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

Saturday, September 29, 2007

2005 Results: Boggs Gets 100%, While Browning and Dawson Receive Hall of Merit Honors, Too!

In his first year of eligibility, legendary third baseman Wade Boggs received 100% of all possible points to become the 14th unanimous selection in Hall of Merit history (past unanimous selections include Hank Aaron, Joe DiMaggio, Lou Gehrig, Lefty Grove, Walter Johnson, Mickey Mantle, Stan Musial, Willie Mays, Babe Ruth, Mike Schmidt, Honus Wagner, Ted Williams and Cy Young).

It was a “slightly” longer wait for star batsman Pete Browning as he was inducted into the HoM in his 107th year on the ballot. He received 28% of all possible points.

Last but not least, All-Star outfielder Andre Dawson claimed the final spot for immortality in his 4th year of eligibility, narrowly besting fellow outfielders Bob Johnson and Alejandro Oms by only a handful of points (the latter two appear to be favorites to enter the HoM themselves in 2006). He earned 25% of all possible points.

Rounding out the top-ten were: Reggie Smith (huge jump!), Bucky Walters, Cannonball Dick Redding, Kirby Puckett (surprising finish after his 2004 showing) and Gavvy Cravath.

Thanks to OCF and Ron for their help with the tally.

RK   LY  Player                   PTS  Bal   1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
----------------------------------------------------------------------------------------
 1  n/e  Wade Boggs              1296   54  54                                          
 2    4  Pete Browning            364   22      6  2  2     3  2  1  3     2           1
 3    7  Andre Dawson             326   23      3  2  2  1  2  1     2  4  2  2     1  1
----------------------------------------------------------------------------------------
 4    6  Bob Johnson              322   25         2  5     2  2  4     2  2     2  1  3
 5    8  Alejandro Oms            316   25      3  2  1  1     1  2  3  2     1  4  1  4
 6   14  Reggie Smith             279   20      1  2  2  2  3     1  3  1  2     1  1  1
 7   10  Bucky Walters            278   18      1  3  3  2  1  2     2  1  1  2         
 8    9  Cannonball Dick Redding  273   15      5  3  1  1     1     2  1  1            
 9    5  Kirby Puckett            269   21      2  1  3  1  1  1     3  2        2  1  4
10   13  Gavvy Cravath            261   21      2  1        2  1  3  2     4  2  1  2  1
----------------------------------------------------------------------------------------
11   12  Tony Perez               261   17      1  5  1     3     1        2  1  3      
12   11  Hugh Duffy               241   16      1  3  1  1  2  2     1  1  3     1      
13   15  Tommy Leach              239   16      1  2  1  3     3  1  2     1  2         
14  n/e  Bret Saberhagen          237   18      2  1  1  2  2     1  1  2     1  1  2  2
15   17  Luis Tiant               232   20      2     1     1  1  2  4        2  1  4  2
16   18  Graig Nettles            232   19         1     1  2  4  3  2        2     3  1
17   19  Phil Rizzuto             211   15         2  2  2  1     2  1  2  1  1     1   
18   22  Ken Singleton            207   18         2  2     1  1        1  2  3  2  2  2
19   21  George Van Haltren       191   12      2  2     1  2  1  1     1     1  1      
20   23  Bus Clarkson             189   14      2     1  4     1  1           1  1     3
21   20  Dizzy Dean               189   12      3  1  1  1  1        1     2  2         
22   16  John McGraw              187   11      3  1  2  2  1                 1        1
23   26  Tommy Bridges            159   10      1  1  4     1     1  1                 1
24   24  Mickey Welch             158   11      1  1  1  1     3  1        1     1  1   
25   29T Burleigh Grimes          153   13      1     1     3              2  2  2  2   
26   28  Dave Concepción          140   11      1     1  1     1  2     1  1  1  1     1
27   32  Larry Doyle              137   10      2     2     1        1        1     3   
28   27  Vic Willis               135    9      1  1     1  1  1  1     2  1            
29   31  Dale Murphy              134   12               2        3  2  1     1     1  2
30   29T Orlando Cepeda           133   11            3     2     1           1  2  2   
31   34  Elston Howard            130   12               3              2  2  3     1  1
32   25  Lou Brock                130    9      1  1  1        1  2     1     1  1      
33   33  Rusty Staub              126   10            1     2     1  2  3     1         
34   37  Tommy John               125    8         3  1        1        2              1
35   36  Bobby Bonds              121   10         1     2           1  1  2  1  1  1   
36   38  Bob Elliott              114   10               1     2     1  2  1  2  1      
37   35  Norm Cash                113    9            1     1  3     1     2        1   
38   39  Ben Taylor               102    8         1        1  2     1     1     1  1   
39   48  Pie Traynor               98   10            1  1     1              1  1  4  1
40   41  Carl Mays                 90    8               2     1        1  1  1  1     1
41   42  Wally Schang              87    6      1     1        1     1  1  1            
42   45  Don Newcombe              86    8                  2     1     1     1  2  1   
43   50T Lee Smith                 82    5         2     1  1                       1   
44   40  Dave Bancroft             77    7               2     1              1  2     1
45   47  Vern Stephens             70    6                  1  1  2           1        1
46T  46  Chuck Klein               67    5         1     1                 2  1         
46T  49  Rick Reuschel             67    5      1           1           2           1   
48   44  Bill Monroe               65    5               2     1  1                    1
49   95T Bert Campaneris           64    5               2  1              1        1   
50   54  Ed Williamson             64    4         1  1           1  1                  
51   52  Sal Bando                 62    5                     2  2              1      
52   50T Frank Tanana              61    4      1        2                             1
53   62T Urban Shocker             60    6                     1     2           1  2   
54   43  Don Mattingly             59    5      1                 1              2  1   
55   61  Johnny Pesky              56    6                           1  1  1     2  1   
56   55  Addie Joss                52    4      1                 1           1     1   
57   53  Thurman Munson            49    5                  1           2              2
58   69T Leroy Matlock             49    4                     2  1              1      
59   71  Jack Quinn                48    4         1                       1     2      
60   60  Ernie Lombardi            46    4               1        1        1        1   
61   59  Wilbur Cooper             46    3         1        1                 1         
62   62T George J. Burns           45    4                        1     2  1            
63   58  Tony Oliva                45    3      1           1                       1   
64   65T Lance Parrish             44    4                     1     1  1           1   
65T  56T Frank Chance              40    4                     1     1           1     1
65T  72T Al Rosen                  40    4                              2     2         
67   67  Tony Mullane              38    3                  1  1              1         
68T  56T Buddy Bell                36    3               1     1                       1
68T  69T Rabbit Maranville         36    3            1                 1        1      
70   64  Lefty Gomez               35    4                  1                    1     2
71   81  Fred Dunlap               33    4                              1  1           2
72   75  Bruce Sutter              32    3                        1  1              1   
73   74  Frank Howard              30    3                           1  1           1   
74T  65T Ed Cicotte                30    2            1           1                     
74T  76  Jimmy Ryan                30    2               1     1                        
76   77  Bobby Veach               29    3                                 2  1         
77   68  Ron Cey                   28    4                                 1           3
78   78  Jim Kaat                  25    2                        1  1                  
79   84T Jim Rice                  24    3                                 1     1     1
80   72T Jack Clark                22    2               1                             1
81   82T Luke Easter               22    1         1                                    
82   80  Sam Rice                  21    2                        1              1      
83   84T Dave Parker               18    2                                    2         
84   82T Brian Downing             17    1            1                                 
85   86  Luis Aparicio             16    1               1                              
86   79  Bill Mazeroski            15    1                  1                           
87T  93T Tommy Bond                13    1                        1                     
87T  87  Sam Leever                13    1                        1                     
87T  88T Carlos Morán              13    1                        1                     
90T  91T Tony Lazzeri              12    1                           1                  
90T n/e  Virgil Trucks             12    1                           1                  
92T  91T Fielder Jones             11    1                              1               
92T  88T Hack Wilson               11    1                              1               
92T  88T Dizzy Trout               11    1                              1               
95T 100  Dick Lundy                10    1                                 1            
95T  95T Jack Morris               10    1                                 1            
97T  93T Brett Butler               9    1                                    1         
97T  95T Mickey Vernon              9    1                                    1         
99T  99  Elmer Smith                8    1                                       1      
99T  95T Jim Fregosi                8    1                                       1      
99T 100  George Kell                8    1                                       1      
99T n/e  Dolf Luque                 8    1                                       1      
103T 100  Charlie Hough              7    1                                          1   
103T 104  Bill Madlock               7    1                                          1   
103T 100  Gene Tenace                7    1                                          1   
103T n/e  Jim Whitney                7    1                                          1   
107T 104  Dutch Leonard              6    1                                             1
107T 104  Dennis Martinez            6    1                                             1
107T 104  Al Oliver                  6    1                                             1
Ballots Cast: 54
John (You Can Call Me Grandma) Murphy Posted: September 29, 2007 at 10:34 PM | 231 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 3 pages  < 1 2 3 > 
   101. Howie Menckel Posted: October 05, 2007 at 03:59 AM (#2561013)
Dan R,
That is a very good post.

I would say this, fwiw.

You are extraordinarily passionate, and remarkably knowledgeable.

I hope this helps, taking a small risk here, but as a lifelong independent, I think I can chance it.

"The positive question has basically been resolved: the overwhelming scientific consensus is that man-made greenhouse emissions are causing the planet to heat up."

I think that's just slightly off-center, and I raise it only because it may be useful for what you're trying to accomplish.

The overwhelming scientific consensus is that the planet is heating up.
Beyond that, there is a broad consensus that humans are causing it.

They're not quite the same statement.

You're recognizing the innate bias of the "head in the sand" crowd.
You're not realizing the opposite bias of "I am convinced either that the world is doomed or that capitalist pigs will ruin it soon," etc.

That does not negate the research findings. Somebody is right. But don't ignore what is there, either.

You have very intriguing conclusions in general.
But they might be wrong.

You're angry with dogmatic global warming denials, understandably so.
But that's in part because you don't like dogmatism.

Are you leaning dogmatic in your HOM views?
Maybe.
But as you show in your last post, you don't want to be.

Intellectually, I picture it as always being on your toes, ready to move in any direction.
You're a lot closer to "tiptoe" than people you disagree with, but you want to keep it that way.

Anyway, if this devolves into some political pro/con, please delete this and other responses. I'm not an advocate for any of that.
   102. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 04:32 AM (#2561082)
Howie, thanks very much for your response.

I didn't mention people who would be on the other extreme of the global warming pseudo-"debate" because they're not germane to the positive/normative distinction. I may disagree with their normative prescriptions--if they were to propose, say, banning all carbon emissions tomorrow or something ludicrous like that--but at least they are responding to the empirical scientific findings.

My conclusions--as in, I think such-and-such a player is a deserving HoM'er, or not--are by definition normative. I couldn't possibly say that David Concepción "is" a HoM'er--he isn't; he hasn't been elected. I think he should be. That is a normative statement. But I make the argument to support that normative conclusion with positive, empirically accurate statements: namely, that both replacement level for SS and league standard deviations were at an all-time low in the late 1970's. Voters are free to disagree with me--and most do--that the confluence of those factors means Concepción should be in the HoM. But if someone were to say that those factors weren't true, I would get annoyed. That was why I reacted--perhaps over-reacted--as I did to sunnyday's remark about 2B around 1910. If you want to say "it doesn't matter that 2B hit like 1B around 1910, I still think Doyle belongs," go right ahead. That's your prerogative. But don't say the defensive spectrum was something it wasn't.

As for whether I'm "leaning dogmatic in my HoM views," I'd only say that I'm a pretty mediocre dogmatist if I am. I've just recently come around to accept that in-season durability does have a certain value that mere longevity doesn't, because my rate-based salary estimator meant that I was effectively using a higher replacement level than the empirically correct one in my voting. I was sufficiently convinced by sunnyday--who I probably spar with more than anyone else on these boards--of the case for Don Newcombe to toss him a vote on my last ballot. My WARP system has benefited tremendously from the feedback and criticism on these boards--Dandy Little Glove Man's astute catch of a discrepancy in my standard deviation regression enabled me to increase the overall accuracy of my system by 4% (which is a ton). My plea for people to do their homework before posting is double-edged--it's not just to save me the trouble of fact-checking them, it's also for them to fact-check and enlighten me! I have learned more about the history, theory, and analysis of baseball from the HoM project than I ever could have imagined--and remember, I only re-entered the voting about 10 "years" ago. Part of the reason I am so passionate about this is that I am thankful to have gotten so much out of it, and hope to offer a few similarly valuable nuggets of insight to the rest of the group. Most of the time it feels like a pretty daunting uphill battle...but I feel tremendously rewarded every time another voter incorporates my findings into their analysis, *particularly* when doing so leads them to support candidates I don't! (Leach leaps to mind).
   103. Howie Menckel Posted: October 05, 2007 at 04:47 AM (#2561102)
Wow, great response.

I think any fear I had of sending this off the rails is eliminated by that answer.

Keep up the good work.

I like the provocative new stuff, actually, and mainly just fear it might get lost from the innately confusing interpetative alternatives.

And this from the oldiest of old fogies, at least voting-wise.
   104. baseball fanatic Posted: October 05, 2007 at 11:54 AM (#2561219)
"The positive question has basically been resolved: the overwhelming scientific consensus is that man-made greenhouse emissions are causing the planet to heat up."


If only scientists on that side of the argument could explain why other planets in the solar system were also warming up (which is indisputable) and still somehow prove that it is our fault, then maybe there would actually be a consensus on the subject and would convince the average person that our global warming is man-made and not caused by the Sun.
   105. TomH Posted: October 05, 2007 at 12:59 PM (#2561244)
(the joys of a 4-day weekend. Woo hoo!)

after thinking for some time about posts 97ff....

While it may be a positive (as opposed to normative) statement that the typical MLB replacement level is in the 75% to 85% of MLB average range, it is much less positive that this always applies, or applies over a long period.

Cleveland Spiders, 1899. A player who only got 3 WS on that team probably actually only DID contribute 1 win (3 WS=1 win) to that team.. whereas on other teams, and according to the 80% metric, he likley had BELOW 0 value. But he was better than the othr gugys the spiders could come up with. Similarly, Babe Ruth's 1921 may have only been worth 30-ish WS on that team; home runs aren't as valuable when you are usully losing by 5.

I don't say this to defend WS use of very very very low replacement level. Bill James himself has since gone back to re-work the whole scheme into Win and Loss Shares (as of almost a year ago).

Replacement level for the 1906-1909 Cubs and the 1995-2007 Yankees is higher than normal. To improve these teams, you better be at least average, or you aren't getting playing time. A young position player on the 75-76 Reds, where is he gonna go?

Some have also studied & concluded that over a very long period of time, replacement level rises a bit.

For the HoM, IN GENERAL I completely agree with DanR's 80%-of-avg take. I use WS minus 11 or 12 per full year (650 PA), or WARP minus 2.5 per full year. But there ARE other answers, depending on the issue at hand.
   106. karlmagnus Posted: October 05, 2007 at 01:15 PM (#2561254)
I agree with baseball fanatic. The overwhelming consenus on global warming appears to be the following:

(i) The planet has warmed in the years since 1970
(ii) Carbon emissions could cause warming, although we don't know to what degree.

Hawever whether carbon emissions actually ARE causing warming isn't established; it could well be a natural fluctuation perhaps caused by solar conditions as it's well within the bounds of such. The problem is bedevilled by "mistakes" like NASA proclaiming for years that the warmest year ever was 1998 when it was actually 1934.

Personally I'm in favour of a medium sized carbon tax on the precautionary principle ("cap and trade" is a license to lobbyists and thus a very bad idea.)

On the question of second basemen, the fact that they hit like 1B for a few years round 1910 is likely just a random fluctuation; it's far too small a sample to prove anything, given the variances between individual players. You have this blip in 1B hitting in 1885-95 too because ABC were 3/12 of the league, and there was Beckley as well. Where there's a "scientific consensus" is that 2B was a mainly bat position in the 1890s and a mainly glove position since 1930; it probably makes sense to assume a smooth curve of change between the two rather than a sudden shift for which there is no apparent explanation.
   107. 'zop sympathizes with the wrong ####### people Posted: October 05, 2007 at 01:37 PM (#2561275)

Hawever whether carbon emissions actually ARE causing warming isn't established; it could well be a natural fluctuation perhaps caused by solar conditions as it's well within the bounds of such.


