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, September 19, 2005

1961 Ballot Discussion

1961 (October 3)—elect 1
WS W3 Rookie Name-Pos (Died)

265 78.8 1942 Vern Stephens-SS (1968)
242 73.1 1946 Ralph Kiner-LF (living)
199 63.0 1943 Sid Gordon-LF/3B (1975)
209 55.1 1939 Eddie Joost-SS (living)
176 60.6 1944 Ed Lopat-P (1992)
147 55.9 1942 Johnny Sain-P (living)
161 47.5 1947 Ferris Fain-1B (2001)
125 39.7 1942 Willard Marshall-RF (2000)
113 38.5 1947 Vic Raschi-P (1988)
104 35.3 1939 Hank Majeski-3B (1991)
108 33.5 1943 Peanuts Lowrey-LF/CF (1986)
107 31.1 1946 Eddie Waitkus-1B (1972)

1961 (October 3)—elect 1
HF% Career Name-pos (born) BJ – MVP - All-Star

00% 38-55 Pee Wee Butts-SS (1919) #6 ss – 0 – 1*

Players Passing Away in 1960
HoMers
Age Elected

87 1917 Fred Clarke-LF
86 1929 Bobby Wallace-SS

Candidates
Age Eligible

91 1913 Tommy Corcoran-SS
89 1905 Kid Carsey-P
80 1912 Noodles Hahn-P
79 1925 Terry Turner-SS/3b
78 1920 Howie Camnitz-P
76 1921 Russ Ford-P
72 1927 Bill Killefer-C
69 1932 Stuffy McInnis-1B
68 1935 Art Nehf-P
67 1932 Everett Scott-SS

Thanks to Dan and Chris!

John (You Can Call Me Grandma) Murphy Posted: September 19, 2005 at 01:21 AM | 187 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. Mark Shirk (jsch) Posted: September 24, 2005 at 10:19 PM (#1640228)
Prelim

1. Wes Ferrell
2. Cupid Childs
3. Joe Medwick
4. Hugh Duffy
5. Dick Redding
6. Charlie Keller
7. Bucky Walters
8. Earl Averill
9. Dobie Moore
10. Ralph Kiner - A conservative placement right now, he could be as high as seven though it is is unlikely.
11. Clark Grffith
12. Quincy Trouppe
13. Joe Gordon
14. Pete Browning
15. Eppa Rixey

Ted Lyons make smy PHOM. My current backlong looks like this:

Duffy
Redding
Keller
Walters
Averill
Moore
Kiner
Terry
   102. sunnyday2 Posted: September 25, 2005 at 01:30 PM (#1641204)
Others have done this too. It seems likely that we will elect a hitter, though if Jennings can go from #5 to #2 then Ruffing and Ferrell are not out of it. Still, here are the top hitters (by consensus + newbie).

In alpha order: Averill (#2 holdover), Kiner (top newbie), Medwick (#1 holdover), Sisler (#8 holdover).

The Big Picture (from TNBJHBA)

14. Averill (CF) 280/33-30-30/143/27.2
18. Kiner (LF) 242/37-35-30/155/26.6
13. Medwick (LF) 312/40-36-33/157/25.5
24. Sisler (1B) 292/33-29-29/135/23.0

Also courtesy of WS: Sisler C- and 23 defensive WS (+ 9 pitching WS), Averill A+ and 56, Kiner C- and 26, Medwick B- and 44, but of course these WS are included in the above totals

I prefer 3 consecutive and 5 non-consecutive myself: Averill 89-146 Kiner 97-155 Medwick 109-157 Sisler 89-143

Why will Kiner and Medwick be at the top of my ballot? It comes down to WS peak, I guess. Well, that and OPS+.

Averill 132/159-49-47-43-37-34-31-19-19-15
Kiner 148/183-82-72-54-45-40-31-21-16-x
Medwick 133/170-57-49-40-38-33-29-28-22-18-15-97
Sisler 124/179-69-63-59-53-37-33-19-9-0-98-90-84

Medwick and Sisler of course have the longer careers.

Sisler's odd career shape obscure this simple fact. His 6th and 7th best seasons (by OPS+) are better than Averill's 6th and 7th best seasons, and his three best years after eye trouble (119-109-100) would not look out of place with Averill's next 3 (119-119-115). That's 25 points of OPS+ to Averill over 3 years versus a 56 point advantage to Sisler over their 3 best years and 35 more over their #4-5-6 years.

And Sisler has about 1600 additonal PAs. Even give Averill 2 years of PCL MLE credit and he remains behind Sisler on career length. So I still like Sisler over Averill. But I still also prefer Kiner and Medwick (in some order) over Sisler, Kiner on an even higher peak and Medwick on a somewhat better career as a whole.

1. Medwick (or Kiner)
2. Kiner (or Medwick)
3. Sisler
4. Averill

OTOH Averill has bounced around my ballot pretty good (from 18-20 to the 30s, lately around #25). Now, on my prelim, #15, though that #15 spot is still up for grabs between Averill and Vern Stephens.
   103. Howie Menckel Posted: September 25, 2005 at 06:38 PM (#1641528)
Good post, sunnyday!
The thing about Sisler is that it is true that Sisler's later years don't have much value.

I wouldn't even mind if someone dismissed those later years completely - IF they then adjusted to picture Sisler as a pure-peak, short-career player. He'd have a monster OPS+ like other short-career candidates, for instance, if you dump the later part.

I have Sisler-Kiner as VERY close.
Close enough that Sisler's 111 career IP at 124 ERA+ tips the balance? Maybe.
Not sure any of the four get an 'elect-me' vote, though, from these corners. Still some work to do.
   104. Chris Cobb Posted: September 25, 2005 at 08:07 PM (#1641811)
While I favor the eventual election of Averill, Kiner, Medwick, and Sisler, none will be near the elect-me spot on the 1961 ballot (and two or three will be off my ballot altogether), mostly because there's a glut of pitchers at the top of the ballot.

We struggle hard enough to measure empirically the differences in value among candidates at the same positions that I don't see any succinct way to make an empirical argument that there are a good half dozen pitchers who are better candidates than some of the much-discussed outfielders, so I'll just remind everyone of a couple of quota-related matters.

We have so far elected, by my count, 29 pitchers and 33.5 outfielders: I expect that those numbers should be about equal. We're not quite there. Moreover, we have two outfielders queued up at the top of the returnees for 1961 and a highly similar third candidate arriving fresh. Unless the electorate starts valuing pitchers more, therefore, this imbalance seems likely to grow rather than diminish over the next several elections.

Rixey and Ruffing have tremendous career value and substantial peak value when the lower average value of pitchers is taken into account.

Wes Ferrell has the best peak among eligible players.

Clark Griffith offers an excellent combination of career value and peak value, especially if it is remembered that he probably lost two two seasons of major-league play to contraction in 1892-3.

These four top-ten returnees merit just as much consideration as the top returning outfielders do. It's beyond me to say how exactly you should weigh them against those outfielders, but the evidence suggests that we haven't been giving pitchers quite enough credit.
   105. Mark Shirk (jsch) Posted: September 25, 2005 at 08:54 PM (#1641951)
My problem with Sisler, 'problem' in that he has been around 20-25 instead of on ballot, isn't the bad years that he put up after his sinus infection. It is that his peak sin't that historic. He has only one thirty WS year, (though he has 29,29,27) and he has only one 10 WARP1 year (though he has plenty of 9's). Those shoulder years are good but I dont' see his peak as historic enough to pu thim from strong candidate to automatic ballot guy.

He has some big OPS+ measures, but remember that he didn't walk much so his OPS+ is a little slugging heavy, and he had the least amount of defensive value.

Sunny,

How about Charlie Keller? Where do you rank him? Since I will have King Kong at #6 this year I don't want him to be left out of the outfielders discussion. With war credit his career and prime look very similar to Kiner and peak depends on what you trust more WS (Keller) or WARP3 (Kiner). If you aren't giving war credit then i guess it is a moot point.

Chris,

I agree somewhat, though I would like to see a diferent set of pitchers looked at. I will have Rixey and Griffith on my ballot, but three of my top 7 are Ferrell (1), Redding (5), and Walters (7). And Ted Lyons is finally making my PHOM.
   106. Mark Shirk (jsch) Posted: September 25, 2005 at 08:59 PM (#1641958)
Oh, I was thinking about the strength of schedule adjustments for WS and shouldn't we be adjusting notfor quality of team but instead for quality of a team's pitching and defense? The advantage that is derived is not that a player isn't facing his own pitching and defense, not that he isn't facing his own hitters. So offense only winners and defense only losers would be distorted by looking only at team record right?
   107. sunnyday2 Posted: September 25, 2005 at 09:25 PM (#1641988)
Sisler

Thought experiment. Instead of starring at the University of Michigan for four years (which he did), but instead of that, let's pretend he turns pro at 19. He puts up two or three of those later years at age 19-20-21. Then he has 7 killer years. Then he puts of two or three decline years.

