 06:32 PM (#1809256)
This to me is exactly wrong. The caliber of players has nothing directly to do with how many major league teams there are. This implies that if there had still been 16 teams in 1985-1995 that Will Clark would somehow have been a different player than who he was. The pool is the pool, opportunities don't change that.

The number of opportunities will generally be proportional to the size of the talent pool. As the population increases the number of ML teams it can support both financially and in terms of talent also increases. If that weren't the case, we'd see huge swings in the quality of play, which we do not observe as we move from 1940-2005. The HoM slots for the whole of the project are set up to be proportional to the number of major leagues teams (unless I am greatly mistaken), so if this idea is "exactly wrong," then the whole project is on highly queestionable footing. I take it as axiomatic for the project that we should expect to elect about twice as many players from a 30-team league as from a 16-team league.
   120. sunnyday2 Posted: January 06, 2006 at 07:01 PM (#1809299)
Only in a general sense. Population change 1960-62 is completely unrelated to opportunity change.
   121. DavidFoss Posted: January 06, 2006 at 07:25 PM (#1809343)
The HoM slots for the whole of the project are set up to be proportional to the number of major leagues teams (unless I am greatly mistaken)

This is true.

Only in a general sense. Population change 1960-62 is completely unrelated to opportunity change.

This is hyperbole. Changes in league size don't track the talent pool *exactly* to the date, but they do correlate with changes in the size of the talent pool. Integration of African-Americans started in 1947 and was in full swing by the mid-50s. Latin Americans started entering MLB en masse in the late 50s and early 60s. Does anyone think the 1964 Cardinals played in a weaker league than the 1946 Cardinals due to expansion? Integration offsets that. In fact, the the HOM was inducting NeL's almost as if it was a third major league a 25% increase in MLB from 1960-62 doesn't fully make up for the "integration contraction" that occurred ten years earlier.
   122. sunnyday2 Posted: January 06, 2006 at 08:14 PM (#1809425)
If the number of ML teams tracks the talent pool it is most certainly by accident. When did MLB actually discuss the talent pool in making expansion decisisions?

Yes expansion tracks population and wealth of "major markets"--which is to say those markets that can support a ML team emerge as a result of population growth and increases in per capita wealth.

But "markets" that commercially support MLB are hardly the same markets that produce players, for the very reasons you list. e.g. African-Americans (within the U.S.) and Latin Americans (outside the U.S.) have produced large expansions of the talent pool while providing very very little of the markets that support MLB, or at least this was true of African-Americans at the time integration took place and it continues to be true of foreign "talent markets" today.

The tracking of the commercial market supporting MLB and the talent market feeding MLB are as I said related "in a general sense." But, again, this is largely accidental as there is really no cause and effect between the two.

So what does it mean? I understand that the rate at which we are electing HoMers is based on the numbers of teams, and I understand the reasoning for that. But it is a large leap of faith to assume that the number of teams tracks the quality of the talent pool and that the 4th 1B today is equal to the 7th then. This is an a priori assumption, not susceptible to any evidentiary hearing (I mean we all know Will Clark could step off our time machine [that is to say, our time{line} machine] and outplay Dan Brouthers and Roger Conner but, again, we don't know what this means, or at least I don't). This is because of what I just said: 1) that growth of the markets and the talent pool are independent events--yes, both driven by population growth, but of different populations; and 2) because great players are random outliers anyway.

It is tautological, then, to say that we will elect players in a certain chronological pattern and that the pattern then should drive or justify how we make out our ballots.

The 1946 versus 1964 comparison is true enough, but its meaning for making out a ballot is not obvious to me. If 1946 vs. 1964 is decisive, then I give you 1878 vs. the 1880s.
   123. sunnyday2 Posted: January 06, 2006 at 08:17 PM (#1809429)
>>Only in a general sense. Population change 1960-62 is completely unrelated to opportunity change.

In other words, to say that the above is not hyperbolic and then to exclaim...

>Changes in league size don't track the talent pool *exactly* to the date, but they do correlate with changes in the size of the talent pool.