Before I quit (before writing my dissertation, but after everything else), I was a PhD student studying climate change.

I can assure you, whether carbon emissions are causing warming is absolutely, 100% established. Already, in 2 posts about climate change in this thread, there were 2 wildly inaccurate statements. There's a whole bunch of misinformation floating around.

I'm also happy to take any questions --off line-- about global warming. Feel free to email me. I'm actually a -gasp- registered Republican, so if you're a skeptic I can point you in the direction of good normative policy discussions about how to deal with climate change.
   108. karlmagnus Posted: October 05, 2007 at 01:45 PM (#2561287)
With respect, 'zop a PhD doesn't impress me. Academia is notoriously closed to political incorrectness. In order to claim that something is "100% established" you have to include the views of highly credentialed skeptics like Patrick Michaels and Fred Singer. The climate science community is economically dependent on global warming being a major problem, so their reserach gets government and environmental group funding. Its research may well have merit, but it cannot be regarded as unprejudiced.
   109. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 01:54 PM (#2561304)
TomH, thanks for your response.

I'm afraid I still have a fundamental disagreement with you about this. The only way you can calculate that a player who got 3 WS on the 1899 Cleveland Spiders contributed exactly one win to the team is if you use a replacement level of precisely 52% of league average offense, which is, I repeat, a completely arbitrary and incorrect selection. Let's take a hypothetical guy who was 3 WS = 1 win in 1899. League average was .214 runs per batting out, so James's 52% baseline is .111 runs per out. Say the guy made 300 outs. If he was 1 win = 10.77 runs in 1899 above that baseline, then he created .111*300 + 10.77 = 44.2 runs. OK, so if you use James's 52% figure that's 3 batting WS. But if you used a baseline of 25% of league average offense, then he'd magically get 44.2 - (300*.214*.25 = 16.05) = 28.15 runs/10.77 runs/out * 3 WS/win = 7.8 batting WS. If you used a baseline of literal absolute zero, he'd have 44.2 runs/10.77 runs/out * 3 WS/win = 12.3 batting WS. And if you used the actual replacement level of 80% of league average offense, then he'd have negative batting WS: 44.2 - (300*.214*.8 = 51.36) = -7.16 runs/10.77 runs/out * 3 WS/win = -1.99 batting WS. (This, of course, would drive James crazy, since he doesn't assign negative WS, so he's forced to unfairly penalize this guy's teammates for his performance). Which of these results is correct?

See, there's nothing inherently "true" or "right" about the Win Shares allocation system. You could use James's approach of dividing actual wins among a team's players any way you want and come up with radically different distributions depending on what replacement level you set. To take sunnyday's example of Al Rosen in 1953, he only happens to be credited with 36.7 batting WS because James chose 52% as his baseline. If James had chosen 0%, Rosen would have only 26.7 BWS. If he had chosen 80%, Rosen would have a monster 54.6 BWS. It all depends on how much value you decide to put "under water." I think the right number is 80%, of course, and I could see a theoretical argument for 0% (although how would you apply that to pitchers? They could theoretically allow 50,000 runs per inning). But 52%? That's just plain strange...and...wrong.

Remember, using a proper replacement level does not "untie" your system from actual wins. If you want, you could simply say 50 wins is replacement level, assign half to defense and half to offense, and distribute those by playing time. So let's say a DH has exactly 1/9 of his team's plate appearances and has negative 1 WARP. Then you could credit him with 3*((50/2*9)-1) = 5.3 WS. That would still be a more conceptually accurate, and flexible, system than James's monstrosity. ;)

As for extreme teams like the 1899 Spiders or the Big Red Machine, I think it's safe to say that the Spiders were NOT combing the nation for the best freely available players they could find--they simply took the first warm bodies they could find to field a team. The entire team WAS well below the freely available talent level. And regarding the 1975-76 Reds, who would they have been forced to play if one of their stars had gone down? A replacement player, of course--one whose production would probably be approximately 80% of positional average.

I don't see what any individual team's won/lost record or roster has to do with the leaguewide freely available talent level.
   110. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 02:07 PM (#2561332)
Can we open another thread to discuss global warming? Maybe I shouldn't have opened that can of worms.

'zop, you're a registered Republican? I didn't know THAT...we may have to take this offline, jack@$$. ;)

Karlmagnus, 2B was NOT a "mainly bat" position in the 1890s, as measured either by worst-regulars or positional average. From 1893 to 1900, 2B positional EqA was lower than that of every other position besides C and SS. It was only from about 1906-15 that 2B was *really* a "bat position," hitting far better than 3B and equal to 1B as measured both by average and worst-regulars. And the presence of Collins and Lajoie has no effect on my replacement level calculation, since I only consider the worst 3/8 of regulars. Could this just be a random blip as you suggest? I'm not sure. I'm dealing with a sample size of 6 players per year times a 9 year moving average is 54 players. Can someone calculate the odds that the average of a group of 54 players would deviate from its "true" level by 1.1 wins, if the standard deviation among players is 2.73 wins?
   111. John (You Can Call Me Grandma) Murphy Posted: October 05, 2007 at 02:27 PM (#2561379)
Can we open another thread to discuss global warming?


Can we not? Please?! :-0
   112. karlmagnus Posted: October 05, 2007 at 02:35 PM (#2561391)
In the real world, Collins, Lajoie and Hornsby must have a real effect on the replacement level of 2B. As the manager/GM sits there glumly after seeing his prize pitcher decimated by one of those three, it must be natural to look down the bench at the light hitting 0-4 second baseman and wonder whether there's a replacement available. Conversely, if everybody else's SS in the '70s is hitting .215, Mark Belanger doesn't look so bad. The long term moves caused by changes in the way the game is played are difficult to detect amidst the "noise" of these kinds of fluctuations.

Grandma, I agree, let's not have a global warming thread. I have work to do!
   113. Juan V Posted: October 05, 2007 at 02:42 PM (#2561404)
But how does the high replacement level of 2Bmen affect global warming?
   114. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 02:44 PM (#2561410)
Karlmagnus, I suppose that's a plausible theory, but it would be helpful if you had any sort of evidence to support it. I'm not aware of any study that shows that the introduction of a few superstars at a position into the league causes an increase in the production of the worst regulars at the position X years later. It would be fascinating if that were the case and you could demonstrate it.
   115. karlmagnus Posted: October 05, 2007 at 02:48 PM (#2561423)
Don't have the data, but what about SS hitting since Ripken entered the league in '82, with an additional boost from Jeter/Nomar/A-Rod in '96-97? Your "worst 25%" data should reflect an upward trend following both of those events, without SS having become less of a "glove" position.
   116. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 02:54 PM (#2561438)
It's worst 3/8, not worst 1/4, but you couldn't be more correct. The average production of the worst 3/8 of starting MLB SS has absolutely skyrocketed, rising virtually every year from an average of 3.7 wins below overall league average per 162 games in the 1982-90 period to just 2.7 wins below overall league average in the 1997-2005 period. In today's game, SS and C replacement level are almost equal, something that hadn't been the case since the deadball era. Now, association is not necessarily the same thing as causation--many things could have caused the rise in SS rep level--and this is only a single data point. But you are right that the data is pretty compelling in this particular case.
   117. karlmagnus Posted: October 05, 2007 at 03:01 PM (#2561464)
Very interesting, thank you. Might be worth looking at 3B too -- was the replacement level already rising before Matthews, between Matthews and Brett/Schmidt or only after Brett/Schmidt? It may be that superstars can emerge from anywhere, but their example often changes the way the game is played by changing the image of the position in GM/manager/scout minds.
   118. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 03:38 PM (#2561530)
The change in 3B rep level was very rapid. From 1939-47, the worst 3/8 of starting 3B averaged 1.9 wins below average per 162 games. From 1948-59, they averaged 1.0 wins below average per 162 games. That does overlap with Matthews' entrance into the league, but the move really appears to have happened virtually overnight between 1947 and 1948 (even though I can't think of any prominent black 3B who entered the league at that time). In 1947 you had Lee Handley, a rookie Eddie Yost, and a motley crew of mediocrity on the Red Sox stinking up the league at 3B, which was typical of the preceding years, whereas in 1948 only Putsy Caballero was meaningfully below average--a much improved Yost, Billy Cox, and George Kell were the next-worst starters, and none of them played at a rate worse than 1.0 wins below average per 162 games. This was typical of the succeeding seasons as well. I don't know if anyone can point to anything specific that happened in the 1947-48 offseason that would make it easier for 3B to hit, but the shift was *dramatic*. In 1951, the season *before* Matthews' rookie year, only three starting MLB third basemen were below overall league average, and only one of them (Fred Marsh) was meaningfully so. The shift definitely took place before Eddie entered the league.
   119. DL from MN Posted: October 05, 2007 at 05:45 PM (#2561778)
It's a shame that the pitcher data doesn't look like it will be done in time to impact anyone's chances. Is there a similar discrepancy in pitcher replacement levels v. WARP & Win Shares that you can discuss?
   120. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 06:39 PM (#2561854)
Yeah man, I'm sorry, it's just really time-consuming and intellectually challenging. It took me about a year to fully develop my position player WARP, and I'm still revising it. And pitchers are more complicated because they a) have to be separated from their fielders b) they have had varying workloads over time and c) their workloads are not directly comparable due to relief pitcher leveraging and "chaining." The best study on pitcher replacement level I've seen was a recent one by Tangotiger, which is referenced at http://www.insidethebook.com/ee/index.php/site/comments/the_replacement_pitchers and suggests that modern starting pitcher replacement level is a .410 winning percentage. You can do a pretty straightforward WARP calculation from that: take a pitcher's RA, adjust it for park and defense, see how many games an average team would win if it added him, and add on 2.0 wins per 200 innings pitched for replacement level. To do a quick example, assuming Roger Clemens had average defense in 1997, he allowed 65 runs in 264 innings with a park factor of 101 (so 64.36 park-adjusted). The league RA was 4.98, so a league-average team plus Clemens would score 806.76 runs and allow (1458-264)*4.98/9 + 65 = 725.68 runs, which translates to 89.1 wins, making Clemens 8.1 wins above average. Add on 2.0*264/200 = 2.6 wins for the gap between average and replacement, and you get Clemens at 10.7 WARP, which sounds just about right for one of the greatest pitching seasons in history. BP has him at 14.9, so in that one example it's off by 4.2 wins--about the same size (and direction) error as it has for position players, once you factor in that Clemens pitched so many innings. WS has him with 31.7 pitching win shares, which suggests that it has pitching replacement level *exactly* right (unlike its hitting replacement level which obviously needs to be corrected).

None of this counts for the (high) standard deviation of Clemens's era, the (low) IP per starter of his era, or the (unknown) changes in pitcher replacement level going backwards in time. As soon as I have conclusive data on any of those issues, I will of course share them with the group.
   121. Mark Shirk (jsch) Posted: October 05, 2007 at 06:52 PM (#2561865)
Hey 'zop,

As someone looking to start a PhD this fall (I would be doing it in International Relations) can I ask you why you quit, especially at that stage?
   122. DL from MN Posted: October 05, 2007 at 07:10 PM (#2561894)
I don't think you necessarily need to get everything right to publish some results. An updated WARP with just the correction to replacement level would inform my voting significantly. There are a ton of pitchers in my spreadsheet with similar PRAA that are only differentiated by their PRAR.
   123. KJOK Posted: October 05, 2007 at 07:11 PM (#2561895)
Some have also studied & concluded that over a very long period of time, replacement level rises a bit.


I hope this is understood to be a fact, and not just 'a bit' but really 'tends towards .500' (or theoretically .517 or so) as X goes towards infinity with X being the TIME variable.

In other words, if a player gets hurt mid-game, the replacement level is very low - you grab someone off the bench who's on the current roster. If a player announces his retirement in April 2007 effective at the end of the year, his replacement level in 2008 will be MUCH higher - you trade for a replacement, or promote a hot prospect, etc.
   124. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 07:16 PM (#2561902)
OK, but trading for a replacement has a cost--the player you give up--and you can only promote a hot prospect if you have one; they're not freely available. The argument that "offseason" replacement level is higher than 80% of positional average makes no sense to me. Sure, you can acquire an above-replacement player, but only for an above-replacement price!

DL from MN, in that case, I would just use 1 win for every 100 innings as the gap between average and replacement for starting pitchers. (This is NOT the case for relievers).
   125. sunnyday2 Posted: October 05, 2007 at 07:29 PM (#2561922)
I think I asked the question "where are the 2B who hit (in the 1910s NL) like Larry Doyle?" and the answer remains, there aren't any...except Larry Doyle. Whether as a group they hit pretty well (relative to 1910s 1B) doesn't mean that the best of them is unqualified for the HoM.

Ironically, the argument was made that Jake Beckley was unqualified by virtue of his membership in an undistinguished group. Now, Larry Doyle is unqualified by virtue of his membership in a distinguished group. One that, BTW, is pretty much identical to the one Beckley belongs to.

My head hurts.

But seriously, Doyle was far and away the best of that group. Where or where is the 1910s NL 2B that hit like Larry Doyle?

Actually I have been working on documenting this whole question and there's more than I can get organized and posted right now but I'll just generalize. There are 10 peer 2Bs with enough innings to have a WS rating and only one of them, Johnny Evers, rates more than one full letter grade better. Doyle was a C+, Evers an A-. Everybody else--Cutshaw, Egan, Gilbert, Huggins, Knabe, Miller, Rawlings, Hornsby and Ritchey--are no better than B+ and most a B and B-. Other 2Bs played more and have letter grades elsewhere, and therefore played 2B for about 1 year each on average--Groh is an A- at 3B, Ivy Olson a C- SS, Doolan an A+ SS, Heinie Zimmerman a C+ 3B, Byrne at C+ 3B, Stock a C at 3B. My point being--was Doyle that bad and was the FAT (freely available talent) that good with the glove?

And is the differential on defense equal to the following: Doyle in the average year was 27 OPS+ points better not than FAT but 27 OPS+ points better than the average 2B. E.g. in 1910 he was 128 (only his 7th best year) and the other regulars (i.e. played the most games at 2B for their team whether they were batting title eligible or not) were Evers 115, Dots Miller 69, Knabe 89, Egan 82, Hummel 115, Shean 71 and Huggins 114. The average (technically the mean) is 114 + 89/2 (obviously). I use 1910 because it is the most typical year (differential of +26 versus mean of +27). Again, it is nowhere near Doyle's best year.

Maybe the reason for the following is that the FAT was so good that teams just couldn't resist going out and getting some of that FAT, year after year.

Pittsburgh 2B (1907-1920) Abbatichio 114-108 Miller 115-69-99 A. McCarthy 84 Viox 142-106-111 Farmer 103 Pitler 75 Cutshaw 116-79-71

Cincinnati (1913-1920) Egan 95 and Groh 107, Groh 120, Groh 123 and Olson 61 and Rodgers 91, Loudan 85, Shean 61, L. Magee 123 Rath 96-82. Prior to 1913 Egan held the fort for 5 years.

Philadelphia (1907-1920) Knabe 114-84-82-89-80-85-85 Byrne 85 Niehoff 77-95-93 McGaffigan 56 Pearce 26 and Paulette 78 Miller 68 and Rawlings 67

These are not atypical.

The idea that E. Collins had some impact on FAT in the NL--or on Larry Doyle's value in an NL pennant race--also does not stick for me. Sorry, more later.

PS. I am also ABD in sociology with a concentration is social research methods.
   126. Mike Green Posted: October 05, 2007 at 07:44 PM (#2561939)
DanR, Tango's article says that the replacement level for starters is .380, not .410. .410 is a what an average non-closer reliever will do in a starting role. An average non-closer reliever is better than replacement level.
   127. 'zop sympathizes with the wrong ####### people Posted: October 05, 2007 at 08:24 PM (#2561992)
121. Mark Shirk (jsch) Posted: October 05, 2007 at 02:52 PM (#2561865)
Hey 'zop,

As someone looking to start a PhD this fall (I would be doing it in International Relations) can I ask you why you quit, especially at that stage?


I was really pissed at my advisor last summer, since I felt I was being shat on in our lab in favor of some other older students whose data he was using for an upcoming NSF proposal (this is about as common in science labs as groundballs in a Fausto Carmona start). So I took the LSAT in a fit of rage and did very well.