Now, is this player any better or different or more valuable than the George Sisler you all know and love.

No, he's not. But he would be doing a lot better on your ballots.

And BTW, four of his seven 20 WS seasons were in the 'teens. Does WS underestimate 1B defense during that period? Consensus seems to be yes.

So there is a very weird confluence of bad luck, bad karma and bad faith working against Sisler.

• Like I said, consider whether the fact that he WAS playing baseball at the college level before MLs is worth any MLE XC. If AAA conversions are .92 and AA is .85, then what is A ball, .75? Then what is Michigan?

• Pretend he has a normal curve.

• Don't forget that he broke in as a pitcher.

• WS probably underestimates his defense. Add 1.5 WS per year x 14 = 21 more WS. Is that close?

• Now he's got a couple 30 WS seasons and a 28.5.

• And now think of all the things you hate about his nemesis, WS and Bill James.

Now does Sisler look any different?
   108. sunnyday2 Posted: September 25, 2005 at 09:27 PM (#1641992)
Ps, j, there is a series of 3 posts on the Kiner thread looking at the "moneyball" guys include Keller.

Charlie moved up my ballot as a result, from about 55ish to about 20. But even if you give him the 1.5 years that he spent in the merchant marine during WWII he is still short of Kiner by a full season. Secondly, his career (like Sisler's) has an odd shape. He spent about 2 season's worth of PAs spread out over a period of 5 years. No way can those 2 years worth of PAs be as valuable as if they really were in 2 actual seasons in terms of pennants-added.
   109. Michael Bass Posted: September 25, 2005 at 09:31 PM (#1641996)
Since career shape doesn't factor in my rankings one ounce, I'm pretty confident saying Sisler's unusual shape isn't hurting him with me. :)

Rather, like jschmeagol above, I find his peak to be not historic. The one year was historic, the rest is not really peak so much as prime. He's *just* off my ballot, so it's not like I hate him, I just see no pressing need to put him over people with longer primes/careers, such as Averill.

(And I still like Sisler a decent amount better than Medwick)
   110. sunnyday2 Posted: September 25, 2005 at 10:07 PM (#1642071)
Averill, of course, didn't have a longer career than Lonesome George. ;-)

Longer prime? As shown above (and below) his longer prime consists of being better than George in their 9th and 10th best seasons and then 119-109 and 115-100. George is better for 7 years and even for the 8th.

Averill 132/159-49-47-43-37-34-31-19-19-15
Sisler 124/179-69-63-59-53-37-33-19-9-0-98-90-84

Now if you believe WS Averill was an A+ fielder (as a CF he couldn't be much worse, I mean Hack Wilson was a C+) while
Sisler was a C-. Was George a worse fielder than Hack? No, we know that WS is too generous to CFers and too stingy with deadball 1B.

But even then, with Averill's big defensive WS edge, all it does is get him within a 292-280 deficit of Sisler for career WS.

Let's see--career, prime, peak. Where was Averill better again? Oh, I remember, while he was in the PCL! ;-)

I just don't get it.
   111. Mark Shirk (jsch) Posted: September 25, 2005 at 10:34 PM (#1642099)
I think that giving college credit would set a dangerous precedent. What about Mark Prior? Mark Texeira (?)? The precedent has become to only give extra credit to plyers in situations that they could not help like being black, being drafted, and stupid talent evaluators. Could Sisler have left colelge early to play baseball? If the answer is yes I cn't imgaine giving him any credit. Averill gets only one year from me.

I think that too much credit is given to late dead-ball era 1B as a counterweight to Win Shares. There is simply no way that Sisler had the defensive value of Averill. Plus it has usually been teh case that it is harder to find offense at CF than at 1B, even during Sisler's time.

There are a myriad of problems with OPS+, not the least being that it favors players who were heavy in Slugging over those that were heavy in OBP, even though OBP seems to correlate better to run prodcution. There is also the fact that during times when the level of competition isn't as high (meaning the average is lower) high OPS+ numbers are easier to accumulate. This counteracts whatever 'advantage' Sisler has in being a dead ball era 1B. That Sisler has some more impressive OPS+ numbers doesn't really carry much weight with me.
   112. KJOK Posted: September 25, 2005 at 10:51 PM (#1642115)
Plus it has usually been teh case that it is harder to find offense at CF than at 1B, even during Sisler's time.

Actually, the two were very close offensively during Sisler's time.
   113. karlmagnus Posted: September 25, 2005 at 11:22 PM (#1642173)
Jschmeagol, there's deadball and deadball. The 1911-20 period had high batting averages, hence high OPS+ was not so easy, since nobody was hitting home runs. Sisler's 1920 and 1922 were both historic, his WS are undervalued because they get 1B wrong pre-1920 (and only decreasingly wrong in the 20s)and for those voters who like Cool Papa Bell, I would suggest that Sisler's contmporary "buzz" should also be a big point in his favor -- not conclusive, but not ignored altogether.

I don't think anyone in 1930-1950 would have claimed that Averill, Klein or Medwick were "better" than Sisler. We should be very careful before simply assuming that contemporaries were out to lunch.
   114. Howie Menckel Posted: September 26, 2005 at 12:20 AM (#1642251)
Interesting responses.

First, I guess all would agree that if Win Shares are wrong about Sisler, then we are making a mistake in not electing him.

Which produces the second question: What if WS are a bit off on a 1910-30 1B?
I promise I am not a slave to OPS+, and accept some of the "OBP-challenged' criticism.

Are you other guys as willing to shine a spotlight on WS?

I'll bet that Bill James himself would welcome my skepticism (not necessarily of the entire approach, but of being too committed to dismissing any player who otherwise might be HOM-worthy).

I agree with karlmangus's take as well.
Clearly Sisler was NOT as good a player as his reputation. But his reputation has him inner-circle HOM. Is it not possible that he is outer-circle instead?

Finally, as has often happened over these years, I concur on Chris's pitching-shortage concerns - and on who can best solve it.
   115. Chris Cobb Posted: September 26, 2005 at 01:08 AM (#1642363)
Two quick comments on Sisler:

1) In judging his peak in win shares, remember that two of his peak seasons came in 1918 and 1919, which were shortened to 128 games and 140 games, respectively. Bump those up and his peak improves.

2) WARP takes, I believe, a fairer view of first base defense pre-1930. In 1920, it looks like it treats an average first baseman as 17 FRAR, an average left fielder as about 20 FRAR, an average right fielder about 19 FRAR, and an average CF about 30 FRAR.

A rough weighting of batting value by position, then, would suggest that, in comparing Sisler's peak to those of corner outfielders from the 1930s (just as random example :-), their hitting should be weighted equally. In comparing a first baseman to a centerfielder, however, the centerfielder could, on average, be 10-15 batting runs worse than the first baseman and be of about the same overall value.
   116. Howie Menckel Posted: September 26, 2005 at 01:11 AM (#1642369)
More good stuff, Chris.

I'm not so much a Sisler advocate as a guy who fears he has not gotten full consideration due to WS.
That's all I ask, and I know that this electorate is willing to consider all angles.
   117. TomH Posted: September 26, 2005 at 01:27 AM (#1642392)
Agree that Sisler is underrated by WS.

Also agree that his contemps were knocked off their socks by his defense, which somehow the metrics have not captured.

Over their careers, Averill was a better rate hitter, regardless of which metric (EqA has EA at +12, OWP has him at +.035) is used. But Sisler's longer career brings them even in BRAR, for example.

1B in Sisler's day hit better, relative to league, than CF in Averill's time. OWP difference of .537 to .509.

BP gives Averill small edge (2.5 career adjusted Wins) in league quality. This seems reasonable.

Sorting best seasons by RCAA we get
Sisler. 89 84 49 48 44 37 21 07 -1
Averill 72 58 53 44 41 36 30 26 22 20

Depends on taste I guess. Cancelling out any negative RCAP seasons, Averill leads 339 to 330.

George pitched. Averill may get some minor league credit.

They ARE very close. As are Beckley, Chance, Medwick, and a ton of other OFers.

I would agree with Chris Cobb's assesment, but say it more strongly. I believe it would be a clear sign of IMbalance if we honored more OFers than pitchers.