...is simply asking whether the glass half full or half empty. They track only in a general and they don't track exactly is a difference without a distinction. And neither one answers any meaningful question about who to vote for.
   124. Chris Cobb Posted: January 06, 2006 at 08:32 PM (#1809456)
...is simply asking whether the glass half full or half empty. They track only in a general and they don't track exactly is a difference without a distinction. And neither one answers any meaningful question about who to vote for.

Changes in league size, which track in a general way with the size of the talent pool, don't answer _all_ meaningful questions about who to vote for, but it answers some questions. For example, if player A is the 30th best player of the period 1900-1920 according to one's system and player B is the 30th best player of the period 1970-1990 according to one's system, that is evidence in favor of ranking player B ahead of player A. There may be other evidence, as for example, Player A may be so far ahead of Player B in a head-to-head comparison that one decides that Player A's head-to-head advantage outweighs Player B's contextual advantage. But if the two players are similar in a head-to-head comparison, the 30th best player from the larger pool ranks ahead of the 30th best player from the smaller pool.

30 is chosen deliberately, because that's how far down in the rankings for most 20-year periods that we have to go to get to the borderline candidates who stay on the ballot from year to year.
   125. sunnyday2 Posted: January 06, 2006 at 09:35 PM (#1809547)
Chris, I would say your example doesn't respond to your set-up, that it "answers some questions." Rather it only very partially answers questions. 30th vs. 30th is purely abstract. What is not abstract is the actual choice that needs to be made, e.g. Will Clark or Jake Beckley? To me, if they rank the same in their respective pools, your "quasi-pseudo-sort of-timeline" theory puts one check mark in the Clark column, but that doesn't finally get me very far down the road to a final choice.

I will allow that the talent pool argument carries more weight the further down into the pool you go. Because greatness is an outlier, for #1 versus #1 I wouldn't use the "timeline" at all, nor probably all the way down to #10 or so. After that we are talking less extreme outliers and the pool begins to be a more credible affect.

Even so all of this (discussion) demonstrates confusion (not necessarily yours) between the talent pool argument and the opportunity argument. If there are more teams and more WS available, does that by itself suggest that we should elect more players from that period? Or are we really appealing to the pool? It is not always clear what is being claimed.

On another point, I could even argue that more opportunity advantages modern players, who stay in the game longer simply because there are more jobs. I'm not sure we shouldn't penalize some modern players in the sense that we should put them into a hypothetical 16 or 24 team environment and ask how long they would have really been in the MLs. This is what we have done for the NeLers, modeled them into a neutral environment. But we have not modeled modern players into that same enviornment.

There has been talk that shorter versus longer seasons and timelines constitute double jeopardy for the old-timers. Expansion makes that triple jeopardy. My point being that all the talk about the talent pool and the opportunity pool seems to some degree to be designed (a priori) to give some of our recent favorites yet another leg up. Maybe it's valid to talk about the talent pool and the opportunity pool, but if so let's ask what we can do to maintain a legel playing field or whether we care whether the playing field is level or not.
   126. DavidFoss Posted: January 06, 2006 at 10:16 PM (#1809617)
There has been talk that shorter versus longer seasons and timelines constitute double jeopardy for the old-timers. Expansion makes that triple jeopardy. My point being that all the talk about the talent pool and the opportunity pool seems to some degree to be designed (a priori) to give some of our recent favorites yet another leg up.

I understand your concerns. I'm against timelining myself.

As far as season length is concerned. I plan on being extra vigilant in noting that a 30 WS season after 1961 is not the same as it used to be. Its going to be a pain in the next to do all the adjusting and its tempting to be lazy, but those extra 8 games a year adds up to an extra season over a twenty year career.

As I said before, I'm not a big fan of timelining.

On the expansion front, that's what the extra elect-me slots are for! Once we get to the 1980s we're going to have more elect-me slots than we know what to do with. I don't think guys like Van Haltren & Sisler will have anything to worry about as they are firmly entrenched on people's ballots so they'll ride in on their previous support. I'm almost more worried about the post-WWII borderline guys as its a deep backlog to insert these guys into.