At that point, it became clear to me that I was going to be leaving, literally, hundreds of thousands of dollars a year on the table if I pursued an academic career instead of going into law.

I really love geology, and I think that climate change is a hugely important issue. But, I don't love it -that- much, and frankly, anything you do as a job sucks in a job-ish way; any job has office politics, people you're forced to work with who you despise, and a large percentage of your time spent on mind-numbing menial drudgery. For the difference between a junior researcher's salary and an associate's salary, I'm willing to deal with a little more suckitude.

Sorry if that sounds cynical. But it's the truth.
   128. jimd Posted: October 05, 2007 at 08:24 PM (#2561993)
I hope this is understood to be a fact, and not just 'a bit' but really 'tends towards .500' (or theoretically .517 or so) as X goes towards infinity with X being the TIME variable.

What is the source for this statement?
   129. Mike Green Posted: October 05, 2007 at 09:08 PM (#2562046)
DanR, it also looks like you didn't make the park adjustment to the runs scored. The average team in the Skydome would have scored 814.8 runs. When you make this adjustment, and the .380 instead of .410 adjustment, Clemens' season comes out at about a 12 WARP, assuming average defence.
   130. sunnyday2 Posted: October 05, 2007 at 09:10 PM (#2562050)
>If a player announces his retirement in April 2007 effective at the end of the year, his replacement level in 2008 will be MUCH higher - you trade for a replacement, or promote a hot prospect, etc.

If OTOH you trade away Luis Castilla, then you replace him with Nick Punto or Alexi Casilla.

IOW, the vagaries of human/bureaucratic decision-making sometimes mean that the (more or less) implied "perfect market" for FAT doesn't function. And the value of a player is relative to the opposition on the field, not the opposition that is FA but in AAA or on somebody else's bench, etc. etc.
   131. KJOK Posted: October 05, 2007 at 09:35 PM (#2562097)
I hope this is understood to be a fact, and not just 'a bit' but really 'tends towards .500' (or theoretically .517 or so) as X goes towards infinity with X being the TIME variable.

What is the source for this statement?


Probably 100's of posts on Fanhome Sabermetics board, "THE BOOK" blog, etc.

The best 'one-stop' place would probably be here:

Baselines - Patriot
   132. jimd Posted: October 05, 2007 at 09:38 PM (#2562106)
Somebody did a study of all player replacements over a 50 year span or something like that?
   133. KJOK Posted: October 05, 2007 at 09:40 PM (#2562117)
OK, but trading for a replacement has a cost--the player you give up--and you can only promote a hot prospect if you have one; they're not freely available.


And as I've argued before, ANY replacement has a cost. You sign someone off the waiver wire, you pay him a salary, he takes up a roster space that could be used by someone else, etc, etc.

The issue is NOT cost, it's talent. And from a talent perspective, no team SHOULD be looking to fill a starting position with a 'slightly above replacement' player for the long term.
   134. TomH Posted: October 05, 2007 at 10:11 PM (#2562191)
when I respond eventually on the WS issue, it will be over on the uber system thread - WS vs WARP
   135. KJOK Posted: October 05, 2007 at 10:14 PM (#2562199)
Somebody did a study of all player replacements over a 50 year span or something like that?


At least read the part beginning at "PROGRESSIVE MINIMUM" if you're not going to read the whole thing to find the answer...
   136. jimd Posted: October 05, 2007 at 11:14 PM (#2562355)
I skimmed the document and read some sections in detail. I don't see what "PROGRESSIVE MINIMUM" has to do with replacement level.

If one defines all players to be replacement players, then of course the "replacement level" will be .500.
   137. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 06, 2007 at 01:18 PM (#2563957)
sunnyday, thanks for your response.

First, I didn't support Beckley. My stance more generally is that the overall level of positional offense has nothing to do with a player's value. Since every team has to field a player at every position, the way you win baseball games is by being better than your opponents at a given position, whether that means having a 90 OPS+ where they have a 50 or a 130 OPS+ where they have a 90. What *does* matter is *how* much better you are than your peers. Pie Traynor was the best white 3B of the 1920's, but the gap between him and the worst regular 3B in MLB in those days was not particularly large. David Concepción was the best SS of the 1970's, and the gap between him and the worst regular SS in MLB in his days was ENORMOUS. That's my definition of Merit. I find Doyle far short of Meritorious because he didn't exceed the baseline (the worst-regulars average) of his position in his time by a sufficient amount. This would be true regardless of whether 2B offense was at an all-time low or all-time high in his era. I mention the strength of 2B offense in Doyle's era because his hitting probably *would* be Meritorious (ignoring fielding for now) if he had been able to produce those OPS+'s in, say, the 1960s.

You note that Doyle exceeded his positional OPS+ by 27 points. He probably gave back about 10 of those points with the glove. Is there a single HoMer with a middling career length (13-14 seasons) who only exceeded his positional OPS+ by 18 points after adjusting for defense? That sounds like an argument *against* him, not for him.

The examples of teams with a string of poor-hitting 2B that you have selected continue well past 1915, which is when both the 2B worst-regulars average and Doyle's own offense began to drop markedly. The only consistently poor hitter at 2B that you mention during the 1907-15 period is Knabe, who was a wizard with the glove.

Eddie Collins does not have any impact on FAT as I calculate it, because he was never one of the worst three 2B in his league.

Mike Green--On the .380 vs. .410 thing, good catch. I suppose that to *really* be accurate, you would have to conduct a "chaining" analysis--the starting pitcher is replaced by the long reliever (costing about 10 runs per 100 innings at a leverage of 1), the mopup man replaces the long reliever (costing another 6-7 runs per 100 innings, but at a very low leverage), and an AAA player is called up to replace the mopup man (costing an untold number of runs at an infinitesmally low leverage). Then you'd add up those costs after accounting for leverage to get replacement level. And then you'd have to update that for changing pitcher usage over more than a century...and you wonder why pitcher WARP is taking me so damn long??? :)

As for the park adjustment, you're doing it wrong. You EITHER adjust Clemens's runs allowed to a neutral park, dividing by 1.01 (which is what I did), OR you increase the Average Team Plus Clemens's runs scored to the Skydome level, multiplying by 1.01, but NOT both. The result is the same no matter which way you do it.
   138. sunnyday2 Posted: October 06, 2007 at 02:20 PM (#2564004)
>Is there a single HoMer with a middling career length (13-14 seasons) who only exceeded his positional OPS+ by 18 points after adjusting for defense? That sounds like an argument *against* him, not for him.

That of course is the question. But I would add that for a peak voter, the key is not the mean of 27, it's the advantages vs. the mean 2B of +54, +47, +45, +40, +37, +32, +28, +26, +23 that counts. IOW for 5 years the edge was not 27, it was ? +37 OPS+ points and up to +54 points, +/- (and in Doyle's case perhaps minus) some defense. But as I also showed, the FAT that was populating the other rosters at that time tended to be B and B- gloves, not a bunch of A+ guys, which is of course consistent with what we know about where the position sat on the spectrum in those days before the DP became de rigeur.

I did cover Doyle's entire career as a regular, if it is inconvenient to consider his period 1915-20 what does that say about the method. His offense began to drop and yet he provided positional advantages of +47 points in 1915, +45 points in 1919, +37 points in 1916, +23 points in 1918--again, this is against the mean 2B as a hitter, not against the worst/FAT. Knabe must have been a wizard, holding a job for 7 years with OPS+ consistently in the 80s, but your point was that Doyle's cohort was a bunch of terrific hitters. My point was, NOT, and the defense wasn't out of this world either. Knabe's WS rating is B+ which is good and better than Doyle's C+ but hardly comparable to an A+ SS. Wizard in relative terms, somewhat deflated terms, in fact. Only 3 of 10 of Doyle's cohort who have a WS ranking are B+ or higher. The other 6 are B and B-. How many runs/how many wins is half a letter grade worth against a C+ 2B when 2B was not a terribly important position? Not many.

I'd be willing to bet that we are close to electing any number of hitters whose positional advantage is less than Doyle's consistent 30-50+ for his peak/prime and 27 for his career. Bob Johnson comes immediately to mind. Does Bob Johnson provide a positional advantage vs. the mean LF (not the FAT but the mean ML talent) of 54, 47, 45, 40, 37, 32, 28, 26, 23 with a mean of 27? I would be totally shocked if that were the case. No wonder we're electing too many bats. We don't even seem to demand a positional advantage, just a high OPS+.

If I have time I will compare Doyle to Concepcion, but time is something I don't have much of right now. I agree with you that we need to be considering positional advantage, not just the raw offensive production, and that a good middle IF or a good C can provide a ton of positional advantage. I just find that Larry Doyle fits that description, along with Rizzuto, Pesky and others. I haven't yet determined to my own satisfaction re. Davey. Why not Fregosi? He looks better to me as of today.
   139. Howie Menckel Posted: October 06, 2007 at 03:08 PM (#2564034)
"David Concepción was the best SS of the 1970's, and the gap between him and the worst regular SS in MLB in his days was ENORMOUS. That's my definition of Merit."

That's my definition of a reason why the REDS won.
Now, do we give 100 percent of the credit to Concepcion himself?

That depends.
If the position of SS in the 1970s was so demanding that it was almost impossible to find a respectable fielder who also could put up a respectable OPS+, then I would give Concepcion full credit. There have been eras/positions where certain spots indeed seem to have been particularly difficult on its performers.
But if the reason Concepcion had such an advantage had a lot to do with general GM/managerial ineptitude - such as the current system that has teams bring in lesser relievers in late-inning tie games while saving their best reliever to save many 2- and 3-run leads - then I will not force myself to automatically elect such a player.

I actually vote for Concepcion on the latter half of my ballot, so I am hardly his enemy.
But where you lose me, and perhaps others, is what I see as locking yourself in.

I imagine you don't agree with my ineptness question.
But what if you did believe it to be true, even to an extreme? Is there a point at which you might say, "Wait, I know this is a mediocre player [not talking Concepcion here] overall, and I know the other teams have mediocre-player alternatives that they are not using, so maybe I hesitate in putting this guy into the Hall of Merit?"

I realize this is an extreme example.
But especially in the 8-team days, just two dopey GMs can really make a difference in the comparative position player pool.
Conversely, I'm not going to be stuck dismissing a candidate just because he played at a time of spectacular quality at his position. Yes, his relative worth is a factor, but I'm trying to be moderate about it.

Maybe I'm missing the point, though. Help me out.
   140. sunnyday2 Posted: October 06, 2007 at 08:13 PM (#2564211)
Howie, you are right. I mean, certainly the idea of "replacement value" can use refinement and FAT is an interesting refinement. But in order for it to define value to ten decimal places, as seems to be the claim, one would have to assume a perfectly functioning market. This seems to be a big assumption or, rather, an assumption for which there seem to be counter-claims.
   141. TomH Posted: October 07, 2007 at 02:28 AM (#2564778)
well, I can't FIND the uber WS vs WARP thread, so let's continue here...

DanR: The only way you can calculate that a player who got 3 WS on the 1899 Cleveland Spiders contributed exactly one win to the team is if you use a replacement level of precisely 52% of league average offense. Let's take a hypothetical guy who was 3 WS = 1 win in 1899. League average was .214 runs per batting out, so James's 52% baseline is .111 runs per out. Say the guy made 300 outs. If he was 1 win = 10.77 runs in 1899 above that baseline, then he created .111*300 + 10.77 = 44.2 runs. OK, so if you use James's 52% figure that's 3 batting WS. But if you used a baseline of 25% of league average offense, then he'd magically get 44.2 - (300*.214*.25 = 16.05) = 28.15 runs/10.77 runs/out * 3 WS/win = 7.8 batting WS.

No, that is not right. You can't merely move the baseline and then neglect to reaccount for the ratio of runs to wins, or there would be far more win shares than wins. WS starts with team wins, figures how many runs it takes to make thsoe wins in the team as a whole, and then distributes them (offensively) but RC/out. The above calculations are not correct; changing the baseline would not give the example batter 4.8 more BWS. Same with the Rosen example.

DanR: See, there's nothing inherently "true" or "right" about the Win Shares allocation system.

Oh, I completely agree. But it makes much sense for the application for which it was designed; because using absolute zero has its own problems, and using 80% of average (or precisely average, the only other measure that has inherent good properties to it) causes wins then that need to be re-allocated by playing time, since there would be many fewer WS earned than wins the team achieved. Using 'average', you could create a system that then adds so many "wins" for each player by playing time (1 win per 150PA or 40 IP or some such thing). But James created a system that did not require that. 52% on offense was a low enough replacement level to make it work. I am not really trying to defend 52% as "right"; is it arbitrary? Sure. Would other numbers work? Sure; IF you wished to go back and subtract or add in fudge factors to make the individual totals match ther team total. WS is a top-down approach, and maybe I should leave it at that; it is DIFFERENT from a bottom-up approach that almost every other system uses. In some applications, it will be a better tool. For others, it is worse. If I had invented it, I might have tried to come up with a player's Win Shares AND Loss Shares.

DanR: And regarding the 1975-76 Reds, who would they have been forced to play if one of their stars had gone down? A replacement player, of course--one whose production would probably be approximately 80% of positional average.

My point about the extreme teams was that great teams tend to have higher freely available talent on the bench (Dan Driessen!), whereas the Spiders do not (duh). Is this important when considering long-term replacement level for the HoM? Maybe not. Again, the question WS was designed to answer was "how do I distribute the 108 wins in Cincinnati's 1975 team among their players?". Babe Ruth, on the 99 Spiders, could have been just as great a player but would have likely won fewer additional games for that team. This is pretty obvious, no? WS captures this. Whether you think this is improtant or relveant can be argued. But it does capture it.
   142. Howie Menckel Posted: October 07, 2007 at 03:19 AM (#2564861)
The extremes can be fun.

Let's give Babe a 195 OPS+, 30 pts higher than the 2nd-highest league total.

The pitchers are even worse than the Spiders, so they go 10-144, basically losing games 6-2 because of Babe where they'd lose 6-0 without him.

Are Babe's contributions nearly irrelevant to this team? Yes, they lose almost every game anyway.

Is Babe the best player in the league?
I say yes.

Anyone say no?
   143. Dr. Chaleeko Posted: October 07, 2007 at 07:16 PM (#2565568)
I'm a dedicated WS voter, as you all know. I've not subscribed to BP WARP for many reasons, ones we've all talked about.

But I'm now in the process of putting Dan R's data into a format I can use and work into my analysis (such as my analysis is...). I guess I'm going behind WS back....

Much to many people's horror, however, I'm going to mix the two of them. And the reason why is precisely the reason that TomH and Howie have enumerated in #s 141-142. Because Dan's system tells me who the best players were from the bottom up, and because Bill's system tells me how much value being the best player had to that player's particular team in terms of their wins (fromthe top down).

I know some will be scandalized, but I like that both systems are fully elaborated and that I can see the steps in both systems and cogitate on them, unlike BP's black box numbers. And I like bringing in more opinion that I trust, not less. I don't agree with everything I know about Bill's system, nor with everything I know about Dan's system, but I think between the two I will arrive at very good answers.

Too bad I won't be able to actually combine them enough to use them before the 2006 and 2007 elections.... But hopefully by 2008 I'll have it all cobbled together. By which time, perhaps I'll finally go back and construct my many-times-aborted pHOM.
   144. Joey Numbaz (Scruff) Posted: October 07, 2007 at 07:24 PM (#2565576)
By which time, perhaps I'll finally go back and construct my many-times-aborted pHOM.


I also need to do this.
   145. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 07, 2007 at 11:08 PM (#2565995)
Wow, in the day since I've been away from the Internet this discussion just got seriously substantive. Great thread, guys. On the Merits:

Sunnyday:

1. You have to take into account that the standard deviation of fielding ability in Win Shares is far below what all the play-by-play systems show it is, and far below the level that is consistent with the standard deviation of team defensive efficiency. (The degree of this underestimation varies by position, and is particularly extreme for corner outfielders and first basemen, but for second basemen the stdev is 38% lower in WS than it is in Chris Dial's system). "How many runs/how many wins is half a letter grade worth" is the *exact* right question to ask. My research finds that for Doyle, it's about 0.9 wins a year below average.