Griffith, Ferrell, Ruffing, Rixey, Walters, NegL/CubanL pitchers, all ought to be strongly considered.
   118. Michael Bass Posted: September 26, 2005 at 01:31 AM (#1642394)
WARP1 (which BTW, takes a dim view of Averill's defense):

Prime Year comparison

Best: 14.2 vs. 12.6 Sisler
2nd best: 11.7 vs. 9.4 Averill
3rd best: 10.5 vs. 9.2 Averill
4th best: 10.2 vs. 9.1 Averill
5th best: 9.4 vs. 9.1 Averill
6th best: 9.0 vs. 8.4 Sisler
7th best: 8.8 vs. 8.2 Sisler
8th best: 8.1 Averill vs. 4.5
9th best: 7.3 Averill vs. 4.5
10th best: 6.9 Averill vs. 4.3

I don't even count years 8-10 as prime for Sisler. So Averill had a longer prime by 3 years, not counting any PCL credit (I give him one prime year for his last year). Over their top 7 years, Averill has more total WARP. From years 8+, Averill kills Sisler, just kills him. Toss in a prime PCL year and...
   119. karlmagnus Posted: September 26, 2005 at 01:44 AM (#1642410)
Michael, I don't think you've pro rated 1918 and 1919 in Sisler's career to 154 games, which you clearly need to do.
   120. Mark Shirk (jsch) Posted: September 26, 2005 at 01:48 AM (#1642413)
It isn't just WS, it is WARP as well. If only we had SLWTS that go back before 2000 or whatever.

For some reason both uberstat systems prefer Alverill over Sisler and both of them agree that Sisler's peak was not historic. We throw out these stats and their other components (Eqa, RC, the defensive metrics) aren't we largely left with the very metrics that overrated Sisler in the first place?

Also, I am not a Sisler hater and am not sure how I got on this side of the argument but I also don't think we are necessarily underrating him. I think we have it right, he is in the backlog and the part of the backlog that has a real chance to get elected.

So I guess it is good we are having this discussion! ;-)
   121. Michael Bass Posted: September 26, 2005 at 05:03 AM (#1642584)
Michael, I don't think you've pro rated 1918 and 1919 in Sisler's career to 154 games, which you clearly need to do.

Not an unfair point, though a decent amount of that would be offset by a quality of competition adjustment for 1918, I think.
   122. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 01:25 PM (#1642853)
I don't even count years 8-10 as prime for Sisler. So Averill had a longer prime by 3 years, not counting any PCL credit (I give him one prime year for his last year). Over their top 7 years, Averill has more total WARP. From years 8+, Averill kills Sisler, just kills him. Toss in a prime PCL year and...

I agree. Averill is the far better choice, Michael.
   123. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 01:29 PM (#1642857)
Also agree that his contemps were knocked off their socks by his defense, which somehow the metrics have not captured.

Win Shares indicates he was a mighty fine defensive player prior to his sinus problems. After that, like his hitting, his defense fell apart.
   124. Howie Menckel Posted: September 26, 2005 at 01:31 PM (#1642869)
I've got a bad feeling that some non-Sisler 'his peak isn't that great' voters are forgetting that his TEAM only played 122 games in 1918 and 139 in 1919.

karlmagnus's point on that is crucial, I say.
   125. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 01:39 PM (#1642885)
I wouldn't even mind if someone dismissed those later years completely - IF they then adjusted to picture Sisler as a pure-peak, short-career player. He'd have a monster OPS+ like other short-career candidates, for instance, if you dump the later part.

Are there voters here who are actually giving Sisler demerits for playing at an average or sub par level after 1922? IOW, in their systems, Sisler would come out better if he had retired after '22?

I keep hearing that this is so, but I don't think anybody has expressed an opinion in that fashion. If they did, they are wrong.

IMO, the number of people who are doing this has to be extremely low.
   126. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 01:41 PM (#1642888)
I've got a bad feeling that some non-Sisler 'his peak isn't that great' voters are forgetting that his TEAM only played 122 games in 1918 and 139 in 1919.

I'm not, Howie. Since I compare him to his peers, he's not hurt by that. He's also not hurt by WS problems with defense at first base in my system for the same reason.
   127. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 01:47 PM (#1642897)
I don't think anyone in 1930-1950 would have claimed that Averill, Klein or Medwick were "better" than Sisler. We should be very careful before simply assuming that contemporaries were out to lunch.

Well, since they were more BA-oriented back then, that's not surprising. Since Sisler just misses having the highest batting average at his position, he would have rated higher than the guys will a little pop (but less contact skills) in the view of the oldtimers.
   128. KJOK Posted: September 26, 2005 at 05:30 PM (#1643367)
Are there voters here who are actually giving Sisler demerits for playing at an average or sub par level after 1922? IOW, in their systems, Sisler would come out better if he had retired after '22?

Well, he doesn't do better in my system than if he had RETIRED, but he certainly gets "demerits" for those 9th thru 14th best seasons vs. his "competition" such as Beckley, etc....
   129. Dr. Chaleeko Posted: September 26, 2005 at 05:46 PM (#1643408)
Talking about the pitchers...

My system adjusts pitchers' WS for usage patterns by attempting to place each season on historically even footing. For season X, I average the WS of the leading pitcher and the fifth-place pitcher and then find an adjustment factor by comparing to the historical average of leading/fifth-place pitchers. Finally I apply the adjustment factor to an individual pitcher's WS in season X to find his adjusted total.

My adjusted totals for some of the pitchers currently receiving support, seasons listed from highest adjWS total to lowest:
 RR  ER  CG  RW  AJ  WF  BW  BG  TB  DT  CM  VW
-----------------------------------------------
 25  25  24  27  26  35  39  27  26  35  31  31
 24  25  23  25  22  29  33  24  20  23  25  28
 24  23  22  25  19  27  31  22  19  22  23  24
 23  21  20  20  18  26  27  21  19  17  21  19
 22  21  20  17  16  24  20  20  18  17  20  18
 22  21  15  16  16  23  14  19  18  16  20  17
 20  20  15  14  13  18  14  16  17  16  19  17
 17  18  12  14  12  17  14  15  16  15  18  17
 16  17  12  13   6  14  14  15  15  13  15  16
 15  17  10  10       6  12  15  15  13   9  15
 15  17   8   6       2  11  10  13  13   9  14
 15  15   8   1       1  10   9  12   9   6   5
 14  14   5   1       1   6   4   6   5   5
 14  14   3           0   3   3   3   3   5
 12  13   1               1   3   1       1
 10  10                   1   2   0
 10  10                   0   1
  9   9                   0
  8   8                   0
  8   6                   0
  7   5
  6   0
  0 
  0 
  0 
  0 
===============================================
336 330 198 188 148 222 250 290 218 216 228 247

RR=Ruffing
ER=Rixey
CG=Griffith
RW=WAddell
AJ=Joss
WF=Ferrell
BW=Walters
BG=Grimes
TB=Bridges
DT=Trout
CM=Mays
VW=Willis


Then to compare them to hitters, I've gone back and figured the same information for batters (yearly 1st and 5th-place totals) and found the average difference between pitchers and hitters. Turns out that the historic difference is about 20% toward hitters. So to compare against hitters, I adjust the adjusted numbers up by 20%. Here's how the guys above come out for their careers as well as in several nonconsecutive increments en route:
    RR  ER  CG  RW  AJ  WF  BW  BG  TB  DT  CM  VW
--------------------------------------------------
 3  88  88  82  91  80 109 123 108  78  96  95  99 
 5 142 139 129 136 121 169 179 163 123 136 145 153
10 251 250 207 217 177 262 260 274 220 223 243 255
15 333 338 237 226 177 266 299 337 262 259 274 291
ca 403 396 237 226 177 266 300 348 262 259 274 291
   130. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 06:43 PM (#1643520)
Well, he doesn't do better in my system than if he had RETIRED, but he certainly gets "demerits" for those 9th thru 14th best seasons vs. his "competition" such as Beckley, etc....

I don't give him much value points for his post-1922 career, but he does get something to tack on to his resume (however negligible it is). If that's what your system is doing, then we agree.
   131. Paul Wendt Posted: September 26, 2005 at 06:50 PM (#1643534)
Over their careers, Averill was a better rate hitter, regardless of which metric (EqA has EA at +12, OWP has him at +.035) is used. But Sisler's longer career brings them even in BRAR, for example.

1B in Sisler's day hit better, relative to league, than CF in Averill's time. OWP difference of .537 to .509.

BP gives Averill small edge (2.5 career adjusted Wins) in league quality. This seems reasonable.