Should make for some lively discussion threads in the coming years.
   127. sunnyday2 Posted: January 06, 2006 at 10:28 PM (#1809637)
And just to clarify, Chris said that the 30th best player from an earlier era probably is not as good as the 30th best player from a more recent era. I accept that as an axiom but how it plays out is a different thing. I'm not making any commitments.

But earlier he (or somebody) said that the 7th best today is equal to the 4th best then (more like 30th versus 20th), and while in practice anything can happen but I am not accepting that as any kind of useful generalization at all.
   128. Daryn Posted: January 06, 2006 at 10:31 PM (#1809640)
I'm not sure we shouldn't penalize some modern players in the sense that we should put them into a hypothetical 16 or 24 team environment and ask how long they would have really been in the MLs.

This is a great point -- offset somewhat or entirely by the fact that the pre-1950 players didn't have to compete in an integrated league, but an excellent point nonetheless. You have convinced me not to give Fred McGriff any credit for his 2004 season. Or Rickey any credit fo his last 4 seasons -- that could drop Rickey down to #1 on the ballot when he is eligible.
   129. Chris Cobb Posted: January 07, 2006 at 01:59 AM (#1809918)
But earlier he (or somebody) said that the 7th best today is equal to the 4th best then (more like 30th versus 20th), and while in practice anything can happen but I am not accepting that as any kind of useful generalization at all.

I said the 7th best first baseman is equivalent to the 4th best first baseman. I agree that there's no predicting the relative rankings of the real top 10 players of an era: they are outliers. But by the time one gets down to the 4th-5th best player at a position, we're in to the part of the curve where there is better predictability in distribution of quality, and I think the comparison of Will Clark to Bill Terry by this method bears out that it is a useful rule of thumb.

This is a great point -- offset somewhat or entirely by the fact that the pre-1950 players didn't have to compete in an integrated league, but an excellent point nonetheless. You have convinced me not to give Fred McGriff any credit for his 2004 season.

There are lots of examples of players hanging around past their having significant value prior to 1969, too. Since McGriff was below replacement level in 2004, he _benefits_ from not having that season counted, anyway.
   130. Mark Shirk (jsch) Posted: January 07, 2006 at 02:14 AM (#1809926)
A 30 WS seasons post 1961 sin't that less valuable than a 30 WS season prior to 1962 (or I shoudl say in any 154 game season). going from a 154 to 162 game scheduel takes a schedule adjsutment of 5.2% (I use 5% because it is easier and even some of the effects of simply extrapolating a player's value). 5% of 29 WS is a little under 1.5 WS. So depending on how you round up (I only round up at .7) a 30 WS season in 1975 is the same as a 29 WS season in 1935. You may dispute this and say that I should round up at .5 but the difference is definitely within the 3WS margin of error that James suggests.

As an aside, my system of schudle adjusting adds 1 WS from 14 to 33 WS and 2 WS from 34-53, and so on. It is actually pretty easy.
   131. OCF Posted: January 07, 2006 at 02:21 AM (#1809927)
There are lots of examples of players hanging around past their having significant value prior to 1969, too. Since McGriff was below replacement level in 2004, he _benefits_ from not having that season counted, anyway.

Getting from Will Clark to talking about players hanging around to the bitter end or beyond is kind of a weird segue, since Clark most certainly didn't hang around. In his last season, he was only 36, and he had a combined 145 OPS+ in >500 plate appearances. He may have had health and injury concerns, but he wasn't done if he didn't want to be done.

I do think that Win Shares overvalues Clark's 1989. Oh, he deserved the MVP (and I said so in 1989), but it wasn't that good.

I haven't put Clark into my own home-brewed offensive system yet, but I suspect he'll do very well in it.
   132. John (You Can Call Me Grandma) Murphy Posted: January 07, 2006 at 03:18 AM (#1809961)
I do think that Win Shares overvalues Clark's 1989. Oh, he deserved the MVP (and I said so in 1989), but it wasn't that good.