2. OK, so the criteria you're looking at is best 5 years of OPS+ above positional average? I don't happen to have positional OPS+ data--where did you get it, out of curiosity?--but I can convert pretty easily from my BWAA. Looking at Reggie Smith, for example, I get OPS+ above position of +56 in 1977, +50 in 1978, +44 in 1974, +41 in 1969, and +37 in 1973, suggesting that the two were similarly valuable hitters at their peaks after adjusting for position (assuming your numbers for Doyle are right). Then once you factor in that Lil' Reggie has none of the fielding problems that Doyle had, and he has much career more outside of his peak, and that he played in integrated leagues with a low standard deviation while Doyle was playing in the EXTREMELY weak teens NL (compared to the star-studded AL), and it seems that Lil' Reggie leaves Laughing Larry in the dust...no?

3. I don't support Bob Johnson, but he very well might have had similar OPS+ above positional average. Shouldn't be hard to look up, no? (Just make sure to knock something off for the war years).

4. Fregosi is a serious candidate, right on my in/out line. He was a no-doubt-about-it superstar from 1963-70, albeit in an AL that was substantially weaker than the NL of those days. Definitely a higher peak than Concepción's. But Concepción has *so* much career on him--Fregosi really has *nothing* outside of those eight years--that he's substantially ahead from me. A pure peak or short-prime voter unconcerned with league strength would probably prefer Fregosi.

Howie Menckel:

I'm not a complete believer in efficient markets, but I do think there are very plausible explanations for the drop in SS replacement level starting around 1970. First, you had the gigantic expansion of 1969. As Nate Silver noted in his original Freely Available Talent article, the reason why SS is the only position where FAT players are meaningfully below-average fielders is simply that it is the hardest position to play, and you have to be a truly gifted athlete to both field it competently and hit at a tolerable major league level. Since there may not be more than 15-20 human beings playing baseball on this earth capable of doing that at any given time, FAT shortstops are below-average fielders as well as hitters. Given that, it seems entirely reasonable that the mega-expansion wouldn't affect all positions equally, and that the gap between the #24 and the #20 shortstop would be bigger than the gap between the #24 and the #20 1B. Secondly, you had the move to turf fields, which has been discussed at great length. The combination of those factors is more than enough to convince me that you really did need a super glove at SS to compete in the '70s (and many of the winning teams like the Orioles, Reds, A's, and Yankees had them), and that Concepción and Campaneris "deserve" all the credit for the pennants they added. Your mileage is free to vary.

TomH:

Mea culpa on the calculations, but does that change the substance of my point? That 52% is an arbitrary number, and that using a different baseline level would lead to a different allocation of Win Shares among batters?

What are the problems of using absolute zero, besides the fact that it's just blatantly not representative of how baseball works? (Neither is 52%, of course).

Using wins above/below average plus credit by playing time would be an INFINITELY better system than the current Win Shares model, in my opinion. Incomparably better.

Moreover, 52% is not a low enough level to make it work, because there *are* still players who hit at worse than 52% of league average. Look at Bill Bergen, who had not a single Batting Win Share in his nearly 1,000 game-long career. The presence of Bergen causes the Batting Win Shares of all of his teammates to be inappropriately reduced.
   146. TomH Posted: October 08, 2007 at 12:12 AM (#2566273)
If you used absolute zero, every batter would have so many runs above baseline, you would have to go back and take away wins per playing time to get the team wins to match win shares. Or make believe it takes oddles of runs to create a 'win'. E.g., if a team only wins 40 games and so gets 120 WS, and if 60 were batting, and they scored a mere 600 runs, that's 10 runs per win share, or 30 runs per win. Wouldn't work.

Your take on WHY the FAT talent level of shortstops dropped in 1970 is a fascinating theory. If SS FAT level drops often with expansions, that would be a large finding (not only for this dicsussion, but for MLB GMs!!). That is actually a question that really piques my interest at the moment.
   147. Dr. Chaleeko Posted: October 08, 2007 at 12:17 AM (#2566307)
Then once you factor in that Lil' Reggie has none of the fielding problems that Doyle had, and he has much career more outside of his peak, and that he played in integrated leagues with a low standard deviation while Doyle was playing in the EXTREMELY weak teens NL (compared to the star-studded AL), and it seems that Lil' Reggie leaves Laughing Larry in the dust...no?

Dan, I think you could be double-dinging Doyle and some other players on the unintegrated question, within the HOM context (not as a general rule). I'm not sure how much sense this will make, but I'll try. In the NgL era (about 1900-1950), there's a big bulge in the number of electees each year and a consequent bulge in the number of total electees concentrated in the the NgL era when the NgLs were at their peak and our records are best in tact for them.

So what's happened, in essence, is that we've elected two parrallel groups: the best MLB players of the NgL era, and the best NgL guys of the NgL era. We essentially avoid the question of what an integrated league would even look like by electing represntative samples from both simultaneously. We elect from three (or four) leagues instead of two, meaning that the comparisons of MLB players to league still works by dint of MLB election slots being apportioned in similar proportions to other 8-team leagues. So applying the unintegrated argument to, say, Bob Johnson may be hitting him with an extra demerit relative to the number of election slots apportioned to his era.

Now your rejoinder seems likely to be about the quality of the league rather than about HOM representation (that integrated leagues were just better, and indeed were), but for the purposes of HOM representation and voting, we could instead posit that Reggie Smith was playing in the near numerically equivalent environment of three to four leagues (24 to 26 teams versus 16) just as Johnson was playing within a three-to-four major league system throughout his career. Our perception is different of those leagues, but the demographic reality is pretty close for each.

Now returning to Doyle. The question is, indeed, murkier. We have less reliable stats before 1920, and we're not entirely sure how many Black teams there were, nor how good they were. But we've nonetheless HOM'ed a bunch of Black players from the Doyle era: Williams, Johnson, Lloyd, Hill, Santop, and pieces of Torriente and Charleston's career (off the top of my head, sorry if I forgot someone), an era in which we were electing 1 player a year for much of the time. So while maybe not as compelling as the Johnson example, the point remains in play: we have built in representation, which has the effect of avoiding the question of whether an integrated league is better or worse than a latter-day integrated league.

Your mileage is likely to vary.
   148. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 08, 2007 at 02:30 AM (#2566804)
Sunnyday:
I just double-checked Smith's vs. Doyle's position-adjusted offense. First, I screwed up Smith's 1973 because I didn't account for the DH, but he also didn't play a full season that year, so I guess those two factors probably cancel out. Second, I did Doyle by the same methodology (BWAA per year minus positional average BWAA per year times 11.43), and get notably lower numbers for his top five than yours: +46, +45, +43, +37, and +36, which is way worse than Smith. How are you getting your results? Assuming that the +54 you cite is 1911, that would suggest that 2B were exactly league-average hitters in the 1911 NL. Given that their EqA was .276, that's just about impossible.

TomH:
Is that why James picked 52%? Because it's the only number that gives you 10 runs a win in the modern game with no tweaking? That would be sort of cute. Still empirically wrong, but a tiny bit less arbitrary--a number selected for convenience rather than for accuracy.

Measured in absolute terms, the FAT level at all positions always drops with expansion, of course. The question here is whether the size of that drop varies by position. It seems to me perfectly reasonable to expect that it would. If you look at my chart, you see SS rep level fall off the table in the 1960s after staying fairly constant from about 1915-60, and it does show another mini-drop right after 1977. But it climbs sharply throughout the 1990s when there were two expansions. Then again, there wre lots of countervailing factors: first, the 1990s were the post-Ripken era, and second, there were many more True Outcomes and fewer balls hit to SS, decreasing the importance of the position. I just think that with so few expansions, there isn't enough data to get any sort of statistical confidence about this question.

Eric--Actually, my integration argument most definitely is about HoM representation. I think we've screwed up by electing a bulge of players from 1900-1950. I think that's being unfair to the pre-1900 and post-1950 eras. I think we should have elected fewer players from both the Negro Leagues and the major leagues in that time period, and more players from both before and after.
   149. Howie Menckel Posted: October 08, 2007 at 03:13 AM (#2566863)
Dan R,
I think you just made the case you need to make in replying to me.
Others may agree or disagree, but you want to get that angle out there....
   150. sunnyday2 Posted: October 08, 2007 at 01:41 PM (#2567044)
For Larry Doyle, yes, his biggest year (+54 OPS+ versus the mean for NL 2B) is 1911. All I did was take all of the NL 2B in order:

Doyle 153 OPS+
Sweeney 120
Hummel 113
Huggins 99
Miller 99
Evers 86
Knabe 80
Egan 75

Or for his most average year(s), 1910 and 1912:

Doyle 128 and 132 (+26 and +28 versus the mean regular NL 2B)
Evers 115 and 139
Huggins 114 and 117
Knabe 89 and 85
Egan 82 and 72

Pitt--Miller 69 and McCarthy 84
Bkn--Hummel 115 and Cutshaw 91
Bos--Shean 71 and Sweeney 133

As noted elsewhere, for his career Doyle averaged +27 OPS+ points versus the mean regular NL 2B each year. Taking a typical year, in this case 1910, you get the following.

Doyle 128 OPS+ 21 offensive WS and 4 defensive in 151 games
Mean 102 OPS+ 12 offensive WS and 4 defensive in 139 games

That is the sort of advantage Doyle gave his team in a typical year.

I wasn't suggesting, BTW, that 5 year consecutive peak is the be-all, it was just one set of numbers to look at. But it is worth noting some of the assumptions that go into this analysis, aside from the fact that it uses WS. It also sets Doyle (or any player) up within his league on the basis that "a pennant is a pennant"--i.e. it is within the NL that Doyle's value has meaning. His value has no meaning versus the other league.

For the record, the AL 2B that year include Eddie Collins and Nap Lajoie, and some competent hitters like Frank LaPorte, Jim Delahanty, Larry Gardner (in 113 games in his first year as a regular and his only year at 2B), but also non-entities like Rollie Zeider, Red Killefer and Frank Truesdale. Net: better than the NL and undoubtedely pushing the mean up a few OPS+ points and maybe 2 WS.

So what's left is 9 offensive WS within the league and maybe 7ish versus the MLs, and a break-even with the glove. Note that as a C+ glove Doyle is worse than every peer for whom we have a letter grade. What evens this out is the simple fact that there are a few others for whom we don't have a letter grade, and the fact that Doyle had more playing time that all of the better opponents and 151-139 games on average. How many WS is a back-up 2B going to get replacing Johnny Evers for 30 games or Jim Delahanty for 50 games? Not too many. So I would say that Doyle's defense actually cost his team relatively little in practice. Not to blow it off completely, but Dan's own calculation says 0.9 versus FAT. It might be 1 to 1.5 WS versus the average team (not Evers or Huggins alone, but Evers or Huggins and their replacements). That seems to me to be pretty much the limit.
   151. sunnyday2 Posted: October 08, 2007 at 01:45 PM (#2567050)
All of this BTW is meant to test DanR's methods in ways that make sense to me. That is, asking how much of an advantage does a player provide versus others at his position (which is what DanR is trying to calculate, I think) but, in my case, within the league, not versus all ML talent. In WS. And comparing the test players versus the mean rather than against FAT.

The obvious problem is I can't calculate these comparisons by hand for more than just a few players, but I am working on a few other players.

Just in the case of Larry Doyle, it seems that he was worth about 27 OPS+ points and 9 offensive WS in a typical year and gave away no more than maybe 1-1.5 WS on average with the glove.

The key question is whether this is a great performance, or whether it is just very good, or maybe only a good performance and all I can do is compare Doyle to a few other players this week. That is what I will be doing.
   152. sunnyday2 Posted: October 08, 2007 at 01:47 PM (#2567052)
PS. 5 years peak is not the be-all or end-all here at all. It was just one of the data sets I happened to look at. What I'm really trying to do is look at careers and then at "typical" seasons for a more finite understanding of what is really going on.
   153. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 08, 2007 at 02:09 PM (#2567081)
Sunnyday, thanks very much for the research and clarification. I think I found your mistake--you're listing Johnny Evers, who had an 86 OPS+ in 155 at-bats, as Chicago's starting 2B, rather than Heinie Zimmerman, who had a 124 OPS+ in 535 at-bats. Replace Evers with Zimmerman and you get a simple average of a 108 OPS+ for NL 2B in 1911, making Doyle 45 points above positional average, not 54. Would you mind re-running your analysis with that correction (and double-checking your other numbers) to see how it affects Doyle's standing?

As for his fielding, the 0.9 wins is vs. FAT *and* vs. average, since FAT second basemen field the position at a league average level (according to Nate Silver). Doyle should be docked 2.7 WS per 162 games for his defense, not 1-1.5 WS.
   154. sunnyday2 Posted: October 08, 2007 at 04:26 PM (#2567300)
The 1-1.5 WS is the actual margin by which the "mean" 2B outperformed Doyle, looking at the actual WS earned by each regular 2B at his position in 1910. He earned 4 defensive WS in 1910 (rounded), Evers and Knabe 6, Huggins 5, Dots Miller 2, and so on. So he was not 2.7 WS worse than average (or than the mean) on a seasonal basis. Of course, playing time enters in. I suppose that with rounding he could be 2.7 WS worse than the best defender for the season but that would be pretty much the worst case.

I'm not arguing that 2.7 is wrong, just that it overstartes his actual performance per season in which neither Doyle nor anybody else actually played 162 games, and many played less than 130-135-140. And recognizing that this was more of an offensive position and as a result, regulars like Egan and Huggins and Dots Miller were B- gloves themselves. Evers was the only A (A- actually) glove in Doyle's NL, and Evers was done after 1916, so while I suppose he might have given back more than 1 or 1.5 WS in some other years, I can't believe it would be a lot different even if his defense was in decline before his offense was (and his OPS+ was 116-137-107 in his final 3 years). One thing about Doyle is that there was little if any ramping-up or decline (OPS+ 83, then 134 his first 2 years). So his career was "shortish" at 14 years, but relatively few players had 13 years of prime like he did. (OK, OPS+ 99 in 1917).

But bottom line, even after giving back 1-1.5 WS per actual season, he is still +9 WS including offense (+10) and defense (-1). Earlier it was stated that if he was +28 OPS+, he probably gave back 10 OPS points with the glove. I had a hard time visualizing what this means, so I wanted it in WS. So a 28/18 ratio seems to be somewhat overstated.

I realize these are WS and not every likes WS. And again, this particular year was selected for a closer look because his OPS+ advantage of +26 in 1910 is about the mean for his career. He advantaged the Giants (or Cubs) by ? 27 OPS+ points in half of his seasons and ? 27 points in the other half, though in his case he was below the mean only one time, in his rookie year. Correcting 1911 to +45 wouldn't change any of that.

And how this compares to other backloggers, I don't know yet. I am out on a limb here, you know. I support Larry Doyle, and I am going to explore whether +27 OPS+ and +9 WS per year turns out to be a good record. I think it will.
   155. Paul Wendt Posted: October 08, 2007 at 07:31 PM (#2567569)
Unanimous HOMers:

Schmidt and Boggs stand alone since the big six outfielders: DiMaggio, Williams, Musial, Mays, Mantle, Aaron. (The big six outfielders, debuts 1936-1951, are the players my dad's generation recognized as the best players in baseball.)

The six elder unanimees are Young, Wagner, Johnson, Ruth, Gehrig, Grove, debuts 1890-1925.

The distribution by position is
7 of (three in center)
3 p
2 3b
1 ss, 1b
0 2b, c
   156. sunnyday2 Posted: October 08, 2007 at 08:04 PM (#2567629)
Well, Paul, I guess I am old enough to be your father.
   157. Paul Wendt Posted: October 08, 2007 at 09:44 PM (#2567790)
before the election I predicted that one or two voters would leave him out of the top spot; however, as the week went along, it seemed more and more likely that he'd pick up the 100%

don't take that too seriously, Chris F!


OCF:
Waddell did appear on my ballots for 27 years, from 1916 through 1942, peaking at #4 in 1929; however, he had slipped to about #31 or so by the year of his election.

by the way, Waddell also slipped to the thirties in the annual results; the difference is he didn't come back on OCF's ballot. Waddell may be the biggest "double reconsideration" here, ahead of Joe Sewell.