Sorting best seasons by RCAA we get
Sisler. 89 84 49 48 44 37 21 07 -1
Averill 72 58 53 44 41 36 30 26 22 20

Depends on taste I guess. Cancelling out any negative RCAP seasons, Averill leads 339 to 330.

George pitched. Averill may get some minor league credit.

They ARE very close. As are Beckley, Chance, Medwick, and a ton of other OFers.


The conclusion of this argument should be that
they ARE very close as batters.

In order to conclude that they ARE very close, you need to dismiss not only the WS evaluation of Sisler as C- firstbaseman but the WS evaluation of Averill as A+ centerfielder.

Anyone who thinks Averill is very close to one of these others as a batter should rank Averill higher --much higher, if they are not both ranked highly, because the rankings are so close (#10 is close to #30).
   132. sunnyday2 Posted: September 26, 2005 at 07:14 PM (#1643605)
Just about every decent CF is an A+ and just about every corner OF who isn't named Manny Ramirez is a C.

Of course Averill earns 56 career WS versus Sisler and Kiner and Keller et al in the 20s.

Career WS: Sisler 292 Averill 280. You don't add the defensive value on top of that, it's already in there. It seems that we often hear this argument--that two players are about equal on CWS but then the one was a better defender. Defense is already in there.

So WS does not conclude that Averill is better than Sisler at all.
   133. karlmagnus Posted: September 26, 2005 at 07:47 PM (#1643666)
Dr. Chaleeko, I think your system has the effect of suppressing dead ball era peaks while keeping longer lively ball careers. In order to be fair, you should first do the calculation you have done, then compare their career length to the fifth longest career of any contemporary pitcher, and adjust the numbers accordingly. Only thus can you give a fair comparison between Joss/Griffith/Waddell and the 30s guys.

Even then, a system that says Joss's peak is the lowest of the 12, except for Griffith's historic 1898 seems to me to leave something to be desired.
   134. Dr. Chaleeko Posted: September 26, 2005 at 08:18 PM (#1643727)
I think any system which compares deadball pitchers to lively ball pitchers will suppress the peaks of deadball pitchers---in fact, that's somewhat the point of what I'm trying to do. Deadball pitchers had bigger peaks by raw WS totals because of the usage patterns of the era. I'm trying to ratchet them down to WS totals that make sense to me while preserving the in-season pecking order.

The fact that Joss ranks 12th in this group is more a matter of the facts at hand: short career to suppress his career totals, pitched in deadball era, so ratcheting him down makes him look more "normal," and IIRC, he wasn't leading his league in WS very often, so he doesn't exactly benefit from dominance either. Compare to Waddell. Same exact era, but Waddell dominated more often than Joss, and he comes out looking a little better.

Meantime, I forgot to mention Mickey Welch.

Here's his adjusted WS, followed by his batter-comparison numbers.
 MW  
---
 29
 25
 24
 19
 19
 17
 16
 16
 15
 12
  6
  2 
  0
===
199

 3  95
 5 140
10 229
15 239
ca 239
   135. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 08:37 PM (#1643763)
Career WS: Sisler 292 Averill 280. You don't add the defensive value on top of that, it's already in there. It seems that we often hear this argument--that two players are about equal on CWS but then the one was a better defender. Defense is already in there.

So WS does not conclude that Averill is better than Sisler at all.


Yes, it does, Marc.

WS/162 Games: Sisler 23.01 Averill 27.18

Hands dow, Averill is the winner using WS.

I can't stress this enough: if you are using straight WS in your analysis, than your analysis will be skewed. Bill James himself acknowledges this in his own ranking system.
   136. KJOK Posted: September 26, 2005 at 08:40 PM (#1643767)
I don't give him much value points for his post-1922 career, but he does get something to tack on to his resume (however negligible it is). If that's what your system is doing, then we agree.

Not sure. Almost all the players Sisler is up against played 15 plus years. Sisler had 6 GREAT years plus 1 very good year. Starting in his 8th best year and down, however, he clearly is WAY BEHIND vs. the 8th best year and lower of the player's he's up against.
   137. sunnyday2 Posted: September 26, 2005 at 08:47 PM (#1643790)
jschmeagol, this is re. your ballot but I didn't want to clog up the ballot thread.

How many years (games, PAs) of war credit are you giving Keller? He only missed 1.5, not 2, not 3.

Give him 200 games and 1000 PAs for that 1.5 years and he is BA eligible 7 times with 5600 PAs. He also had 6 partial seasons and 300 games of decline which are not as valuable as 2 150 game seasons.

Even then, best case he is short of Kiner by 2 really productive years and 600 PAs. Or, only counting the truly productive years, he's still short 1700 PAs, even counting 1.5 years of WWII XC.

In order to get Keller over Kiner IOW, you've probably got to give him 2 years of WWII XC instead of 1.5 and then pretend his last 300 games were all in 2 productive years.
   138. John (You Can Call Me Grandma) Murphy Posted: September 26, 2005 at 08:48 PM (#1643794)
Starting in his 8th best year and down, however, he clearly is WAY BEHIND vs. the 8th best year and lower of the player's he's up against.

Which I don't disagree with, Kevin. But when it gets to the point that his crappy post-1922 seasons whittle away his pre-sinus condition value, then I can't agree with the system (like TPI does with Pete Rose and Brooks Robinson).
   139. sunnyday2 Posted: September 26, 2005 at 08:49 PM (#1643801)
K,

Sisler is clearly better than Averill for 7 years, then equal in the 8th and only then behind (and not way behind). If Averill had another 5-6 average years I could see him eventually accumulating an advantage. But considering he only has 2 years when he was better, and then marginally better, than Sisler's declining years...well, to say he was better is a stretch. To say he was WAY better is a WAY stretch.
   140. Michael Bass Posted: September 26, 2005 at 09:50 PM (#1643955)
Sisler is clearly better than Averill for 7 years, then equal in the 8th and only then behind (and not way behind).

Clearly? I'd take Averill's 7 best years over Sisler's (though it is *very* close, enough so that on second thought I'd call them even through 7), and once you get to 8, Averill blows Sisler out of the water.
   141. sunnyday2 Posted: September 26, 2005 at 10:07 PM (#1643980)
Michael,

maybe you're a WARPie where I'm a WS guy. But if they're equal for 7 then Averill blows Sisler away, where does Sisler's 292-280 career WS edge come from (and this is after the defensive WS are figured in).

Maybe it comes from the time Sisler spent hangin' around after sinusitis, except that most of us say that time had no value at all, so that can't be it.
   142. Michael Bass Posted: September 26, 2005 at 10:27 PM (#1644013)
I have no explanation for WS, I don't really even use them. WARP1 prefers Averill in exactly the manner I described however. (And, FWIW, Averill also as mentioned above gets from me a year of prime credit for his last minor league season as well)

As for the WS lead, I agree with Mr. Murphy above. It comes from Sisler piling up a whole lot more worthless WS (is there *anyone* who thinks WS's replacement level is high enough?) than did Averill.

Give me

33-30-30-27-26-26-26-24-24-22

over

33-29-29-27-25-24-22-19-16-15

any day.

And yes, I know 1918/1919 need some adjustment (though a straight-line adjustment on 1918 is simply too much), but it's not enough to get Sisler over the hump.
   143. sunnyday2 Posted: September 26, 2005 at 11:03 PM (#1644057)
Let's at least look at the real numbers (and I don't know why a straight line adjustment for 1918 is inappropriate). There were relatively few MLers who didn't play that year, but those who did only got about 122 games. I mean, if you give Averill credit for a year in the MiLs, this especially only seems fair. So here is Sisler with 1918 and 1919 season-length adjustments.

Averill 33-30-30-27-26-26-26-24-24-22-11-1
Sisler 33-29-29-27.5-27-26-22-19-16-15-13-11-11-10

After 6 years they're 0.5 apart. Then Averill gets the edge for 4 years. But if you're gonna knock Sisler for hangin' around (at 13-11), then we should penalize Averill for hangin' around at 11-1. Sisler's 10 is in half a season (his 14 seasons are really more like 13.5), so at a fine rate. So what you're left with is one extra season compared to Averill at 11 WS.

Rather than in quality order, here is the comparison by age, beginning at age 22.

Averill x-x-x-x-PCLMLE-26-24-30-30-26-33-22-27-24-26-11-1
Sisler 10-25-29-27.5-26-33-27-29-x-11-19-11-16-15-13-8

Sisler is better at age 22-26 (Averill not in MLs) plus 27-28, Averill (by 1) at 29, Averill at 30 (Sisler is out), Averill at 31-38, though miraculously at age 37 playing with one eye Sisler only trails the declining Averill 11-8. Even counting Averill's 1 WS at age 38 he is better 10 times (2 of them by 1 WS) and Sisler is better 7 times.