Didn't he hit extremely well with runners in scoring position that year?
   133. DavidFoss Posted: January 07, 2006 at 03:38 AM (#1809973)
Didn't he hit extremely well with runners in scoring position that year?

.389/.481/.688

Mitchell was:

.281/.457/.612
   134. Paul Wendt Posted: January 07, 2006 at 03:50 AM (#1809979)
Getting from Will Clark to talking about players hanging around to the bitter end or beyond is kind of a weird segue,

what he said.

Go back four decades or so, dividing salaries by 100 or so, and Clark plays in the majors as long as he can, because he needs the money.

Chris Cobb #130
But earlier he (or somebody) said that the 7th best today is equal to the 4th best then (more like 30th versus 20th), and while in practice anything can happen but I am not accepting that as any kind of useful generalization at all.

I said the 7th best first baseman is equivalent to the 4th best first baseman. I agree that there's no predicting the relative rankings of the real top 10 players of an era: they are outliers. But by the time one gets down to the 4th-5th best player at a position, we're in to the part of the curve where there is better predictability in distribution of quality, and I think the comparison of Will Clark to Bill Terry by this method bears out that it is a useful rule of thumb.


This contemporary-rank-at-position metric is extremely sensitive to nearly-irrelevant variations. If Greenberg arrives one year earlier, Terry's score is 5/16 rather than 4/16. If Killebrew, Stargell, and Carew move to first base a little earlier, and get counted as 1Bmen, where is Tony Perez?

the part of the curve where there is better predictability in distribution of quality
Granted, that plaudit is not well-defined, but I feel that it is "true enough" in the vicinity of "4th-5th at fielding position" only with continuous (playing-time-weighted) rather than discrete (yes or no) definitions of both contemporaries (the problem illustrated by Greenberg and Terry) and fielding positions (the problem illustrated by Killebrew and Perez).
   135. Paul Wendt Posted: January 07, 2006 at 04:06 AM (#1809994)
I see that 2006 voting is open to all comers.

Among sabrmetric favorites, I see that Bobby Grich is in his last year of (write-in) eligibility. Ted Simmons, Darrell Evans, Kent Tekulve, and Dan Quisenberry have a few years of life.
   136. OCF Posted: January 07, 2006 at 05:16 AM (#1810030)
What I remember from 1989, working with some version of Runs Created that I'd gotten from some previous Abstract, is that I thought that there was very little offensive difference between Clark and Mitchell, and it didn't seem likely (1B versus flank OF) that there was an important defensive difference. Obviously, the Win Shares calculation finds differences - in part, the clutch hitting stats - that weren't visible to me in 1989.

There are a number of predictable biases in BBWAA voting for MVP. Probably the most powerful such biases involve team record; those do not apply when we're comparing teammates. But there are some others that do apply to the Clark/Mitchell choice. One is the "new story" bias: the writers tend to like surprises and breakout years. You have to undertand that the writers aren't really analysts, they're storytellers. They got their jobs by telling stories; they like telling stories. And "Guy who's been really, really good for a while has another good year" isn't the most compelling story. Another bias is RBI's - we know they like RBI's. That leans in Mitchell's direction. Another, tangentially related, is "protection." The writers were fully aware that Clark had had a good year even compared to his own lofty expectations, but I saw a number of suggestions that attributed Clark's production to the superior "protection" that Mitchell offered by batting behind him. (I saw nothing about how Clark, by constantly being on base, contributed to Mitchell's season by enhancing his opportunities.)

My statement that I thought Clark deserved the MVP was in part recognition that Clark had indeed enhanced Mitchell's season by providing him RBI opportunities, but was mostly a conscious reaction against the "new story" bias - I'm more impressed by those who stay at the top than by the guys with skyrocket seasons. If you ask the question, "Who had the better 1989 season," then that may be a close call (even though Win Shares doesn't think it's close). But suppose that at the end of the 1989 baseball season, you asked the questin this way, "Based on all currently available evidence, who is the better baseball player right now: Mitchell or Clark?" Then you'd have to say Clark. It's not that hard a concept: ask who had the better 2005 season, Albert Pujols or Derrick Lee, then you'll get an argument - the answer isn't obvious. But ask which of those two is a better baseball player right now, and the answer is clearly Pujols.