Of course some single reconsiderations have been much greater in magnitude, with most of the really big ones concerning players widely credited with play outside the majors. Even outside the major Negro Leagues, because Dobie Moore and Alejandro Oms must be the numbers and two --ordering that I may reverse if Oms goes in. Oms scored no points in his first year, Moore scored none when the backlog was strongest (the mid-1950s after 15 years of headliners).


DanR
sunnyday2:
What on earth do you mean that you "don't agree with giving a player credit for what a replacement player might have done while he was out of the lineup?"


Let me summarize or caricature a position that I have gleaned in 100 years of reading here, maybe not from Marc.
A player who needs replacement mid-season should be debited some, not credited with replacement level play (0) for the games he misses. Suppose you denominate in wins and you generally count wins against a replacement benchmark. A player with +5 wins in the first half of the season, then DNP, should be credited with something like +4 wins (less than +5) for that season.
This may be a good fit with KJOK position on replacement and the time horizon, although he contrasts only two extremes: replacement mid-game and replacement with one year advance notice (next).


====== above this line, quotations are from the preceding page ======

123. KJOK Posted: October 05, 2007 at 03:11 PM (#2561895)

> Some have also studied & concluded that over a very long period of time, replacement level rises a bit.

I hope this is understood to be a fact, and not just 'a bit' but really 'tends towards .500' (or theoretically .517 or so) as X goes towards infinity with X being the TIME variable.

In other words, if a player gets hurt mid-game, the replacement level is very low - you grab someone off the bench who's on the current roster. If a player announces his retirement in April 2007 effective at the end of the year, his replacement level in 2008 will be MUCH higher - you trade for a replacement, or promote a hot prospect, etc.

124. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 03:16 PM (#2561902)
OK, but trading for a replacement has a cost--the player you give up--and you can only promote a hot prospect if you have one; they're not freely available. The argument that "offseason" replacement level is higher than 80% of positional average makes no sense to me. Sure, you can acquire an above-replacement player, but only for an above-replacement price!</i>

I believe there is a stochastic component of the argument which you have a missed.
KJKS announces his retirement in April effective immediately, so he is replaced by freely available talent, say 80% of league average for sake of argument. If KJMO announces that this is his last season, so that replacement is needed one year hence, you have a chance to work with and observe three replacement players for a year, this season and next winter. After that year, one appears to be a 70, one an 80, and one a 90. So our retiring player gets replaced by a 90, someone no longer freely available because he has been selected as the best of three players (in this simple example). An enthusiastic proponent of this argument might use 60, 80, and 100 for numerical example.


Marc sunnyday2 on Larry Doyle in the 1910s
FWIW I don't like the limited "within-league" comparison.


DanR #146
Looking at Reggie Smith, for example, I get OPS+ above position of +56 in 1977, +50 in 1978, +44 in 1974, +41 in 1969, and +37 in 1973, suggesting that the two were similarly valuable hitters at their peaks after adjusting for position (assuming your numbers for Doyle are right).

But Reggie Smith notably better than numerous HOM outfielders --measured by rate OPS+ 137, a context provided for me by DanR and sunnyday.


DanR #148
Eric--Actually, my integration argument most definitely is about HoM representation. I think we've screwed up by electing a bulge of players from 1900-1950. I think that's being unfair to the pre-1900 and post-1950 eras. I think we should have elected fewer players from both the Negro Leagues and the major leagues in that time period, and more players from both before and after.

In the first sentence, at least, 1910-1950 seems more accurate. What do we have, Frank Grant at peak in 1889, Grant Johnson in 1899, Rube Foster in 1904.
   158. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 08, 2007 at 09:52 PM (#2567797)
Paul Wendt--And *why* should he be docked a win for the rest of the season? Replacement level is replacement level--it's freely available! By definition!

Sunnyday--By your own calculations, doesn't Reggie Smith have a higher OPS+ relative to position, both on a peak and career basis, than Doyle?
   159. Paul Wendt Posted: October 08, 2007 at 09:58 PM (#2567805)
Well, Paul, I guess I am old enough to be your father.

I think not, not even a member of his generation. Maybe you qualify for five of six, all but DiMaggio. At that, you may have been slow to pick up on the declines of Musial and Williams, as a still youngish fan c. 1960(?). I was thinking something like age 10 and up. Anyone born in the thirties would plausibly see DiMaggio as the best for at least two seasons, 1949-1950.
(My dad, another Robert, knew of "Rapid Robert" Feller's sensational age 17-18-19-20 seasons, at least by 1937, Feller's age 18. I never thought to ask him about Ducky Medwick ;-!
But he was not yet ten and he is by definition in the middle of his generation.)

I think of we baby boomers as a generation. Maybe I wouldn't think that way if I were near one end or the other.
   160. sunnyday2 Posted: October 09, 2007 at 02:30 AM (#2568779)
>Sunnyday--By your own calculations, doesn't Reggie Smith have a higher OPS+ relative to position, both on a peak and career basis, than Doyle?

Don't know yet. Like I say this is a time-consuming project. But: I had Reggie2 on my ballot early but that was based too much on OPS+ with not enough recognition of playing time issues.

>At that, you may have been slow to pick up on the declines of Musial and Williams, as a still youngish fan c. 1960(?).

Not. You had the young turks, Mantle and Mays, and you had the grand old men, Musial and Williams. Even a 10 year old could tell the difference.

>I think of we baby boomers as a generation. Maybe I wouldn't think that way if I were near one end or the other.

I think that's right. My sister (b. 1946) hated the Beatles. And the late boomers weren't even born when JFK was killed, and were not yet in K when I was graduating from HS. I was 18 in 1968, the youngest ones in 1982. Vastly different worlds.
   161. Howie Menckel Posted: October 09, 2007 at 03:04 AM (#2568958)
yeah, I have siblings born in 1951, 1956, 1958, and also 1961

different animals
   162. John (You Can Call Me Grandma) Murphy Posted: October 09, 2007 at 11:34 AM (#2569572)
And the late boomers weren't even born when JFK was killed, and were not yet in K when I was graduating from HS.


That sounds like me (I was born in 1965), but I'm not always grouped with the boomers. Sometimes I see graphs that end in '65, but sometimes earlier. I have no idea where I belong (not that I really care).
   163. sunnyday2 Posted: October 10, 2007 at 12:31 AM (#2570331)
I've always seen it 1946-1964 but that's arbitrary. Some kids born after 1964 are cooler than some born before. Hard to believe....
   164. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 11, 2007 at 02:00 PM (#2571602)
Here's a comparison of the post-1893 MLB prime backlog bats, along with one easily-elected HoM'er for reference.

The usual glossary and charts:

The following numbers are all standard deviation-adjusted. SFrac is the percentage of the season played (compared to a player with league average PA/G playing every game). BWAA is batting wins above league average, BRWA is baserunning wins above a league average runner, FWAA is fielding wins above positional average, Replc is wins above average a replacement player at the same position would have accumulated in the same playing time, and WARP is the first three minus the fourth (wins above replacement). Note that Rep is 0.6 wins lower in the AL than in the NL to account for the DH after 1973. All seasons are adjusted to 162 games. I add 10 to the park factor and subtract 2 FRAA per 162 games for 1943 and '45, and add 20 to the park factor and subtract 4 FRAA per 162 games for 1944. I use the data in the relevant players' discussion threads for minor league and Japan credit. aTTL (where included) is career totals excluding sub-replacement seasons. Sorry for the goofy formatting, but the PRE tag doesn't seem to be able to handle consecutive whitespaces anymore.

Reggie Smith

Year SFrac BWAA BRWA FWAA Replc WARP
1967 00.94 
+0.3 +0.3 +0.7 -1.10 +2.3
1968 00.96 
+2.7 -0.2 -0.4 -1.10 +3.2
1969 00.89 
+3.6 -0.3 +0.0 -0.90 +4.2
1970 00.94 
+2.8 +0.1 +0.3 -1.00 +4.1
1971 01.04 
+3.1 +0.3 -0.3 -1.20 +4.2
1972 00.86 
+3.8 +0.3 -0.3 -0.60 +4.5
1973 00.72 
+3.0 +0.0 +0.5 -1.40 +4.9
1974 00.87 
+4.5 +0.1 +0.7 -0.60 +5.9
1975 00.81 
+2.8 -0.5 -0.2 -0.60 +2.6
1976 00.64 
+1.1 +0.0 +1.2 -0.40 +2.6
1977 00.88 
+6.0 -0.1 +0.3 -0.60 +6.8
1978 00.79 
+5.1 +0.2 -0.5 -0.50 +5.3
1979 00.40 
+1.2 -0.1 +0.4 -0.20 +1.7
1980 00.54 
+3.4 -0.2 +0.9 -0.30 +4.4
1982 00.59 
+2.1 +0.1 -0.1 -0.30 +2.4
1983 00.60 
+2.5 +0.1 -0.1 -0.50 +3.0
TOTL 12.47 48.0 
+0.1 +3.1 -11.3 62.1 


3-year peak: 18.0
7-year prime: 36.0
Career: 62.1

Bob Johnson

Year SFrac BWAA BRWA FWAA Replc WARP
1931 00.69 
+1.7 +0.0 +0.0 -0.50 +2.2
1932 00.77 
+2.6 +0.0 +0.0 -0.60 +3.2
1933 00.95 
+3.2 +0.2 -0.8 -0.60 +3.1
1934 00.91 
+3.5 +0.0 +1.0 -0.70 +5.2
1935 01.00 
+3.0 -0.1 -0.1 -0.80 +3.5
1936 00.97 
+2.5 -0.1 +0.5 -0.80 +3.6
1937 00.86 
+3.9 -0.1 +0.7 -0.70 +5.3
1938 00.99 
+3.9 +0.0 -0.1 -1.00 +4.7
1939 00.98 
+5.1 +0.2 -0.2 -0.80 +5.9
1940 00.90 
+2.6 +0.2 -0.2 -0.70 +3.2
1941 00.97 
+3.1 +0.0 +0.5 -0.70 +4.3
1942 00.98 
+3.5 +0.0 -0.1 -0.80 +4.3
1943 00.78 
+1.8 +0.2 +0.6 -0.60 +3.2
1944 00.96 
+4.8 -0.3 -0.3 -0.70 +4.9
1945 00.93 
+1.7 +0.1 +0.0 -0.70 +2.5
TOTL 13.64 46.9 
+0.3 +1.5 -10.7 59.1 


3-year peak: 16.4
7-year prime: 34.6
Career: 59.1

Kirby Puckett

Year SFrac BWAA BRWA FWAA Replc WARP
1984 00.85 
-1.7 -0.1 +2.6 -1.80 +2.5
1985 01.09 
-0.7 +0.1 +0.9 -2.20 +2.5
1986 01.05 
+3.1 +0.0 -0.6 -2.10 +4.6
1987 00.97 
+2.4 +0.0 -0.4 -1.80 +3.8
1988 01.02 
+4.5 +0.2 +0.6 -1.80 +6.9
1989 01.01 
+2.2 +0.1 +0.3 -1.70 +4.4
1990 00.90 
+1.2 +0.3 +0.0 -1.60 +3.1
1991 00.95 
+1.1 +0.5 +0.1 -1.70 +3.4
1992 01.02 
+3.1 +0.4 +0.6 -1.80 +6.0
1993 00.99 
+1.5 +0.2 -1.0 -1.90 +2.6
1994 00.98 
+2.3 +0.2 -0.6 -1.70 +3.5
1995 00.97 
+2.6 +0.2 -0.8 -1.70 +3.5
TOTL 11.80 21.6 
+2.1 +1.7 -21.8 46.8 


3-year peak: 17.5
7-year prime: 32.7
Career: 46.8

Atanasio Pérez

Year SFrac BWAA BRWA FWAA Replc WARP
1965 00.45 
+0.2 -0.1 -0.1 -0.10 +0.1
1966 00.41 
-0.7 +0.0 +0.0 -0.10 -0.5
1967 00.96 
+2.1 -0.1 -0.1 -1.20 +3.1
1968 01.03 
+2.5 +0.1 +1.2 -1.30 +5.1
1969 01.04 
+3.6 +0.1 +0.6 -1.10 +5.4
1970 00.99 
+5.5 +0.1 +0.2 -1.10 +6.5
1971 00.98 
+1.8 +0.1 +0.5 -1.20 +3.6
1972 00.89 
+3.8 +0.1 +0.4 -0.00 +4.3
1973 00.95 
+5.2 -0.3 -0.1 -0.00 +4.8
1974 00.97 
+1.8 -0.3 +0.4 -0.00 +1.9
1975 00.84 
+2.1 -0.5 +0.6 -0.10 +2.2
1976 00.87 
+1.7 +0.2 +0.7 -0.20 +2.8
1977 00.92 
+2.0 +0.0 +0.7 -0.30 +3.0
1978 00.88 
+1.8 +0.3 +0.3 -0.30 +2.7
1979 00.79 
+0.3 -0.3 +0.0 -0.30 +0.3
1980 00.93 
+0.0 -0.2 -0.5 -1.00 +0.3
1981 00.75 
-0.5 +0.0 -0.4 -0.80 -0.1
1982 00.31 
-0.1 +0.1 +0 0 -0.10 -0.1
1983 00.42 
-0.1 -0.2 +0.2 -0.20 +0.1
1984 00.22 
-0.6 +0.0 -0.2 -0.10 -0.8
1985 00.31 
+1.5 +0.0 -0.1 -0.10 +1.4
1986 00.34 
-0.1 -0.1 -0.3 -0.00 -0.6
TOTL 16.25 33.8 
-1.0 +3.6 -9.50 45.5
aTTL 14.22 35.8 
-1.0 +4.5 -8.50 47.6 


3-year peak: 17.0
7-year prime: 32.8
Career: 47.6

Ken Singleton

Year SFrac BWAA BRWA FWAA Replc WARP
1970 00.33 
+0.3 +0.0 +0.0 -0.30 +0.5
1971 00.54 
+1.3 -0.1 -0.2 -0.40 +1.4
1972 00.90 
+1.7 -0.8 +0.2 -0.70 +1.9
1973 01.02 
+4.2 -0.1 +1.0 -0.70 +5.8
1974 00.89 
+1.0 +0.2 -0.6 -0.60 +1.1
1975 01.05 
+5.1 -0.1 +0.5 -1.40 +6.8
1976 00.93 
+2.5 +0.3 -0.1 -1.20 +3.8
1977 00.96 
+5.3 -0.2 +0.2 -1.20 +6.4
1978 00.90 
+4.0 -0.2 -0.4 -1.10 +4.5
1979 01.01 
+4.5 +0.0 +0.1 -1.20 +5.9
1980 00.99 
+3.6 -0.2 -0.9 -1.20 +3.7
1981 00.95 
+2.2 -0.2 -0.5 -1.20 +2.7
1982 00.96 
+0.1 +0.1 +0.0 -0.00 +0.2
1983 00.89 
+2.2 -0.1 +0.0 -0.00 +2.0
1984 00.59 
-2.4 +0.1 +0.0 -0.00 -2.4
TOTL 12.91 35.6 
-1.3 -0.7 -11.2 44.3
aTTL 12.32 38.0 
-1.4 -0.7 -11.2 46.7 


3-year peak: 19.1
7-year prime: 36.9
Career: 46.7

Larry Doyle

Year SFrac BWAA BRWA FWAA Replc WARP
1907 00.41 
-0.3 -0.1 -1.0 -0.50 -0.9
1908 00.67 
+2.3 +0.1 -0.8 -0.80 +2.5
1909 01.01 
+3.8 +0.2 -0.8 -1.10 +4.1
1910 01.04 
+3.0 +0.3 -1.3 -1.00 +3.0
1911 00.95 
+4.6 +0.2 -1.0 -0.90 +4.7
1912 00.98 
+3.2 +0.2 -0.1 -1.00 +4.3
1913 00.87 
+1.7 +0.2 -0.2 -1.00 +2.7
1914 00.98 
+1.5 +0.0 -0.9 -1.20 +1.8
1915 01.01 
+4.5 +0.0 -1.0 -1.20 +4.9
1916 00.83 
+2.1 +0.1 +0.9 -1.20 +4.4
1917 00.84 
+0.5 -0.1 -0.1 -1.30 +1.5
1918 00.58 
+1.3 +0.0 +0.1 -1.00 +2.5
1919 00.75 
+2.6 +0.0 +0.3 -1.30 +4.2
1920 00.82 
+1.0 +0.0 -1.1 -1.60 +1.5
TOTL 11.74 31.8 
+1.1 -7.0 -15.1 41.2
aTTL 11.33 32.1 
+1.2 -6.0 -14.6 42.1 