Best 5 Sisler vs. Averill

Sisler age 22-26: 117.5-0 (or whatever Averill's PCL MLE is)
Averill age 30-34: 138-72 (including one year Sisler inactive)

So everybody is focused on Sisler's decline when the real story of Sisler's 292-280 WS edge is Sisler was a productive MLer at an age when Averill was not, by a wider margin.
   144. John (You Can Call Me Grandma) Murphy Posted: September 27, 2005 at 12:08 AM (#1644229)
So everybody is focused on Sisler's decline when the real story of Sisler's 292-280 WS edge is Sisler was a productive MLer at an age when Averill was not, by a wider margin.

I'll try this again, since I keep getting ignored when I post it :-):

WS/162 Games: Sisler 23.01 Averill 27.18

Yes, their number of Win Shares are close, but Sisler created his in 387 more games. Sisler created 4% more WS, but Averill's rate is 18% better. That's not even including his PCL seasons.
   145. karlmagnus Posted: September 27, 2005 at 12:29 AM (#1644268)
The guy with the shorter career will tend to have the higher rate stats. Also, WS underrates pre-1920 1B.
   146. sunnyday2 Posted: September 27, 2005 at 12:45 AM (#1644308)
Two comments, John.

1. Looking at the rate stats says to me that Averill was the better player "on average." Averill had a fairly normal career curve, so his "average" is reasonably descriptive of him as a player. Sisler never had an "average" year, so I'm not sure it's very enlightening to describe him as such.

2. If Averill is in the MLs instead of the PCL, his rate goes down. It seems that he gets double credit--high rate plus extra credit for PCL. In the real world one of them would have to give.

'Course as a result of all this discussion of Averill versus Sisler and Kiner and whomever, I kept the other guys where they were but moved Averill up. He's on my ballot for the first time. Just not as high as Sisler, Kiner, Medwick.
   147. John (You Can Call Me Grandma) Murphy Posted: September 27, 2005 at 12:52 AM (#1644318)
The guy with the shorter career will tend to have the higher rate stats. Also, WS underrates pre-1920 1B.

As you know, I don't disagree with your second sentence.

As for your first sentence, for Averill to have his rate go down to Sisler's, he would have to play quite a few games. This, of course, would increase his WS considerably so that he would easily dominate Gorgeous George in that category. IOW, Averill will still win the competition.

They both have roughly the same amount of WS, while Averill beats Sisler badly in rate. Your higher rate theory doesn't really apply here, while giving Sisler a first base boost in WS for his Deadball years is not going to make up the difference between the two, IMO.
   148. EricC Posted: September 27, 2005 at 12:56 AM (#1644322)
Given that OPS+ is not the optimal measure of a player's offensive contribution (for example, it overrates slugging-heavy players), and that defense does matter, I thought I'd compare Keller and Kiner's peaks and careers using various comprehensive measures. I gave war credit to the players as follows: Keller 247 games in 1944-45 at his 43-46 rate; Kiner 144 games in 1945, duplicating his 1946 season. In the following chart, Keller 41-45 peak is followed by Keller career, and Kiner 47-51 peak is followed by his career totals. WWII deductions are problematic, but most of Keller's playing time in 43-46 was in the year with no war (1946) and in 1943, with a relatively small war discount. In any case, WARP3 is corrected for WWII competition levels. PA is plate appearances; TPR-I is total player rating from Total Baseball I.

                  Keller        Kiner

PA               3168/5687    3345/6835
Win Shares        171/276      155/257
WARP1            53.9/85.6    48.6/79.0
WARP3            48.0/79.0    47.5/76.1
TPR-I            22.1/30.8    23.0/27.0


In nearly every advanced measure, war-credit Keller beats war-credit Kiner both in peak and career. And yet the Keller advantage is even bigger than it looks in the above chart. Zero win shares or zero WARP is worth than useless, yet Keller has more peak and career WARP and WS in fewer projected PA. The career length argument can often be valid, but, in this case, arguing that Kiner is superior because he had a longer career is kind of like arguing that a player who goes 200 for 683 in a season was a better hitter than a player who went 201 for 569 because the former player had more at bats.

As pretty as the HR titles are, I have to conclude that the notion that Kiner was greater than Keller is erroneous. What would Kiner's career have looked like if he had to play with the balata ball and lost more than a year and a half at his peak due to the war?

As an aside, WARP3 for Joe Gordon 1939-1943 and career are 48.0/78.6. By at least this one measure, Gordon had a greater peak and greater career than Kiner without even giving Gordon any war credit.
   149. Dr. Chaleeko Posted: September 27, 2005 at 12:59 AM (#1644334)
OK, new question.

Where's anybody think Cepeda and Coimbre fit into the mix? On the ballot? Just off? Fab 50? Hot one hundred?
   150. John (You Can Call Me Grandma) Murphy Posted: September 27, 2005 at 01:14 AM (#1644374)
2. If Averill is in the MLs instead of the PCL, his rate goes down. It seems that he gets double credit--high rate plus extra credit for PCL. In the real world one of them would have to give.

Here's Brent's MLE's for Averill, Marc.

Year BWS fWS WS
1926 19.8 05.1 25
1927 15.9 05.2 21
1928 25.1 05.5 31
Total 60.8 15.8 77

Adding these WS increase Averill's total to 357.

Brent gave him a total of 439 games for his PCL years, so that will now give us a total of 2107 for his entire career.

Averill's rate would now be 27.44. His rate would actually go up slightly, while his WS skyrocket, if we add in his PCL numbers.

Averill now destroys Sisler in both categories when we factor in the former's minor league stats.
   151. Chris Cobb Posted: September 27, 2005 at 01:52 AM (#1644514)
One problem with assessing Earl Averill using win shares is that it is hard to say exactly what his defensive rating means.

Sunnyday2 claimed, somewhat hyperbolically:

Just about every decent CF is an A+ and just about every corner OF who isn't named Manny Ramirez is a C.

So I thought I would do a little study to try to give us a bit more traction with the win shares evaluation of centerfielders' defense relative to other centerfielders, esp. with reference to Earl Averill.

To try to get a sense of the spread of defensive value for centerfielders and Averill’s place therein, I identified the centerfielders in the fielding win shares outfielders table in James’ _Win Shares_ book, focusing on the centerfielders whose playing time was roughly similar to Averill’s. James estimates Averill as having played 13661 innings in the outfield, so as an upper bound I took 16,000 defensive innings, and as a lower bound 11, 322 defensive innings, which is as far below Averill’s total as 16,000 is above. This works out to a range of 8-12 154-game seasons in the field. There are 136 outfielders in this range overall. Of those 50 are identified by James in NBJBHA as primarily centerfielders, though many probably spent some time at other outfield positions.

Of those 50, 29 played a significant portion of their careers before 1961 (our historical present) and 22 were, like Averill, pre-WWII players.

Averill’s 3.99 ws/1000 defensive innings ranked 15th out 50, 10th out of 29, and 7th out of 22. In this cohort of players, considered for all-time, to the historical present, or to Averill’s time, he consistently ranks as better than about 70% of the centerfielders and worse than about 30% of them.

By this study, he certainly looks like an above-average centerfielder.

It is harder, of course, to rank him against players with significantly longer careers. They could have been better than Averill over a 13,000 inning prime, but their career rate may be lower than his. For the record, there are 18 centerfielders with more than 16,000 defensive innings listed in James. Of those, 5 have a career rate higher than Averill’s (Mays, Speaker, Carey, Otis, and Willie Wilson). Even if we assume all 18 of these players were better defenders than Averill (an excessively conservative assumption), Averill would rank 33rd out of the 68 centerfielders with at least 8 seasons worth of defensive innings: he still is above average within that group.

So win shares evaluates Averill as an above average defensive centerfielder.

This study also provides a way to calibrate win shares’ letter grades for centerfielders with respect to positional average (at least for long-career players). Of the 50 players with between 11,322 and 16,000 defensive innings, here is their breakdown by letter grade:

19A+
7 A
9 A-
5 B+
7 B
1 B-
2 C+

If the 18 long-career centerfielders are added in, those totals become

25 A+
10 A
12 A-
9 B+
7 B
3 B-
2 C+

By this study, it looks like a centerfielder with a career letter grade of A or A+ is above average for a centerfielder, while a center-fielder with a career letter grade of A- or lower is below average. I wouldn’t trust this rough estimate as the basis for a single-season zero-point for average, since a lot of the players whose careers are considered here had playing time at other positions, but it’s a start.

