77 Responses to “100+ Career “Runs” From Fielding”

1. Steve Lombardi Says:
   June 17th, 2011 at 9:35 am

   FYI, Andy looked at the other end of this group earlier this year:

   http://www.baseball-reference.com/blog/archives/10207

2. Scott Says:
   June 17th, 2011 at 9:48 am

   Andruw Jones will almost certainly be the biggest Hall of Fame snub of all time, or at least top 3.

3. kds Says:
   June 17th, 2011 at 10:06 am

   This list is only 1950 and later?

4. Chuck Says:
   June 17th, 2011 at 10:29 am

   "Andruw Jones will almost certainly be the biggest Hall of Fame snub of all time, or at least top 3."

   Not even close.

5. Genis Says:
   June 17th, 2011 at 10:36 am

   The only players who were primarily first basemen on this list:

   Keith Hernandez
   Albert Pujols

   And I love any list that includes Craig Counsell!

6. Larry R. Says:
   June 17th, 2011 at 10:37 am

   @2,4

   Chuck's right. I don't think Jones or Bell will get close enough to even sniff the HOF.

7. Steve Lombardi Says:
   June 17th, 2011 at 10:46 am

   Kds - yes, I do not believe the earlier data is out there.

8. Lawrence Azrin Says:
   June 17th, 2011 at 10:53 am

   @2/ Scott Says: "... Andruw Jones will almost certainly be the biggest Hall of Fame snub of all time, or at least top 3."

   Scott, I think that Jones needs to get in line behind Barry Larkin, Ron Santo, Jeff Bagwell, Tim Raines, Bobby Grich, Bill Dahlen, (and several others at least), for a HOF-snubbing argument.

   A lot of Jones' case is how much value you give to his defense - with 239 Rfield, about 40% of his total WAR is from defense, a higher % than Ozzie Smith. If you trust that, he's got a decent but hardly overwhelming case. All the players I mentioned have more career WAR than Jones, although at 59.2, Jones is certainly in the discussion.

   His reputation was hurt a lot by his decline at age 30, then his Wiley Coyote-type fall off the cliff at age 31. He may not play a full season as a regular again, but if he sticks around for 5/6 more years, he could build up some impressive counting stats (about 500 HRs, 1500 RBI).

9. Devon Young Says:
   June 17th, 2011 at 11:12 am

   So, you're telling us that not only were Robinson & Belanger two of the best infielders ever, but that they have a legit argument for being the two best infielders ever? And they played on the same team for about 10 years?!? WOAH! Not to miss... Paul Blair is also on this list?!?

   How did Jim Palmer not win 300 with that defense eating up half the infield and a piece of the outfield?!?

10. Johnny Twisto Says:
    June 17th, 2011 at 11:19 am

    There is earlier data, which you can see on each player's WAR table. Because there is no play-by-play data yet for years prior to 1950, those numbers are calculated by a different form of Total Zone and necessarily have even bigger error bars. For whatever reasons, B-R doesn't make those numbers searchable in the PI. But you can look them up. A few notables:

    Joe Tinker +180
    Rabbit Maranville +130
    Johnny Evers +127
    Jimmy Collins +121
    Tris Speaker +92

11. Austyn Says:
    June 17th, 2011 at 11:24 am

    Two days in a row with a post that includes the great Gary Gaetti.

12. Johnny Twisto Says:
    June 17th, 2011 at 11:27 am

    Incidentally, for people who are skeptical of "advanced" defensive stats (and skepticism is warranted), does Steve's list not look like a pretty good collection of the best defenders of the past 60 years? Of course you could quibble about X being higher than Y, or why is A on this list and where's B? But if 10 serious fans were to sit down and list who they thought the best defensive players were, they'd come up with a list not unlike this one. So I think we can do away with the comments about the numbers being useless. Can they be improved? Yes. Should they be complemented by contemporary, subjective opinion? Yes. But the above seems like a great starting point, at the very least. If there is a player about whose defensive you know nothing, TZ/Rfield is not a bad place to start. Sure, it might be wrong (and is, for some players), but it's more likely right.

13. Johnny Twisto Says:
    June 17th, 2011 at 11:29 am

    Devon/9, and Bobby Grich (+73 for his BAL years, +82 overall) was also there for a spell.

14. John Q Says:
    June 17th, 2011 at 11:31 am

    @2 Scott,

    Jones looked like he was well on his way to the HOF before his career came to a screeching and dramatic halt around 2006. The way it looks now, Rolen, Edmonds, and Lofton will be bigger HOF snubs than Jones in 5-7 years. I don't think Jones will even break the top 10-15 among HOF position player snubs when he's eligible.

    At the moment leaving off Mcgwire and Palmeiro, these would be my top 10 HOF Snubs among eligible position players.