The "new story" bias worked against Clark even though he had never previously won an MVP. The same didn't happen to Pujols in 2005. Two comments there: for Pujols vs. Lee, there are team biases that apply. But even more importantly, it's clear that the writers see Pujols as a player of historic importance who must be honored when possible. They never saw Clark that way. Of course they probably never adjusted to the offensive downturn that followed the 1987 high-water year.

Put "new story" bias and RBI bias together, and there's no doubt in my mind that in an election conducted under modern eligibility rules, Lou Gehrig would have won the 1927 AL MVP.
   137. sunnyday2 Posted: January 07, 2006 at 06:23 AM (#1810060)
>You have to undertand that the writers aren't really analysts, they're storytellers. They got their jobs by telling stories; they like telling stories

I said exactly this several years ago: As much as I admire Miggie Tejada, his MVP award is the example par excellence. It's not really the MVP award, it's the Best Story award. Not just Miggie but every year, Miggie just being the most pronounced case.
   138. Joey Numbaz (Scruff) Posted: January 07, 2006 at 09:21 AM (#1810216)
"First, there were, on average, 28 major league teams in Clark’s era to 16 in Terry’s, with a correspondingly larger player pool, so we should expect the 4th best 1B of the 30s to be equivalent to the 7th best 1B of the 90s, acknowledging that both eras were boom times for slugging 1Bmen.

This to me is exactly wrong. The caliber of players has nothing directly to do with how many major league teams there are. This implies that if there had still been 16 teams in 1985-1995 that Will Clark would somehow have been a different player than who he was. The pool is the pool, opportunities don't change that."

I think you are way off there Marc. The population of the world has been growing at a pretty steady rate throughout the 20th Century. It's tripled in the last 100+ years. Our national population has almost quadrupled. You have to account for that, and being the 7th best 1B of 1993 is certainly equivalent (if not better) than being the 4th best 1B of 1929. It's entirely reasonable that there are 30 opportunities now instead of 16. If anything it's tougher now than it was then.
   139. sunnyday2 Posted: January 07, 2006 at 01:44 PM (#1810274)
Obviously this debate is going nowhere, and I am not expressing my point clearly. But I think you are conflating two independent variables there. The talent pool is independent of the number of ML teams. Both are bigger today than they were in 1960 or whenever, but there is no direct relationship between the two. Neither one is a cause or an effect of the other.

I mean, when you say "it's tougher now" what does it really mean? In at least one sense it is easier now. 30 opportunities mean that there are 14 1B playing this year who would have been in the minor leagues or retired in 1960.

This goes back to the old question of whether we are trying to honor players for their value or their ability.

• The larger talent pool suggests though it does not in any way prove that players today have more ability (though I would say that if there is more ability it has more to do with nutrition, conditioning, coaching at a younger age, etc., than it does purely with the talent pool). Injecting this into the debate means we are talking about ability. If we want it to be about ability and we want to say #7 today ? #4 then, don't we have to know what the actual size of the talent pool really is?

• The larger opportunity pool simply says that there are 7,290 WS available in a modern ML season (162 X 30 X 3/2) versus 3,696 100 years ago. If this is the point, then we are talking about value and we are saying that there is twice as much "value" to be won today. And this of course would be true. Does this mean that we should elect twice as many players from today? What about the idea that a pennant is a pennant, which militates quite exactly against a HoM membership that tracks the opportunity pool. I mean, in that sense, the opportunity is always 1.000 world championships.

So my questions are these: Aren't these separate phenomenon? What does each or both of them mean for the numbers of players we should elect from different eras? Are we saying we should elect more players from today than previous eras because of these concepts? And if so, which of these two phenemenon are we appealing to?