One last tidbit. James lists 490 outfielders in his table: 58 above 16,000 innings (18 cf), 136 between 11,322 and 16,000 (50 cf), and 296 between 6695 and 11,321. I haven't gone through and identified the centerfielders in this group, but we can safely assume that the proportion is about 1/3, or 98. A hand count shows that 16 of the 296 outfielders in this innings range have a fws rate higher than Averill's (4.00+), or I estimate, 16 of 96 centerfielders. This is a lower percentage than for the players with more innings, suggesting that Averill is probably, by win shares, around the bottom of the top 25% of all centerfielders, as win shares sees it. This lesser performance by shorter career players suggests that the rule-of-thumb assessment of centerfield defense might go overall like this:

A/A+ above average
A- around average
B+ and lower below average

That's how it looks to me, anyway.
   152. Howie Menckel Posted: September 27, 2005 at 02:34 AM (#1644654)
Kiner was just telling a story on the Mets game, wish I saw the whole thing.
They had a picture up of HOM candidate Chuck Klein, and Kiner was talking about a corrugated fence (Baker Bowl?). Pitcher Beck was complaining about getting taken out of a game, and threw the ball against the fence - making a "Boom Boom" sound, which is how he got his nickname.

Wish all you guys could hear ol' Ralph tell tales..
   153. jimd Posted: September 27, 2005 at 02:53 AM (#1644727)
IOW I remain unconvinced that corner OF defense really matters much.

One ball not caught on defense that goes for a single (either due to error or lack of range) cancels one offensive single. One offensive single is approximately one point of OPS+. One tenth of a point of range factor is about 15 hits (for a full-time player). Like all things in baseball, little things add up over the long season.

If you're hitting 200 OPS+, you can mess up a lot of flies, but they still lower your value. If you're hitting 130 OPS+, two tenths of a point of RF (30 points of OPS+) drops you close to replacement.
   154. Howie Menckel Posted: September 27, 2005 at 03:19 AM (#1644844)
jimd,
I wish RF measured defense as accurately as that.
But I don't believe it, or any other defensive stat, does that.
   155. Chris Cobb Posted: September 27, 2005 at 03:26 AM (#1644882)
I wish RF measured defense as accurately as that.
But I don't believe it, or any other defensive stat, does that.


The metrics may not be that accurate, but jimd's point about the value of a hit being the same offensively and defensively is incontestable. We should treat fielding metrics with healthy skepticism, but the attitude that "corner outfield defense doesn't really matter" is clearly mistaken.
   156. sunnyday2 Posted: September 27, 2005 at 03:45 AM (#1644948)
The big problem I have with defense (even by corner OFers) is not that it is unimportant but, as Howie says, that it is exceptionally difficult to tell who is good/better/best using the available numbers and analysis.

Given as I am to hyperbole, let me hyperbolize it this way:

1. We tend to make evaluations of defensive prowess mostly as the inverse of hitting prowess, with a touch of body type thrown in, and

2. Then we tend to add defensive performance on top of uber-stats, so the defense is double counted (or double mis-counted per hyperbolithis #1 above [that's a hyperbolic hypothesis]).

(Of course, through much of the 19C it should be double counted, even if it is indeed mis-counted and probably more "mis" than as we get into the 20C.)

That is my take on much of the defensive discussion, not just here at HoM of course. And that is why I tend to ignore defense numerically most of the time, though in the 19C that wasn't the case and in the 20C I just try to compensate by being sure that I have some players on my ballot from the right side of the spectrum.
   157. Brent Posted: September 27, 2005 at 04:21 AM (#1645079)
In # 137 Sunnyday2 wrote regarding Keller:

Give him 200 games and 1000 PAs for that 1.5 years and he is BA eligible 7 times with 5600 PAs.

Not that it matters much, but with an extra 200 games during 1944-45, Keller would have been eligible for the BA title 8 times (1939-46), not 7. For 1939, Keller would have been eligible for the BA title under both the standard that was used at the time (100 games) and the standard that is used now (3.1 plate appearances per team game, or 477 plate appearances in a 154 game schedule).

Regardless of his eligibility for the BA title, however, Keller's 1939 season was surely better than either of Kiner's 1946 or 1954 seasons, so I don't see that the fact that he missed a few games that year should hurt him in the comparison with Kiner.
   158. Mark Shirk (jsch) Posted: September 27, 2005 at 04:46 AM (#1645169)
Doc,

I mentioned when I posted my ballot that I don't have eithe rin my top 50 because ther are huge holes in their careers. Unlike most NeL players where we do ahve some numbesr, there are huge swaths of time where we have nothing. I don't feel comfotable simply going back and extracting those parts of their careers. However, if and when we do get soem data I am willing to put them on my ballot.
   159. Howie Menckel Posted: September 27, 2005 at 01:07 PM (#1645355)
No argument from me on that statement, Chris Cobb...
   160. KJOK Posted: September 27, 2005 at 04:56 PM (#1645864)
Sisler/Averill - Another View

20 Year MLB Positional Analysis:

1B - 1910/20:
George Sisler
Jack Fournier
Jim Bottomley
Ed Konetchy
Joe Judge
Jake Daubert
Lu Blue
George Kelly
George Tioga Burns
Stuffy McInnis
Hal Chase
Fred Merkle
Fred Luderus

1B - 1920/30:
Lou Gehrig
Jimmie Foxx
Hank Greenberg
Bill Terry
George Sisler
Jack Fournier
Dolph Camilli
Jim Bottomley
Hal Trosky
Joe Judge
Phil Cavarretta
Lu Blue
Ripper Collins
George Kelly
George Tioga Burns

CF - 1920/30:
Max Carey
Earl Averill
Hack Wilson
Edd Roush
Wally Berger
Earle Combs
Ben Chapman
Cy Williams
Sam West
Lloyd Waner
Pepper Martin
Baby Doll Jacobson
Doc Cramer

CF - 1930/40:
Joe DiMaggio
Earl Averill
Wally Berger
Ben Chapman
Dom DiMaggio
Stan Spence
Sam West
Lloyd Waner
Andy Pafko
Pete Reiser
Pepper Martin
Barney McCosky
Terry Moore
Vince DiMaggio
Doc Cramer
   161. jimd Posted: September 27, 2005 at 06:43 PM (#1646171)
I wish RF measured defense as accurately as that.

I wish that it did too. It's subject to all of the illusions found in OPS and ERA, plus other illusions.

Both Win Shares and WARP attempt to transcend RF. WS is very conservative, refusing to go where the evidence indicates whenever it thinks that the result is too extreme (there are caps on calculations almost everywhere). WARP is the reverse. IMO, the truth lies in between (though closer to WARP).
   162. sunnyday2 Posted: September 27, 2005 at 06:49 PM (#1646186)
K, your positional analysis needs a few words of explanation. Earl Averill was the #2 CF of the 1920s? He only played one year of the '20s? So that can't be what you meant.
   163. KJOK Posted: September 28, 2005 at 05:50 AM (#1648135)
K, your positional analysis needs a few words of explanation. Earl Averill was the #2 CF of the 1920s? He only played one year of the '20s? So that can't be what you meant.

Sorry, meant out of CF's who played primarily in either the 20's or 30's (20 year period), Averill had the 2nd best career (behind only Max Carey).

Also, out of CF's who played primarily in the '30's and '40's 20 year period, he had the 2nd best career behind DiMaggio.
   164. Joey Numbaz (Scruff) Posted: September 28, 2005 at 11:58 AM (#1648184)
Hey guys, great discussion this week - I'm still back around post 120 or so . . . but, just in case I don't finish . . . .

I'm going to be out of town for the next two weeks, I'll be leaving tomorrow morning, won't be back until October 13. I think I'll have internet access, but not sure that I'll have much time. I'll get my ballot in tonight.

I cede any and all authority to John if I can't be reached for any urgent matter that may come up. Best to email me and alert me to something if you think it's important.

**********

Getting to ballot stuff, I definitely agree with Chris on the pitchers. I have Ruffing and Rixey and Griffith placed quite prominently on my ballot.

Good discussion on Charley Jones, I have nothing to add, my position is out there, but it was an interesting conversation.

I'll take another look at Sisler before I vote tonight.

In exchange, I'd like everyone to take another look at Jake Beckley, who would have about 3400 hits with great power if seasons were 162 games in his day. Very good for a very long time has enormous value.
   165. KJOK Posted: September 28, 2005 at 03:55 PM (#1648516)
Sorry, meant out of CF's who played primarily in either the 20's or 30's (20 year period), Averill had the 2nd best career (behind only Max Carey).

Also, out of CF's who played primarily in the '30's and '40's 20 year period, he had the 2nd best career behind DiMaggio.