3-year peak: 14.0
7-year prime: 29.6
Career: 42.1

Dale Murphy

Year SFrac BWAA BRWA FWAA Replc WARP
1976 00.11 
+0.0 +0.0 -0.1 -0.20 +0.2
1977 00.11 
+0.0 -0.1 -0.1 -0.20 +0.0
1978 00.87 
-1.3 -0.1 -0.2 -0.30 -1.4
1979 00.63 
+1.0 +0.0 -0.9 -0.30 +0.6
1980 00.93 
+3.3 -0.2 +1.3 -1.40 +5.8
1981 00.93 
+0.3 +0.1 -0.2 -1.30 +1.5
1982 01.03 
+4.3 +0.4 +0.3 -1.50 +6.5
1983 01.02 
+4.8 +0.5 +0.1 -1.50 +6.9
1984 01.02 
+4.4 +0.2 +0.7 -1.50 +6.8
1985 01.05 
+5.2 +0.5 -0.7 -1.50 +6.5
1986 01.02 
+2.2 -0.1 -0.7 -1.40 +2.8
1987 01.02 
+5.4 +0.2 +1.1 -1.00 +7.7
1988 01.00 
-0.2 +0.0 -0.4 -1.10 +0.5
1989 00.96 
-0.5 +0.0 -0.7 -1.10 -0.2
1990 00.93 
-0.4 -0.1 -0.5 -1.10 +0.0
1991 00.89 
+0.3 +0.1 +0.0 -1.00 +1.4
1992 00.09 
-0.7 +0.0 -0.2 -0.10 -0.9
TOTL 13.61 28.1 
+1.4 -1.3 -16.5 44.5
aTTL 10.65 31.0 
+1.7 +0.5 -13.7 47.0 


3-year peak: 21.4
7-year prime: 43.0
Career: 47.0

Orlando Cepeda

Year SFrac BWAA BRWA FWAA Replc WARP
1958 00.99 
+2.7 -0.2 -0.3 -0.40 +2.5
1959 00.99 
+3.4 +0.1 +0.0 -0.30 +3.8
1960 00.95 
+3.1 +0.1 -0.3 -0.60 +3.6
1961 00.98 
+4.9 -0.1 +0.5 -0.30 +5.6
1962 00.99 
+2.8 +0.0 +0.5 -0.20 +3.5
1963 00.94 
+5.2 +0.1 -0.3 -0.20 +5.3
1964 00.88 
+4.0 +0.0 -1.0 -0.20 +3.3
1966 00.84 
+2.6 -0.3 -0.6 -0.20 +2.0
1967 00.96 
+5.8 +0.2 +0.0 -0.20 +6.1
1968 00.99 
+1.1 -0.1 -0.1 -0.10 +1.0
1969 00.94 
+1.1 +0.1 +0.6 -0.10 +1.8
1970 00.91 
+3.0 -0.2 -0.1 -0.10 +2.8
1971 00.41 
+0.7 -0.2 -0.1 -0.00 +0.4
1972 00.15 
+0.4 +0.0 +0.0 -0.00 +0.4
1973 00.89 
+0.6 -0.7 +0.0 -0.00 -0.1
1974 00.17 
-0.7 +0.0 +0.0 -0.00 -0.8
TOTL 12.98 40.7 
-1.2 -1.2 -2.90 41.2
aTTL 11.92 40.8 
-0.5 -1.2 -2.90 42.1 


3-year peak: 17.0
7-year prime: 31.2
Career: 42.1

Rusty Staub

Year SFrac BWAA BRWA FWAA Replc WARP
1963 00.87 
-0.3 -0.1 +0.7 -0.20 +0.5
1964 00.48 
-0.7 +0.0 +0.2 -0.10 -0.5
1965 00.69 
+1.3 +0.1 -0.7 -0.50 +1.2
1966 00.92 
+2.2 +0.1 +0.0 -0.70 +3.0
1967 00.92 
+4.8 -0.2 -0.9 -0.70 +4.3
1968 01.01 
+4.0 +0.0 -0.1 -0.10 +4.0
1969 00.99 
+6.0 -0.1 -0.3 -0.80 +6.4
1970 01.00 
+4.3 -0.2 +0.8 -0.80 +5.6
1971 01.02 
+4.7 +0.0 -0.8 -0.80 +4.7
1972 00.43 
+1.8 +0.1 -0.3 -0.30 +1.9
1973 00.98 
+2.2 +0.1 +0.6 -0.70 +3.6
1974 00.95 
+1.6 -0.1 +0.2 -0.70 +2.3
1975 00.98 
+3.3 +0.1 -0.1 -0.70 +4.0
1976 01.02 
+2.9 +0.0 -0.8 -1.30 +3.4
1977 01.02 
+0.2 -0.1 +0.0 -0.00 +0.2
1978 01.09 
+1.2 -0.2 +0.0 -0.00 +1.0
1979 00.57 
-0.3 -0.1 +0.0 -0.00 -0.4
1980 00.57 
+1.2 -0.3 +0.0 -0.00 +0.9
1981 00.42 
+2.1 -0.3 -0.5 -0.20 +1.4
1982 00.37 
-0.4 -0.2 -0.4 -0.20 -0.8
1983 00.20 
+0.5 -0.1 -0.4 -0.10 +0.2
1984 00.12 
-0.2 +0.0 -0.2 -0.00 -0.3
1985 00.08 
+0.3 -0.1 -0.2 -0.10 +0.1
TOTL 16.70 42.7 
-1.6 -3.2 -9.00 46.7
aTTL 15.16 44.3 
-1.3 -2.8 -8.70 48.7 


3-year peak: 16.7
7-year prime: 32.6
Career: 48.7

Bobby Bonds

Year SFrac BWAA BRWA FWAA Replc WARP
1968 00.52 
+1.9 +0.1 +0.4 -0.40 +2.7
1969 01.06 
+3.6 +0.9 +1.4 -0.80 +6.8
1970 01.08 
+3.6 +0.6 +0.2 -0.80 +5.3
1971 01.02 
+4.5 +0.3 +1.1 -0.80 +6.6
1972 01.07 
+2.0 +1.1 +0.6 -0.80 +4.5
1973 01.09 
+4.3 -0.1 +0.9 -0.80 +5.9
1974 00.98 
+2.3 +0.3 +0.7 -0.70 +3.9
1975 00.93 
+4.0 -0.2 +0.4 -1.20 +5.4
1976 00.63 
+1.3 -0.1 +0.5 -0.80 +2.5
1977 01.00 
+3.0 -0.1 +0.1 -1.30 +4.4
1978 00.97 
+2.4 +0.2 -0.1 -1.20 +3.6
1979 00.93 
+2.3 -0.5 -0.2 -1.10 +2.7
1980 00.40 
-0.5 +0.2 -0.2 -0.20 -0.4
1981 00.42 
-0.1 -0.2 -0.3 -0.60 +0.1
TOTL 12.10 34.6 
+2.5 +5.5 -11.5 54.0
aTTL 11.70 35.1 
+2.3 +5.7 -11.3 54.4 



3-year peak: 19.3
7-year prime: 38.9
Career: 54.4

Norm Cash

Year SFrac BWAA BRWA FWAA Replc WARP
1959 00.20 
+0.6 +0.0 -0.2 -0.10 +0.4
1960 00.66 
+3.2 +0.0 +0.3 -0.20 +3.7
1961 00.98 
+8.0 +0.0 +1.0 -0.30 +9.3
1962 00.92 
+3.3 -0.1 +0.9 -0.20 +4.3
1963 00.88 
+3.7 -0.2 +0.7 -0.20 +4.3
1964 00.82 
+2.3 +0.0 +0.6 -0.20 +3.1
1965 00.82 
+4.1 -0.2 +0.8 -0.20 +4.9
1966 01.02 
+4.1 +0.0 +0.3 -0.20 +4.6
1967 00.87 
+3.0 +0.0 +0.5 -0.20 +3.7
1968 00.69 
+2.7 -0.1 +0.7 -0.10 +3.4
1969 00.82 
+2.6 +0.0 +0.5 -0.10 +3.1
1970 00.67 
+1.8 -0.1 +0.0 -0.10 +1.8
1971 00.78 
+4.0 +0.0 -0.2 -0.00 +3.8
1972 00.79 
+2.6 -0.4 +0.8 -0.00 +3.1
1973 00.61 
+1.3 -0.2 +0.2 -0.40 +1.6
1974 00.25 
+0.4 -0.2 +0.0 -0.20 +0.3
TOTL 11.78 47.7 
-1.5 +6.9 -2.70 55.4 


3-year peak: 18.8
7-year prime: 34.9
Career: 55.4

Willie Stargell

Year SFrac BWAA BRWA FWAA Replc WARP
1963 00.49 
+0.5 -0.1 -0.7 -0.30 +0.2
1964 00.66 
+1.5 +0.0 -0.7 -0.50 +1.4
1965 00.86 
+2.5 +0.0 +0.9 -0.60 +4.1
1966 00.81 
+5.0 -0.1 -0.1 -0.60 +5.4
1967 00.80 
+2.6 +0.1 -0.4 -0.60 +2.9
1968 00.74 
+2.2 +0.2 +0.1 -0.50 +3.1
1969 00.88 
+5.0 +0.1 -0.4 -0.70 +5.4
1970 00.77 
+1.8 -0.1 +0.4 -0.60 +2.7
1971 00.90 
+6.8 -0.1 -0.3 -0.70 +7.1
1972 00.88 
+5.4 -0.1 -0.6 -0.00 +4.6
1973 00.90 
+6.7 -0.2 -0.1 -0.60 +7.1
1974 00.88 
+5.4 -0.1 -0.4 -0.60 +5.5
1975 00.77 
+3.4 -0.1 -0.2 -0.10 +3.2
1976 00.72 
+2.0 +0.0 -0.5 -0.10 +1.7
1977 00.32 
+1.3 -0.1 -0.1 -0.10 +1.1
1978 00.67 
+3.4 +0.0 -0.3 -0.20 +3.3
1979 00.71 
+2.4 -0.3 +0.0 -0.30 +2.4
1980 00.34 
+1.0 -0.1 -0.3 -0.20 +0.8
1981 00.15 
+0.2 +0.0 -0.2 -0.10 +0.0
1982 00.13 
+0.2 +0.0 -0.2 -0.10 +0.0
TOTL 13.38 59.3 
-1.0 -4.1 -7.50 62.0 


3-year peak: 19.7
7-year prime: 39.2
Career: 62.0

Ranked by 3-year peak: Murphy, Stargell, Bonds, Singleton, Cash, Smith, Puckett, Pérez, Cepeda, Staub, Johnson, Doyle

Ranked by 7-year prime: Murphy, Stargell, Bonds, Singleton, Smith, Cash, Johnson, Pérez, Puckett, Staub, Cepeda, Doyle

Ranked by career: Smith, Stargell, Johnson, Cash, Bonds, Staub, Pérez, Murphy, Puckett, Singleton, Cepeda, Doyle
   165. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 11, 2007 at 02:00 PM (#2571603)
What to make of this:

1. Self-described career voters tend to give lots of love to Staub and Pérez, but they didn't actually have particularly valuable careers. They played for an exceedingly long time, enabling them to rack up the counting stats, Win Shares, and WARP, but virtually everything outside their primes is right around replacement level. All of Staub's seasons above 2 WARP were in his 1966-76 prime, and all of Pérez's in his 1967-78 prime. You could really just ignore all their other years and not miss much of anything. When you strip away their countless stat-padding seasons that did virtually nothing to help their teams win games, they sink to the bottom of the pile.

2. I don't support Dale Murphy, but the pure peak voters should really be gravitating to him. From 1980-87 he was an absolute beast--he had six seasons when he was a serious MVP candidate. That's a peak and prime that is almost certainly above the HoM median. He had absolutely nothing outside those years, of course, but he was a truly elite player for the better part of a decade. The votes that go to guys like Al Rosen should be headed his way.

3. Let's get Reggie Smith right--he is the true career candidate among the offense-first backloggers. By contrast to guys like Staub and Pérez, he was actually well above average every year of his career, making a meaningful contribution to his team's victories. His peak and prime are middling--Murphy blows him away on that basis, and Bonds was certainly also superior--but he just offers the most total value above replacement.

4. Bob Johnson is just the poor man's Reggie Smith. His profile is virtually identical to Smith's (lower rate, better durability, similar package), but a lesser version: no legit MVP-caliber year (once a war discount is applied to 1944), slightly inferior peak and career. If you factor in the fact that Johnson played in segregated leagues and Smith didn't, then the gap seems much larger, but on pure MLB value Johnson still just falls a nudge short of Lil' Reggie.

5. Like Murphy, Bobby Bonds should be taken more seriously by the electorate, although I don't support him. He was that rarest of breeds, a true five-tool player--seasons like his 1969 were wonderful do-it-all years. His peak and prime are meaningfully superior to all the backlog save Murphy, and he has much more career than Murphy. There's certainly a case for him. Singleton is just a marginally weaker version of him, as Johnson is of Smith.

6. It's not that you can't make my PHoM playing 1B. It's just that you have to do something more than hit the ball at a merely above-average rate. Either you tear the snot out of it, or you bring something else to the table. Norm Cash can match the competition like Cepeda and Pérez as a hitter, but he was also the best-fielding 1B of his era, giving him a substantial advantage over them. And he has a magical peak season for the ages that they lack. Again, he doesn't have enough to make my ballot, but he's worthy of consideration.

7. Doyle, Puckett, and Cepeda are joke candidates as far as I'm concerned (as well as Staub and Pérez). They come out near the bottom on all measures. Puckett's presence in the top 10 returnees is a disgrace.

I hope this is helpful. Please let me know if you have any questions or if anything seems out of whack.
   166. DL from MN Posted: October 11, 2007 at 03:31 PM (#2571719)
You left Oms and Cravath out of your analysis. Is it due to credit issues?
   167. Joey Numbaz (Scruff) Posted: October 11, 2007 at 03:48 PM (#2571746)
When you strip away their countless stat-padding seasons that did virtually nothing to help their teams win games, they sink to the bottom of the pile.


Actually, they don't. They sink to the middle of the pile in your career numbers. They rank 6 and 7 of the 12 players listed.

Question regarding Johnson - do the numbers above apply the war discount you mention in point #4? They appear to, but I want to confirm. Also, does your new spreadsheet include this war discount for all appropriate player seasons?
   168. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 11, 2007 at 04:12 PM (#2571789)
DL from MN, that's right. It's one thing if it's 1 year (Smith) or two (Johnson), but when you're talking about a half a career or more, it's just in the eye of the beholder.

Joe Dimino--I meant they sink to the bottom of the pile looking at the peak/prime/career package. As far as I'm concerned, you can't have a claim as a "career" candidate if you don't even have 60 WARP2.

I applied the war discount to Johnson for this post. There is definitely no war discount in my spreadsheet--it's up to every voter to decide how and how much to discount.
   169. Jim Sp Posted: October 11, 2007 at 08:22 PM (#2572088)
As someone looking to start a PhD this fall (I would be doing it in International Relations) can I ask you why you quit, especially at that stage?

I did the same thing, passed the PhD qualifying exams in Applied Math then took a software job. Many years ago in my case though...

Academic politics are horrible, much worse than what I found in the working world. Bottom line is that working for a company is more pleasant, more interesting, and pays (much) better with shorter hours. Of course my experience is in software, becoming a lawyer would not likely be more pleasant. I have one friend from college who became a happy lawyer, they exist but the odds are against you. However a law background will be useful whatever you end up doing, lots of lawyers get off the big law firm track and do something more agreeable down the road.

If you go the PhD route, remember that you're not obliged to stick it out, and that finishing the PhD can actually hurt you in the job market (sticking it out shows that you really want to be an academic and just can't find a job).

Becoming an academic is like becoming a minister, the people who really should go that way have usually known it for a while.