I am not being clear I guess but neither (to me) are those who keep appealing to these two concepts as if they were only one.
   140. Mark Shirk (jsch) Posted: January 07, 2006 at 04:23 PM (#1810337)
If population didn't increase would we have had 14 new MLB markets since 1960? I don't realy think so. And while MLB players don't usually come from the segment of the popultion that holds season tickets (though more do than with say, basketball or football) both segments (those that fund expansion and those that will be playing MLB baseball) are growing.

So no, there are not 30 teams now simply because the talent pool has grown, but the talent poll growth and the growth of new markets are both connected to an overall population growth.
   141. sunnyday2 Posted: January 07, 2006 at 04:51 PM (#1810353)
>but the talent poll growth and the growth of new markets are both connected to an overall population growth.

Yes, nobody is denying that, but that is exactly as far as the connection goes.

The question is, so what? How do you use that information to evaluate players (to justify the timeline, basically)? And which insight is it that you're (the royal you) using--that the talent pool is bigger or that there are more teams? All I'm saying is the appeals to one or the other or both are usually sorta muddled together and I can't seem to follow the actual logic of the argument. Maybe it's me.
   142. John (You Can Call Me Grandma) Murphy Posted: January 07, 2006 at 09:48 PM (#1810634)
FWIW, I'm an anti-timeliner who believes they're are more great players today.
   143. sunnyday2 Posted: January 07, 2006 at 09:59 PM (#1810645)
Well, again, John,I don't disagree, but the question is still so what?

i.e. do you mean there is more ability in the pool? Or more value?

And should we therefore elect more players from recent years?

I'm just trying to ask, so what are we gonna actually DO with this information?
   144. John (You Can Call Me Grandma) Murphy Posted: January 07, 2006 at 10:32 PM (#1810687)
Well, again, John,I don't disagree, but the question is still so what?

I wasn't attempting to answer anybody's question(s). That's why I typed "FWIW." Obviously, you think it's worth diddly-squat. :-)

i.e. do you mean there is more ability in the pool? Or more value?

I mean there are more great players now than from earlier generations, even after taking into account the disadvantages that the earlier generations faced.

And should we therefore elect more players from recent years?

Yes, though not at the expense of the earlier generations by timelining. That's why I pushed for more players to be elected in the later elections for the HoM.
   145. Paul Wendt Posted: January 08, 2006 at 05:27 PM (#1811239)
And should we therefore elect more players from recent years?

Yes, though not at the expense of the earlier generations by timelining. That's why I pushed for more players to be elected in the later elections for the HoM.


Yeah, but you should push for that structure even if you think we should elect more earlier players (eg, it's a Hall of Historical Significance and the frequency of historically significant events or people tends to decline).

Given the annual structure, proceeding according to time's arrow, the time pattern of inductions sets a minimum number of honorees from every "early time period" (every period that begins at the beginning) but it sets a maximum only for the whole period (the beginning to 2003 or so). The accelerating induction rate permits the selection of more later players but also permits the selection of more early players. George Van Haltren gets 100 opportunities for election, Tim Raines only 1.
   146. sunnyday2 Posted: January 08, 2006 at 05:32 PM (#1811245)
I agree with the logic of the numbers of players elected and when.

As has been stated here, as we get into the 3X years, we will have the opportunity to select the best backloggers regardless of era. That's a good thing.

I do have a problem with the notion that, to the contrary, the point is not to select the best backloggers regardless of era, but rather, to elect 3X newbies when that time comes, that the number of players we elect and when we elect them is also some sort of directive as to who to elect.

These are two quite different concepts of what our structure means to our balloting.
   147. Paul Wendt Posted: January 08, 2006 at 05:34 PM (#1811249)
Quoting the first line, revised to match the terminology of the whole:

- you should push for that time pattern even if you think we should select more earlier players

Well, you should not, if you think the structure should enforce selection of more earlier players. But if you believe the electorate should be permitted to go either way, you should support the accelerating induction rate.
Page 2 of 2 pages  < 1 2

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Andere Richtingen
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.8095 seconds
68 querie(s) executed