Now that I think about it, maybe it makes more sense to look at one 30 year period with the player's career centered in the middle instead of two, 20 year overlapping periods....
   166. Daryn Posted: September 28, 2005 at 05:34 PM (#1648724)
I just wanted to let Wes Ferrell fans know that I have re-evaluated him and he is going from 25 to top 10 on my ballot next year. If I recall correctly, Ferrell was the guy who had some extreme advocacy on his behalf. I think that must have turned me off his case, because I am debating putting him as high as number 4 on my ballot now. A 100 OPS+ is pretty rare and pretty valuable for a pitcher.
   167. Joey Numbaz (Scruff) Posted: September 29, 2005 at 12:33 AM (#1650011)
"Now that I think about it, maybe it makes more sense to look at one 30 year period with the player's career centered in the middle instead of two, 20 year overlapping periods.... "

I would go even further Kevin. I would tack 8-12 years onto the beginning and end of the player's career, to really minimize the selective endpoints issue . . .

So for a player who played 1915-30 let's say, I'd look anywhere from 1907-38 to 1903-42 to really get a good idea for how he fits in.

Of course, by making the period that big, you run into the issue of crossing eras that aren't similar, etc..

It's much more art than science when ranking these players. How you decide to weigh different factors can make all the difference in the world when splitting hairs, so IMO you really need to look at different things. Bill James makes a great point about this in the Cobb/Mantle discussion in the NHBA. If you only look at 5-year peak or 10-year prime or whatever, you are going to seriously overrate anyone who just barely meets your threshhold.
   168. KJOK Posted: September 29, 2005 at 06:02 PM (#1651539)
would go even further Kevin. I would tack 8-12 years onto the beginning and end of the player's career, to really minimize the selective endpoints issue . . .

Well, except that then it brings in all kinds of issues about the player's your comparing AGAINST. For example, Musial's last year is within 12 years of Jack Clark, or Early Wynn's last year is within 12 years of Dennis Eckersley, but I don't really think they should be compared as "contemporaries". The goal is to try to compare a player with contemporaries at the same position while ALSO using a large enough timeframe to minimize issues of temporary positional scarcity, like SS's in the 1920's or 1980's..
   169. Joey Numbaz (Scruff) Posted: September 30, 2005 at 12:58 PM (#1653548)
Agreed Kevin - I wasn't saying to compare every player that played a game in that period - I was saying look at players who played the bulk of their career in that period . . . sorry for any confusion.
   170. Paul Wendt Posted: September 30, 2005 at 02:28 PM (#1653671)
Here is the distribution by principal decade for the A+ fielders at each position, rated by Bill James. 20th century only. No one classified as a left or right fielder is graded A+ so I have given the decade-distribution of {A A- B+} handful.

A+ fielders by decade
[u] principal decade [/u]
    0 1 2 3 4 5 6 7 8 9 
'0' means 1900-1909'1' means 1910-1919etc  

C   0 1 0 1 0 0 0 0 0 2 
:  top 4 by DWS/k

1B  0 0 0 1 0 0 0 0 1 3 
:  top 5 by DWS/k  

2B  0 0 2 2 0 1 1 0 1 1 
:  top 4 by DWS/
(the four more recent guysplus #9,11,14,19 

3B  3 0 0 0 0 0 1 0 0 0 ;  top 3 by DWS/
(the three 1900s guysplus Clete Boyer

SS  2 3 0 1 2 1 0 3 1 0 
13 top 13 by DWS/k

LF  no A
fielders
    3 0 0 1 0 0 0 0 0 0 
;  top 4 by DWS/kgraded A A AB+

CF  1 2 1 3 5 1 2 3 2 4 24 top 23 by DWS/k plus #26 

RF  no Afielders
    0 0 0 0 1 0 0 1 1 2 
;  top 5 by DWS/kgraded A AABB


Note the left fielders at the beginning of the century and right fielders at the end of the century, indicating a change in the demands of the two positions that is commonplace here.

Half of the corner outfielders played significantly in centerfield. Ordered by their ratings, with CF share of outfield time (unlisted means about 1%), here are the nine nominal corner men with A to B+ ratings.

LF: Al Simmons (35%), Jimmy Sheckard, Fred Clarke, Sam Mertes (20%)

RF: Larry Walker (4%), Reggie Smith (50%), Jesse Barfield (6%), Tommie Holmes (35%), David Justice

Here is the distribution by decade for the five relatively pure corner outfielders. It sharpens the point about the changing demands of LF and RF.
LF2 0 0 0 0 0 0 0 0 0
RF
0 0 0 0 0 0 0 0 1 2 


Justice played right for six seasons, thereafter about 60% left 40% right. I don't know whether the letter grade is based on his full career. Anyway, he's the B+. In other words,
corner outfielders with long 20c careers
A+: none
A : Jimmy Sheckard LF, Larry Walker RF
A-: Fred Clarke LF, Jesse Barfield RF
B+: David Justice RF (73% LF 27%)

Relative to others at the same fielding position, B in the corners corresponds to A+ in center. Of course, the center fielders are much better fielders on average.
   171. sunnyday2 Posted: September 30, 2005 at 03:02 PM (#1653747)
One of my "rules" is I give extra credit to players whose entire cohorts are disadvantaged in some way--they are black prior to 1947, they fought in WWII, etc. (Well, I shouldn't say etc. because that is pretty much it.)

Now we had thought and Paul has confirmed that there is another cohort that is uniquely disadvantaged and that is corner OF. If Joe Blow is a corner OF there is almost nothing he can do to make himself an A, much less an A+, fielder.

What to do?

Well I am not going to argue that corner OF are as skillful (though skills, per se, are not really relevant here) or as valuable as CFs. But it seems to me that if we compare a corner OF (like Medwick or Kiner) to a CF (like Averill) using WS, that 1) the difference between the two in defensive value is likely to be exaggerated, and 2) the corner OF is being "punished" for a managerial decision that he should be in the corner, in just the same way that Averill is "punished" for the managerial decision that he should be in the PCL.
   172. Mark Shirk (jsch) Posted: September 30, 2005 at 03:34 PM (#1653839)
Sunny, I dont' think that Kiner, Medwick, et al. should get a boost because they cannot be A+ fielders under this system. What should happen is that the letter grade system should be recalibrated for corner outfielders instead of having them lumped in with CFers. They will still have fewer WS per 1000 innings, season, etc. It won't change anything really, tt would just makes things easier on us because then every position would be on the same scale.

Also, that Averill was an A+ CFer, means that he was one of the top 24 defense CFers of last century. That isn't as impressive as an A+ guy anywhere else, but it still says he was a real asset out there. Can the same be said of Kiner or Sisler?
   173. Chris Cobb Posted: September 30, 2005 at 03:38 PM (#1653842)
It's clear that using one letter grade scale for all outfielders makes it hard for the top corner guys to get the A's they deserve relative to their peers. The same would be true if SS/2B/3B were all graded on the same scale: the SS would get all the glory.

All this discussion helps us to understand how to interpret outfielder defensive letter grades as part of a simple characterization of what kind of a player somebody was.

Paul's excellent data gives us the numbers that we would need to recurve the outfielders' grades to give a separate set to centerfielders and corner men so that we could interpret their letter grades as comparing them to their positional peers and showing their relative greatness in comparison to their peers.

All this new information about the letter grades tells us _nothing_, however, about whether actual _numerical_ fielding win shares exaggerate the difference in value between center fielders and corner outfielders. We have presented no evidence whatsoever that center fielders are getting too many fws or that corner men are getting too few.

If your system pays attention to fielding letter grades and not to the numbers, then, some adjustment for the outfield situation is needed. If your system pays attention to fielding win shares, however, there is as yet no evidence that the win shares system consistently exaggerates the difference in defensive value between centerfielders and the guys on the corners.
   174. sunnyday2 Posted: September 30, 2005 at 04:12 PM (#1653919)
Chris, right. A putout is a putout, an assist is an assist. If a CF gets more then he is more valuable.

We were talking about letter grades however, and I just thought the letter grades (taken alone) may exaggerate the differences in skills or even value from the corner to the middle.

Also I've said before that I sometimes hear this type of argument:

Sisler 292 WS Averill 280 but Averill was a better fielder and that moves him ahead of Sisler.