On the other hand joining a PhD program is a great way to get someone else to pay for your Master's degree. If you can pull that off, it's an excellent value.
   170. andrew siegel Posted: October 11, 2007 at 08:47 PM (#2572106)
On the Ph.D vs. law question, I've got a foot in both camps. I started a Ph.D in history and stuck around through my coursework and exams but ultimately left b/c/ I didn't feel the same calling for my subject that others did. I went to law school unsure of whether I wanted to practice or have an academic career in an area more connected to the modern world. I was lucky to do well enough in law school that I could pursue the dream job of teaching and writing about constitutional law. I genuinely feel that I have the best of both worlds. If I had to pick between being a non-law school academic or being a practicing lawyer, it would be difficult for me. If you want to discuss this further, feel free to email me privately.
   171. Dizzypaco Posted: October 11, 2007 at 08:52 PM (#2572110)
I have followed the debates over HOF voting closely since near its beginning, although I haven't voted for reasons discussed in other threads. I still find them interesting, except when Dan R. dominates the discussion, combining Baseball Prospectus like arrogance with highly questionable opinions based on vast over-confidence in a flawed methodology. I generally don't make comments on other posters, and I suppose I'm an outsider here, so sorry for my intrusion, but I couldn't resist adding my two cents.
   172. andrew siegel Posted: October 11, 2007 at 09:16 PM (#2572127)
I for one have no problem with Dan's tone. I do however remain agnostic or opposed to him on a variety of methodological questions (primarily involving how to compare positions and eras). However, whatever you think about his tone and his position on the more controversial methodological questions, I think it is very important that we acknowledge that his system is the only comprehensive system we have that does 3 important things:

(1) Acknowledges the wider variance in fielding performances that most empirical studeis find.
(2) Tries to account for non-SB baserunning.
(3) Bases its evaluation of annual (and therefore career) value on a reasonable baseline, threby elminating the career value that WS and WARP attribute based on useless hanging around time.

I have found that, no matter what system you work off of, if you make these adjustments your results end up looking quite a bit like Dan's.
   173. Dr. Chaleeko Posted: October 11, 2007 at 09:45 PM (#2572146)
Diz,

Would you mind elaborating just a bit on what assumptions you find flawed in Dan's system? I don't ask because I think either of you is right, I ask as someone interested in incorporating Dan's work into my own thinking.
   174. sunnyday2 Posted: October 11, 2007 at 10:04 PM (#2572159)
Diz,

I don't know if we are actually electing anybody different than we would have anyway. But now after 100 years, explorations that used to ride the wind for 100 posts or so can now be successfully concluded in a post or two without any of those annoying twists and turns. The mathematicians have routed the philosophers. I sorta feel like Dave--you know, "Open the pod bay doors, Hal," and Hal says, "I'm sorry, Dave, I can't do that."

The HoM used to be a journey. Now it's a destination.
   175. Chris Cobb Posted: October 12, 2007 at 12:02 AM (#2572212)
I also have no problem with Dan's tone. I wouldn't say he talks about his results any differntly than, say, Joe Dimino talks about his pitchers' pennants added results, except than Dan has the time to say a lot more. If Joe had more time to present his results in a season-by-season fashion and comparatively as Dan does, I think we might be moving towards more consensus on the pitchers. The fact that Joe's careful and thorough data analysis hasn't had more influence shows that Dan's heavily-engaged approach is necessary to jog the electorate into working through new evidence and examining carefully the metrics that we rely upon to develop our rankings.

Moreover, it seems to me that Dan has given our recent discussions a vitality and a purposefulness that they were beginning to lack as the project has been reaching its end, and we face the borderline players who are very difficult for us to distinguish from one another. I often feel bogged down by the weight of so many players to consider, so much data to crunch anytime I want to take a rigorous look at a problem. Dan's _data_, as much or more than his argumentation, has been immensely valuable.

Maybe I'm not troubled by Dan's tone because he seems to me generally to be right, but it looks to me like he is tirelessly willing to make reasonable arguments, bacls up his claims with evidence, and owns and corrects errors when he makes them.
   176. sunnyday2 Posted: October 12, 2007 at 12:42 AM (#2572241)
Doc,

Dan's method assumes that the bottom 3/16 of ML regulars at a given position is the right baseline or (what is usually referred to as) "replacement" value and more to the point, it posits that the levels at which those players perform is driven, mechanistically, by the nature of the game at any given time. He says that it is easier or harder for players to dominate, based on factors intrinsic to how the game is played. IOW the FAT level is not subject to random fluctuations in the way that the top level of talent (think ABC or Gehrig-Foxx-Greenberg-Mize or Ripken-Yount or Brett-Schmidt or Berra-Campanella) is quite obviously subject to random fluctuations in the talent pool. There's a circularity to it, I think. It is what it is. Well, yeah, but are we sure that the bottom 3/16 aren't just as random as anything else in the talent pool?

And despite a rigorous mathematical exercise which makes very strong claims, DanR still applies a pretty stiff coating of bullshirt over the top. Why, if the numbers are so good? And what IS in that bullshirt filter, and why?

Chris,

But does he have to "correct" every other post on every thread? 75 years ago, somebody said you were the "soul" of this project and your posts always left a lot of room for the rest of us to (insert idea) or (insert data) here. Today we are all superfluous. Anything you or I can say, Dan can say it better.

So maybe this project is better off, if you think we are more correct. But even if that's the case, like I say, the scenery along the journey has gone from a mountain path to an autobahn. The poetry has given way to the equation. And I'm arguing that the poetry is not just more fun, it is sometimes more true.
   177. sunnyday2 Posted: October 12, 2007 at 12:44 AM (#2572245)
PS. I will shut up now. No more comments about this topic. (My consensus score on this is probably just as low as on my ballot. After all, Chris has always been the consensus guy.)
   178. rawagman Posted: October 12, 2007 at 02:40 AM (#2572380)
I haven't been very active recently (moving back to Canada, finding a job, starting a new job, moving into oru own apartment, etc.) As one who by nature is the underlying purpose, does not beleive in the concept of a master number of value measurement (neither WS, nor WARP, WAR nor PA, or anything else), I am not a proponent of the DanR mwthodology. I read all of the posts, of course, in an attempt to decipher (in sunnyday's term) some poetry in the theory. And there is poetry there. It might be that I think in an opposite way to most people, not just here, but in most cases. As Bill James said, baseball is about winning games, and to win games you need to score runs. These things can, and are counted in numbers. These numbers come from a plethora of other numbers that, in normal circumstances amount to a value of runs. The other numbers come from context. The poetry of the game.
Of course, the poetry and the numbers are inextricably and irretrievably linked. Growing up, the back of every baseball card was a poem. The poem told me of the greats that can be accomplished by man with talent and hard work. Beautiful, epic poetry.
It is the type of poetry that inspires children to dream, and maybe to act. Maybe to play baseball.
It seems to me that most here, if not all, look at the numbers of baseball, those recorded actions made historical on scorecards, and try to find a way of simplifying the cluster of digits into a single value. That is, to my mind, a way of distancing the game from the context and the magic of what started me on this obsession.
A poem can inspire a dissertation, or it can inspire another poem. What is better?
I have attempted to explain my units of measurement in my ballot introductions and will attempt an additional condensation here. I take a number of numbers that are basic, and to me, retaining of the poetry of the game and try to find a way to get back from them to the field itself. The number should help me make fanciful managerial decisions. What type of manager am I? I don't care for elaborate strategy. I want to let the players play ball. Let them win it from the mound, at the bat and with the glove. Win it away from the dugout - on the field. I want to watch good baseball.
   179. Brent Posted: October 12, 2007 at 03:23 AM (#2572467)
I have some sympathy for Dizzy's comment in the sense that as this project approaches its end, it seems to me that it's become a lot less fun. I don't think Dan should be singled out, because I think it's partly just that the stuff I found most interesting--learning about baseball's history and the stuff that isn't measured in the formulas--has become less important as we've moved into the modern era. Also, as time runs out on our pet candidates, too many of the discussions have become strident. We've always had our vigorous debates and over-stated opinions (I'll never read Jake Beckley's name again without cringing), but in elections past they were tempered by discussion of issues that none of us fully understood and that we explored together. That was a lot more interesting to me than endless repetition of arguments on replacement levels and standard deviations.

In many ways, I think Dan has a great system. If I were going to limit myself to using any single ubersystem, I'd choose his over the alternatives. But I'd also appreciate it if Dan would recognize that reasonable voters can make different assumptions and draw different reasonable conclusions. For example, in post # 166 Dan says, "Puckett's presence in the top 10 returnees is a disgrace." I can assure Dan that when I read this comment I didn't suddenly exclaim, "how foolish of me to have supported Puckett!"

In fact, while largely agreeing with his analysis, but making two (IMO) reasonable changes in assumptions, I think its perfectly defensible to rank Puckett ahead of Reggie Smith. The first is that I'm skeptical of all of the available fielding systems, including zone ratings. The evidence I've seen suggests that the measurement errors of these systems are large. Because I think the most reliable measure of fielding is opinion of knowledgeable observers (including Gold Glove awards), my assessment of Puckett's fielding is much more positive than Dan's.

The other difference is that (as discussed by Paul in # 157 and by others earlier in this thread) I regard the cost of lack of in-season durability to be a bit higher than implied by the single replacement level. Teams may not have ready access to a replacement player at every position. Often, there are hidden costs as other players on the team are forced to play out of position or rosters are juggled. And every dip into the replacement pool is, to some degree, a crap shoot. Although I can't mathematically prove it, my experience as a baseball fan teaches me that the cost of an in-season replacement can be quite a bit higher than the simple model of replacement level would suggest.

If we added, say 1 or 1.5 fielding wins a year to Puckett's prime seasons, and subtract maybe 0.5 wins a season from Smith's partial seasons, it suddenly becomes very reasonable to place Puckett ahead of Smith. I won't argue that everyone has to agree with my assumptions, but I don't appreciate being told that a ballot based on my alternative assumptions is "a disgrace."
   180. 'zop sympathizes with the wrong ####### people Posted: October 12, 2007 at 04:40 AM (#2572563)
I was lucky to do well enough in law school that I could pursue the dream job of teaching and writing about constitutional law.

If Google speaks the truth, then this is the understatement of the century. Tip o' the hat to you, sir.
   181. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 12, 2007 at 05:23 AM (#2572578)
Yeesh! This is supposed to be the 2005 results thread, not the Dan's-worth-as-a-human-being thread...I had no idea I was such a polarizing figure! I suppose bad press is better than no press, right? Anyways...

I'm not really sure how to respond to Dizzypaco, except to say that's a pretty classy ad hominem attack from someone who doesn't even participate in the HoM. If you have a substantive criticism, I'm all ears--I am always looking to make improvements, and indeed my research has been tremendously helped by the thoughtful comments of numerous posters. But if you're just looking to provoke a flame war, I'm afraid I prefer to bow out.

Andrew Siegel, as above, I'd love to hear your criticisms of my methodology, particularly if you have suggestions for how it might be improved. I am extremely glad to hear you have found some aspects of it valuable.

Chris Cobb, nothing to say but thanks and that I very much appreciate your support. You are right that I have invested an immense amount of time in this, and it is extremely rewarding to know that thoughtful voters like yourself feel I have made meaningful contributions to the project.

sunnyday, first, it's the bottom 3/8, not 3/16, not that that's relevant. On the substantive points:

1. You have to distinguish between the FAT calculation and the standard deviation calculation. I do NOT argue that changes in the worst-regulars average are always, or even ever, caused by changing conditions or league factors. I do think they are less subject to random fluctuation than star gluts, because each year's result comes from a sample of at least 40 player-seasons and as many as 100 player-seasons, so the odds that it would bounce around wildly by chance are pretty small, but of course they're not zero. I have done no regression analysis on FAT and thus I have absolutely no idea about what factors might be driving its evolution, only educated guesses (like turf lowering SS rep level in the 1970s). The changes in FAT over time could be caused by anything, or by nothing at all. My contention on FAT is that regardless of what causes its movements, it is the appropriate baseline for determining Merit, but that is of course a normative rather than a positive issue.

2. I DO argue that leagues are easier or harder to dominate based on factors intrinsic to how the game is played, since that is something I have studied and gotten highly statistically significant results on. Again, the correction I make for standard deviation is NOT based on whether a league was actually dominated or not--it is simply determined by the league factors that I have found correlate to ease of domination. If you tell me a league's run scoring, years since expansion, the winning percentage of its worst team etc., I will tell you how easy or hard it was to dominate, without knowing anything about how the players in it actually performed. Some leagues (like the 1926 NL) were easy to dominate but the players didn't take advantage, while others (like the 1992 NL) were hard to dominate, but many players had phenomenal seasons in spite of those obstacles. There is no circular logic here, and no confusion about which way the arrow of causality points. That's what regression is for--to separate out signal from noise.

3. What "bullshirt" are you referring to? For post-1893 MLB position players (the ones I have numbers for), the only adjustment I make to the results my system gives me is for segregation. I think we have elected too many players from both the Negro Leagues and the major leagues that existed alongside them, and not enough from the game's other eras. Babe Ruth would have had a lower OPS+ if he had been in the same league as Oscar Charleston, but Charleston's MLE's would also be lower if he were being translated to a more difficult, integrated league. My system can't account for that factor, since it doesn't "know" the Negro Leagues existed, so I have to input it ex post facto. Other than that, I vote straight down the line of my what my numbers tell me for the players covered by my system. I certainly handle pre-1893 players, Negro Leaguers, and pitchers just by guesswork, but that's because I don't yet have a rigorous quantitative system to evaluate them.

4. I "correct" factually inaccurate statements when I see them. I would never say that you are "wrong" to have a certain player in a certain position on your ballot, or to weight peak heavily, or to use Win Shares--those are normative, voter's prerogative choices. I would, and did, "correct" you when you suggested that because 2B was an offense-first position when Larry Doyle played it, the magnitude of the gaps between his fielding and that of the rest of the league would be lesser than they were for defense-first positions, since that contention is simply at odds with the extensive data we have on the standard deviation of fielding at the various positions. It may be true that before I returned to the project, empirically false statements such as that one would have gone by unremarked-upon, and might even have influenced other voters' decisions. I hope you're not arguing that things were better that way.

Brent, I obviously think Puckett would be a terrible selection, and I am mustering every argument at my disposal to convince the group of that position. We all express strong opinions on the candidates, and many voters often devote screeds to bashing one player or another, often calling them a potential "worst mistake." Why is my ripping on Puckett any different?

Moreover, I think it is a benefit to the HoM to have you clarify for all of us your methodology and reasoning on putting Puckett ahead of Smith. It not only contributes to the group's analysis of those two players and the underlying issues, but it also shows you to be a methodical, logical, rigorous, and consistent voter, which may lead other members of the electorate to place more faith/weight in your analyses and opinions. All this discussion seems positive to me. I don't understand what you're saying the problem is--my saying Smith is infinitely superior to Puckett, and your counter-arguing quite effectively, seems to me to be a great example of the system working perfectly.
   182. Howie Menckel Posted: October 12, 2007 at 01:06 PM (#2572672)
"Chris has always been the consensus guy."

a-HEM!
:)

Just kidding.

It's not that I have a problem with Dan's tone myself, as much as I wonder if it's as effective as it could be, given the data.

I did have a specific question above (139), and got a specific answer (145).
   183. Dizzypaco Posted: October 12, 2007 at 01:53 PM (#2572714)
I suppose I should follow up - I'm not a voter, but I have participated in many threads throughout the "years". I should also be more specific - I disagree with some specific elements of Dan's methodology pretty strongly - for example, I believe he sets replacement value too high, causing him to undervalue players who are either durable (Kirby Puckett), or who have long careers in which they put up several years of good but not great performance. I also question some of the defensive statistics used, as well as the level of confidence in the defensive statistics (see the Darrell Evans discussion).

People disagree with each other all the time in these threads - its what its all about. But I stand by my comment that Dan's typical comments are often set up as "I'm right, and I am here to educate all of you", when his opinions are no more valid (and no more based on statistical validity) than anyone else's. Post 165 has several good examples. And he dominates many of the recent threads. Its really about tone.

The example is the discussion of Kirby Puckett and Reggie Smith is a good example. Puckett is pretty clearly superior to Smith using win shares. Comparing their best seasons, top two, top three, top four, and so on, Puckett wins out. Puckett's career was relatively short, so after the best 12 seasons, Smith starts to pull ahead I believe. In my mind, a very good argument can be made that Puckett is better than Smith, and I would place him higher.