The defensive WS and letter grade are already factored into the 292-280. So it's a question of what the various data point mean.
   175. KJOK Posted: September 30, 2005 at 04:29 PM (#1653965)
Yes, to be clear, lumping RF/LF with CF is no different than if James had lumped 1B with SS, then assigned letter grades. If he had done so, even relatively poor SS's would get B+ grades, while even great fielding 1Bmen would get B- grades, but that wouldn't impact AT ALL how many Win Shares they're receiving for defense...
   176. Paul Wendt Posted: September 30, 2005 at 06:02 PM (#1654198)
Working without guile, I really was interested in the distributions of high-grade fielders by decade, not in the number of high-grade fielders at each position. At the tiptop by Bill James letter grade, the LFs and 3Bs are from the beginning of the century, the left fielders and thirdbasemen RFs and 1Bs are from the end of the century, and the 2Bs tend to be from the second half of the century.


Chris Cobb
Paul's excellent data gives us the numbers that we would need

I should have said that KJOK provided and probably compiled the basic data including assignment of each player to a decade.

to recurve the outfielders' grades to give a separate set to centerfielders and corner men so that we could interpret their letter grades as comparing them to their positional peers and showing their relative greatness in comparison to their peers.

comparison to their transpositional peers?

If the letter grades were "recurved" to have the same distribution at each fielding position, I wouldn't use them as a component of transpositional relative greatness (and the distribution at different positions may differ without affecting positional peers). With possible exception of Fred Clarke and predecessors, a left fielder would be a center fielder if he were a better fielder.
   177. karlmagnus Posted: September 30, 2005 at 06:42 PM (#1654304)
This surely means that counting fielding WS and the fielding grade is counting double, when comparing fielders from different positions. In other words, if M is a A- center fielder, and N is a C+ first basemen, they may acutally both be at say the 35th percentile (0 is best) at their respective fielding positions, so should get no additional bonuses/penalties beyond those assigned by WS to the fielding positions.

Also, in a case such as 1880-1920 1B, where WS gets it wrong, a 1B at the 35th percentile of 1B is acutually much MORE valuable than WS admits him to be.

In concrete terms, if Sisler and Beckley's C+ (or whatever) puts them at the 35th percentile of 1B, and Averill's A- puts him at the 35th percerntile of CF, and WS takes the CF/1B differential as at 2005, or even 1955, then Sisler and Beckley dsserve a substantial bonus compared to Averill, because 1B in 1900 or 1915 was much more important relative to CF than WS, based on 1955/2005, admits it to be.

You can agree or disagree with me in two ways, (i) fault my logic, or (ii) tell me my percentiles are wrong for Averill/Sisler/Beckley. I have no access to information that would confirm the latter, but I believe the logic to be correct.
   178. KJOK Posted: September 30, 2005 at 07:01 PM (#1654335)
Karl - I'm not sure I follow you enough to determine if your logic is faulty, so let's use a concrete example

Sisler's defensive win shares are 24.6 career, 1.35/1000 Innings, and his letter grade is C-.

Beckley's defensive win shares are 38 career, 1.82/1000 innings, and his letter grade is (I think, don't have it here) B.

Averill's defensive win shares are 54.5 career, 3.99/1000 innings, and his letter grade is A+.

Now, the letter grade, as far as I'm concerned, is irrelevant. What I might argue, however, is that because win shares does not apportion enough of the TEAM DEFENSIVE WIN SHARES to 1st basemen during the careers of Sisler and Beckly, that maybe Sisler should have 30 career Defensive Win Shares and Beckly 45 Defensive Win Shares, while during the career of Averill too many TEAM DEFENSIVE WIN SHARES are apportioned to CF'ers, so his career defensive Win Shares (adjusted) should only be 52...(adjusted numbers just made up for example purposes...)
   179. karlmagnus Posted: September 30, 2005 at 07:12 PM (#1654353)
KJOK, yes, that looks right. I'd forgotten Averill was an A+, in which case he's obviously in the top x percent (whatever x is). The problem is more acute for a CF who was an A, A- or B, in which case even though he would appear a better fielder than Sisler/Beckley (and if catching fungoes or some other "level playing field" activity, might intrinsically be) could nevertheless be a worse fielder in terms of his percentile ranking among position contemporaries.

I suspect that no 1900 fielder would rank very well if dumped by time machine into 2005, because of lack of training/unfamiliarity with new equipment, but a 1900 1B might well if dumped into 2005 be able to field 2B, 3B or even SS just as well as he fielded 1B.
   180. Dag Nabbit is a cornucopia of errors Posted: September 30, 2005 at 09:06 PM (#1654733)
Well, I went through the Win Shares book and

1) typed in all 490 OFrs listed in the back, their FWS, and their IP. Then I let excel figure the exact FWS/1000 IP (sure beats typing it in).

2) Then went through the 'cyclopedia and put in their primary OF position for all 490, and what their main decade was at the time the WS book came out (in other words through 2001, so there's no 2000s decade guys).

The results for average FWS by position by decade (note - some decades, especially the 1880s and 1940s have very small sample size. Half the CFs from the 1940s were all named "DiMaggio" for example):

.......RF....CF....LF
1880s
..3.28..3.62..2.84
1890s
..2.33..3.47..3.05
1900s
..2.40..3.60..2.76
1910s
..2.54..3.46..2.78
1920s
..2.50..3.07..2.46
1930s
..2.54..3.73..2.65
1940s
..2.54..4.11..2.54
1950s
..2.53..3.43..2.46
1960s
..2.62..3.68..2.24
1970s
..2.52..3.25..2.36
1980s
..2.57..3.30..2.38
1990s
..2.41..3.48..2.51
[b]TOTAL
..2.53..3.46..2.54[/b] 


Earl Averill comes off like a good CF, but not A+. Pete Browning was a butcher. Medwick was an above average guy at his position. Bob Johnson below average. Edd Rousch & Hugh Duffy come out about the same - both very good, but neither had the best rate of his day. Leach is the only CF under serious consideration who'se listed as the best at CF position for his decade. Van Haltren & Ryan both had below average rates for a CFr. Keller was average. Veach a very good fielding LFr. Klein was very bad. Ditto Hack Wilson. Babe Herman even worse. (Note this is all overlooking time spent in other, secondary outfield positions, and avoiding any differences between career rate and peak/prime defensive rate).
   181. Dag Nabbit is a cornucopia of errors Posted: September 30, 2005 at 09:18 PM (#1654769)
Here's a breakdown of grades by position:

....RF..CF..LF
A
+..1...43..1
A
...2...18..4
A
-..3...27..6
B
+..8...21..5
B
...12..22..14
B
-..23..16..20
C
+..30..11..27
C
...32..6...31
C
-..17..5...25
D
+..22..3...16
D
...3...0...8
D
-..4...0...2
F
...1...0...1 

My alma mater had a GPA system where a + grade was +0.3 extra and a - was -0,3 (in other words 3.3 B+, 3.0 B, 2.7 B-) using that system (only unlike ol' siwash making A+s 4.3), here's the average GPA by position:

RF - 2.10
CF - 3.41
LF - 2.20

FWIW, the one A+ RF was Hugh Nicol, who is also the reason why the average score for 1880s RFs is so dang high.
   182. Paul Wendt Posted: June 01, 2007 at 10:38 PM (#2387716)
This Ballot Discussion thread lacks the usual list of newly eligible candidates.
   183. John (You Can Call Me Grandma) Murphy Posted: June 01, 2007 at 11:19 PM (#2387792)
Paul, that info was tossed into the Summary section of this thread in the ExpressionEngine. This must have happened with one of those annual changes here at BBTF, since that info was originally there. There would have been a number of posts asking where it was if it hadn't been there at the time.
   184. Dr. Chaleeko Posted: June 02, 2007 at 12:45 PM (#2388763)
in the ExpressionEngine.

This must be a software of somesort, but it makes me feel like maybe we need a little plutonium for the flux capacitor.
   185. John (You Can Call Me Grandma) Murphy Posted: June 02, 2007 at 06:37 PM (#2389112)
This must be a software of somesort, but it makes me feel like maybe we need a little plutonium for the flux capacitor.


Dr. Chaleeko=Doc Brown? ;-)
   186. Paul Wendt Posted: June 02, 2007 at 10:49 PM (#2389804)
182. Paul Wendt Posted: June 01, 2007 at 06:38 PM (#2387716)
This Ballot Discussion thread lacks the usual list of newly eligible candidates.

I think I have everything I was looking for, mainly year eligible for several players. I'll use the current Results thread and the yahoo group when I have data to go.
   187. Dr. Chaleeko Posted: June 03, 2007 at 12:13 AM (#2390092)
Dr. Chaleeko=Doc Brown?

You know, John, in 2000 an expressionengine might be available at your corner drug store, but in 1985 it's a little hard to come by!!!!
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
James Kannengieser
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.7738 seconds
43 querie(s) executed