At the same time, I can understand why people would prefer Smith as well - it can be argued either way. Some people don't like win shares for various reasons. The statement, "Puckett's presence in the top 10 returnees is a disgrace." is the type of statement that is a little insulting to those who support Puckett. My issue was more about tone than substance.

Finally, I've said it before, but replacement value is a theoretical concept. There is no right or wrong answer. People differ on how to define the issue, which is fine, but I feel there should be more acknowledgement that this is not something people need to be educated about.
   184. John (You Can Call Me Grandma) Murphy Posted: October 12, 2007 at 04:41 PM (#2572903)
I actually like Smith more than Puckett using Win Shares, FWIW.
   185. Dr. Chaleeko Posted: October 12, 2007 at 05:44 PM (#2573004)
Diz, thanks for supplementing your previous comments.

I agree with everyone, amazingly. I think Dan has at times expressed his opinion in black/white ways that are a little edgier than I like, but so have I (as he says), and so have others. I agree that talk of SDs and Replacement levels isn't much fun, but I also think those concepts are worth discussing, and that they offer different insights into how players build value/merit. I'm glad that Diz, for instance, challenged him on this because it's important to remind ourselves of the fluid nature of ideas and definitions.

I think Dan's been very open with his system, and he's always solicited ideas for revisions. As someone who has pushed a system or two (MLEs, two different versions, and a little on my Keltner-based system), I understand Dan's position well. And his system returns results that feel idiosyncratic in certain instances...just as any new system does...but that are well documented and leave room for me to decide if I agree.

And, yes, his system is currently based on extrapolation of certain information, but based on his actions, I think if presented with new information that gave a fuller picture of the relationships his system is based on, he would revise his assumptions and his system to match the empiraclly observed relationships. I think he's working with the best set of assumptions he could make and is willing to reconsider and that's what I ask for in the moment.

The fact that Dan is currently the most dominant poster in the group is neither good nor bad, it just is. As Chris points out, he's got the time and the data to do it. I say that's fine, we are all entitled to our opinions, and we are entitled to challenge one another's points of view. Calling the kettle black, I'd encourage Dan to avoid sharp language in making normative statements (as in the Puckett instance above), or to reframe them as "I believe it would be a disgrace if..." but I can also challenge him back as Brent did and tell him to cool it a bit if I find the tone too much as has already happened.

But overall, Dan's a very valuable poster, he brings important ideas up, he challenges us all to reconsider our opinions and provides reasons to do so. That's good, and many of us have delivered responses that are helping to build a great dialogue.
   186. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 12, 2007 at 05:50 PM (#2573012)
Dan is currently the most dominant poster in the group


But is the group easy to dominate? Let me run a regression and I'll get back to you. ;)
   187. Chris Cobb Posted: October 12, 2007 at 05:59 PM (#2573024)
Finally, I've said it before, but replacement value is a theoretical concept. There is no right or wrong answer. People differ on how to define the issue, which is fine, but I feel there should be more acknowledgement that this is not something people need to be educated about.

I see several conceptual shortcomings to this statement.

1) Replacement level is not just a concept. Every time a player is replaced, there is a replacement level, in that the replacement player's performance will be of some quality. That quality is the level of that replacement.

2) The replacement levels that we use in our analyses of merit are statistical estimations of the typical quality of replacement players of a given type. They can therefore be more or less accurate, and more accurate estimates will be better than less accurate ones. This is, to use the helpful terms Dan R has introduced, an empirical matter.

3) What kind of replacement level to estimate in order to use it as a baseline for evaluating merit is a normative question, not an empirical one. However, insofar as one views merit as corresponding to a player's value to his team, there are, again, better and worse arguments about which replacement levels most accurately capture the baseline for a player's value. For example (and this has been said before) win shares is not empirically incorrect to set its zero point for batting win shares at the level it does. However, the case that this zero point corresponds to the level of offense at which a position player becomes valuable to his team is a weak one.

If the electorate takes the view that "replacement level is a theoretical concept, so there are no right and wrong ways to define it. Therefore, any setting of replacement level is a matter of personal taste that cannot be the subject of reasoned argument," then it is likely that a) we will measure kind of replacement that we care about less accurately, b) that we will make poorer choices about which kind of replacment level to use as a baseline for value, and c) that we will misapply the data that we have, because we have not recognized the need to think carefully about the meaning of the numerical values it offers us.
   188. Brent Posted: October 13, 2007 at 12:11 AM (#2573534)
I guess my opinion on replacement levels is between Dizzypaco’s and Chris Cobb’s.

Replacement level is not “just” a concept—when a player’s performance drops to a certain level, he loses the opportunity to play in the major leagues. Major league benches and AAA teams have dozens of players who can play just as well as the worst regulars in major league baseball.

On the other hand, replacement level is a concept. The idea that there exists a clear, fine line splitting players who deserve major league jobs from those who don’t plays the same role in sabermetrics that “perfect competition” plays in economics or the “frictionless plane” plays in physics. Replacement level is a model of reality. In the real world, there are costs to acquiring “freely available talent,” replacement decisions depend on comparisons of real players rather than mathematical lines, and the decisions vary depending on the circumstances facing the team.

For many roster changes, replacement level has little relevance. For example, I assure you that this winter when the Yankees and the Mets reorganize their rosters, they will decide to shed a number of players whose performance was well above replacement level.

To determine a replacement level (that is, to make the concept operational), we must first make some assumptions either regarding which type of players are likely to be replaced, or which actual replacements should figure into our determination of replacement level. Making these assumptions will inevitably involve some degree of arbitrariness. Nevertheless, I think it is possible to develop a consensus about a range of reasonable assumptions, which should lead to a range of I guess my opinion on replacement levels is between Dizzypaco’s and Chris Cobb’s.

Replacement level is not “just” a concept—when a player’s performance drops to a certain level, he loses the opportunity to play in the major leagues. Major league benches and AAA teams have dozens of players who can play just as well as the worst regulars in major league baseball.

On the other hand, replacement level is a concept. The idea that there exists a clear, fine line splitting players who deserve major league jobs from those who don’t plays the same role in sabermetrics that “perfect competition” plays in economics or the “frictionless plane” plays in physics. Replacement level is a model of reality. In the real world, there are costs to acquiring “freely available talent,” replacement decisions depend on comparisons of real players rather than mathematical lines, and the decisions vary depending on the circumstances facing the team.

For many roster changes, replacement level has little relevance. For example, I assure you that this winter when the Yankees and the Mets reorganize their rosters, they will decide to shed a number of players whose performance was well above replacement level. To determine a replacement level, we must first make some assumptions either regarding which type of players are likely to be replaced, or which actual replacements should figure into our determination of replacement level. There will inevitably be some degree of arbitrariness. Nevertheless, I think it is possible to develop a consensus about a range of reasonable estimates of the replacement level itself.
   189. Brent Posted: October 13, 2007 at 12:16 AM (#2573575)
Oops--sorry about the duplicative text. A cut and paste must have gone askew. An example of a case where I really wish I could edit what I just posted.
   190. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 13, 2007 at 12:53 AM (#2573676)
It seems to me that the burden of proof should fall pretty strongly on Brent and Dizzypaco to show--with real data--that the consensus placement of replacement level at 80% of positional average is incorrect. The fact that both Keith Woolner's study of bench players and Nate Silver's study of minimum-salary veterans got roughly the same result is what makes me consider replacement level "settled law," to quote John Roberts, and not an open question. If there IS a well-executed study with a different conclusion, however, it would tremendously benefit the group--and the entire baseball analysis community!!--if you guys would present its findings.
   191. rawagman Posted: October 13, 2007 at 02:30 AM (#2573916)
Dan - I don't think that Brent is saying that 80% is incorrect. I think his deeper point is that it is irrelevant. It is not a quantifiable measurement that a GM can actually take into account of its own accord when forced to make a mid-season decision.
   192. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 13, 2007 at 03:28 AM (#2574058)
But it can't be irrelevant--you simply have to use SOME replacement level, whether it's absolute zero (counting stats), 50% of average (Win Shares), 80% (me), or 100% (BRAA, RCAP, etc.). There's no way to just punt on the question. The only issue is what replacement level you use. And Brent's statement that "Although I can't mathematically prove it, my experience as a baseball fan teaches me that the cost of an in-season replacement can be quite a bit higher than the simple model of replacement level would suggest" sounds pretty flimsy to me. Of course it CAN be quite a bit higher. It can also be quite a bit lower (as Brent notes when referring to the Yankees and Mets). What is usually called replacement level is simply the weighted average of all those possibilities.

Brent is of course free to use whatever replacement level he sees fit. But I do think he opens himself up to criticism by subtracting half a win (which is a LOT) from partial seasons without providing *any* hard evidence to suggest that replacement players would likely perform at far worse than the widely recognized consensus of 80% of positional average.

As far as I can fathom, the only two empirically valid approaches to determining a replacement level are:

1. Use the average of replacement players at the position, historically around 80% of positional average.
2. Use the rate production of the actual player(s) who did replace the player in question when he wasn't on the field.

The latter option, which would be more of a Win Shares-type approach, has two fatal flaws: you can't use it when someone plays every game, and when the bench player is good it will give the starter negative value (Wally Pipp, anyone?). So that leaves #1, which is the path I've chosen to take in my WARP.
   193. Brent Posted: October 13, 2007 at 03:35 AM (#2574067)
Two studies with similar results makes the estimate "settled law"? In the social sciences, that would qualify it as "an interesting hypothesis with some empirical support."

Rawagman is right that I'm not disputing the 80% figure (though I'd be more likely to couch it as a range, say 75 to 85%). "Irrelevant," however, is to strong. Replacement level is certainly something GMs should be aware of and sometimes plays a role in their decisions. But I'll bet most roster moves don't involve replacement calculus; management is simply asking if player A is available and is better than player B. If we think that a GM can easily pick up a replacement player every time someone goes on the DL, I think we're kidding ourselves. I'd prefer to say that the importance of replacement level is being oversold.
   194. Brent Posted: October 13, 2007 at 03:55 AM (#2574089)
Let me give an example of why I think in-season replacements are so tricky to measure. In late June, Nationals shortstop Cristian Guzman was injured and sidelined for the rest of the season. Felipe Lopez was moved from second to short and Belliard took over at second. To the researcher, it looks simple--Guzman left the regular lineup and Belliard came in; Belliard played pretty well (certainly above replacement level); little net cost to the Nats. What that misses, however, is that Lopez was having difficulty covering shortstop and--as often happens when players are struggling in the field--went into a slump with the bat. A simple statistical analysis will just say that Guzman's slot was filled by an above-replacement player and Lopez had a bad season. My intuition, though, tells me that Lopez's bad year may have been an indirect result of Guzman's injury.

It seems to me that most real-life substitutions have complexities that would be difficult for statistical studies to control for. Before we declare something "settled law," let's consider that the issues are pretty complex.
   195. Brent Posted: October 13, 2007 at 04:09 AM (#2574112)
Also, I'd like to remind people that Bill James devoted a 3-page section of his Win Shares book to arguing that win shares are not based on value above a replacement level. Since I believe he introduced the concept of replacement level to sabermetrics, I have to believe he knew what he was talking about. Therefore, I think it's misleading to keep saying that WS uses a replacement level of 50%.
   196. Paul Wendt Posted: October 13, 2007 at 04:09 AM (#2574113)
158. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 08, 2007 at 05:52 PM (#2567797)
Paul Wendt--And *why* should he be docked a win for the rest of the season?

That is a numerical illustration, of course.
I tried to help in #157, commenting on #123-124.

Replacement level is replacement level--it's freely available! By definition!

That isn't merely a matter of definition.
It may be understood as mere definition; if so, that is a misunderstanding.

Call it an axiom of your system, which implicitly defines unequivocal replacement level, without any scope for time horizon, or various short and long runs. Your axioms define replacement level implicitly as the axioms of Euclid, Peano, and Zermelo-Frankel define lines, natural numbers, and set membership.

There is a position in economic theory that "cost is cost" without any scope for time horizon. I don't know how widely it now dominates professional economics. Even if it does dominate, and one defers to professional authority or "the market for ideas" without question, replacement level in baseball is on shaky ground. Because it is not supported by research grants or academic salaries, baseball analysis (by theory or by statistical inference) is relatively expensive and the number of practitioners who merit expert status is tiny. Peer review is informal or nonexistent.

--
DanR #190
It seems to me that the burden of proof should fall pretty strongly on Brent and Dizzypaco to show--with real data--that the consensus placement of replacement level at 80% of positional average is incorrect. The fact that both Keith Woolner's study of bench players and Nate Silver's study of minimum-salary veterans got roughly the same result is what makes me consider replacement level "settled law," to quote John Roberts, and not an open question.

Rather than be so audacious, say merely that
you believe Keith Woolner and Nate Silver have dismantled the concept of time horizon in baseball analysis, and that is an axiom of your own player rating system?
   197. Paul Wendt Posted: October 13, 2007 at 04:25 AM (#2574129)
I tried to help in #157, commenting on #123-124.

I messed up the quotation format in #157.

== quote & reformat ==

123. KJOK Posted: October 05, 2007 at 03:11 PM (#2561895)
[quoting TomH #105]
> Some have also studied & concluded that
> over a very long period of time, replacement level rises a bit.

I hope this is understood to be a fact, and not just 'a bit' but really 'tends towards .500' (or theoretically .517 or so) as X goes towards infinity with X being the TIME variable.

In other words, if a player gets hurt mid-game, the replacement level is very low - you grab someone off the bench who's on the current roster. If a player announces his retirement in April 2007 effective at the end of the year, his replacement level in 2008 will be MUCH higher - you trade for a replacement, or promote a hot prospect, etc.


124. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 05, 2007 at 03:16 PM (#2561902)
OK, but trading for a replacement has a cost--the player you give up--and you can only promote a hot prospect if you have one; they're not freely available. The argument that "offseason" replacement level is higher than 80% of positional average makes no sense to me. Sure, you can acquire an above-replacement player, but only for an above-replacement price!

[Paul Wendt #157]
I believe there is a stochastic component of the argument which you have a missed. KJKS announces his retirement in April effective immediately, so he is replaced by freely available talent, say 80% of league average for sake of argument. If KJMO announces that this is his last season, so that replacement is needed one year hence, you have a chance to work with and observe three replacement players for a year, this season and next winter. After that year, one appears to be a 70, one an 80, and one a 90. So our retiring player gets replaced by a 90, someone no longer freely available because he has been selected as the best of three players (in this simple example). An enthusiastic proponent of this argument might use 60, 80, and 100 for numerical example.

== end quote/reformat ==


P.S. I don't know
whether TomH refers to anyone in particular ("Some have studied");
whether .517 is someone's estimate or KJOK's off-the-cuff illustration;
whether KJOK or anyone else in particular believes in short and long runs based on this "stochastic component".

There are other things I don't know but I am too tired to keep listing them.
   198. Paul Wendt Posted: October 13, 2007 at 04:31 AM (#2574141)
I downloaded #190 and returned to it after the end of the ballgame.
Looks like Brent did the same but he was a little quicker.

Don't let me shock you severely at this late hour so stop reading before I say that I agree with much of #193-195 but especially with the first paragraph.
   199. Paul Wendt Posted: October 13, 2007 at 02:07 PM (#2574471)
There are other things I don't know but I am too tired to keep listing them.
;-)

This space is too small to contain everything I don't know, so let me go.
;-)


> Replacement level is replacement level--it's freely available! By definition!

That isn't merely a matter of definition.
It may be understood as mere definition; if so, that is a misunderstanding.

Call it an axiom of your system, which implicitly defines unequivocal replacement level, without any scope for time horizon, or various short and long runs. Your axioms define replacement level implicitly as the axioms of Euclid, Peano, and Zermelo-Frankel define lines, natural numbers, and set membership.


This isn't a mathematical science like geometry, arithmetic, and set theory. I don't suppose that much axiomatization of its theory would be a fruitful goal, as it was in mathematics (unattainable yet fruitful). Indeed, its formal engagement with data is relatively mature. Any project to make its theory rigorous by selecting a useful set of axioms would require a highly artful engagement with a cornucopia of statistical descriptions and findings.
   200. karlmagnus Posted: October 13, 2007 at 02:56 PM (#2574498)
Bump
Page 2 of 3 pages  < 1 2 3 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
The Ghost is getting a Woody
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 1.5558 seconds
49 querie(s) executed