    1-J. Bagwell
    2-B. Dahlen
    3-R. Santo
    4-B. Grich
    5-A. Trammell
    6-B. Larkin
    7-E. Martinez
    8-L. Walker
    9-L. Whitaker
    10-D. Allen

    And among eligible pitchers:

    1-K. Brown
    2-R. Reuschel
    3-D. Cone
    4-L. Tiant
    5-J. Koosman
    6-D. Steib
    7-B. Saberhagen
    8-B. Pierce
    9-T. John
    10-L. Jackson

15. Lawrence Azrin Says:
    June 17th, 2011 at 11:37 am

    @9/ Devon Young Says: "So, you're telling us that not only were Robinson & Belanger two of the best infielders ever, but that they have a legit argument for being the two best infielders ever? And they played on the same team for about 10 years?!?... .. How did Jim Palmer not win 300 with that defense eating up half the infield and a piece of the outfield?!?"

    Devon, he missed most of 1967 and all of 1968 with assorted arm troubles, surgery, and rehabbing in the minors. Fill in the missing years with 16/18 wins (he won 16 every year from 1969 to 1978, except for one year), and he's got his 300 wins (268 actual total).

16. Detroit Michael Says:
    June 17th, 2011 at 11:38 am

    Yo, Adrian! You made this post too.

    Barry Larkin isn't a Hall of Fame snub. He got 62% of the BBWAA vote in his second year of eligibility. He's going in.

17. Detroit Michael Says:
    June 17th, 2011 at 11:40 am

    Also, check out Jim Palmer's career K/BB ratio. He was relying on his superlative defense (plus probably some skill at producing weak contact) to a great extent.

18. Anon Says:
    June 17th, 2011 at 11:41 am

    @9, Devon - the unbelievable defense behind Palmer is why he has a career .251 BABIP. Even for that era (when BABIPs were lower than now), that is an incredible number

19. Dvd Avins Says:
    June 17th, 2011 at 11:45 am

    In case anyoen's unaware, Rfield estimates how many runs a player was batter than the average player at his position. If you believe (as I think anyone serious about the question must) that an average shortstop is providing more value with his glove than an average left fielder does, then it follows that this list, while interesting, is not a list of the players whose fielding was most valuable.

20. John Q Says:
    June 17th, 2011 at 11:46 am

    @ Devin,

    Not only were Robinson and Belanger the two best infielders of all time and on the same team, they were also on the SAME side of the infield. Those two guys made lefty pitchers Mike Cueller and Dave Mcnally stars.

21. Chuck Says:
    June 17th, 2011 at 11:53 am

    John @#14,.

    Very good list.

    Not a Hall of Fame list, however.

22. Liam Says:
    June 17th, 2011 at 11:55 am

    Andruw Jones should be in the Hall, he was both a superstar and a compiler. Here is a man who in his prime 2000 -2006 he hit 262 home runs and was by far the best centerfielder in the game. There is no one playing now who even comes close to his power at his position. Also at the age of 34, he still has a chance to hang around, get another starting job and hit 500 home runs. His similarity scores are a joke #1 Dale Murphy had nowhere near the defense Jones had #2 Jimmmy Edmonds was nowhere near the hitter. He is closer to Ken Griffey Jr. than anybody else. He was a phenom as a kid and then really lost a significant step as he got older. But still HOF

23. Johnny Twisto Says:
    June 17th, 2011 at 12:04 pm

    Huh? Think you better look up Edmonds again. Much better hitter than Jones. And if Jones had offers to be a starter this season, why would be settle for a 4th OF job? If he didn't have the offers this season, when is he going to get them?

24. TapDancingTeddy Says:
    June 17th, 2011 at 12:13 pm

    @1 thanks for the link to the negative side. I didn't see that post and it is fascinating to see that Derek Jeter and Bernie Williams were both more negative in the field than Manny Ramirez. I would've never expected that conclusion. What a terrible thing to have to admit if you're a Yankee fan as I am.

    In regard to this post: I saw Blair, Belanger and Brooks Robinson play, and this is one place that my memory matches the numbers. I always thought that they were some of the best defenders ever, and to this day, I get blank stares when I try to tell people that Mark Belanger was as good or better than Ozzie Smith.

25. Lawrence Azrin Says:
    June 17th, 2011 at 12:25 pm

    @21/ Chuck: I could make a good HOF argument for most of the position players that John Q listed in #14, but not so much for the pitchers (maybe Tiant and John).

    @19/ Dvd Avins Says: "In case anyone's unaware, Rfield estimates how many runs a player was batter than the average player at his position... ...this list, while interesting, is not a list of the players whose fielding was most valuable.

    Ok, then how can we tell from the WAR data who has the most _total_ defensive value?

    I'm surprised that no one has mentioned the two "DE's" from the 1970s/80s: Dwight and Darrell Evans.

26. Rob Says:
    June 17th, 2011 at 12:30 pm

    @25

    I was just about to mention Dwight Evans. If someone like, say, Gary Carter is in the HOF, why isn't Dewey?

27. Chuck Says:
    June 17th, 2011 at 12:31 pm

    Lawrence,

    Yes, I agree, the list of position players is significantly better than the pitchers.

28. Bastaducci Says:
    June 17th, 2011 at 12:35 pm

    @ at post 19

    I could not agree more.

    Any leftfielder or right fielder is a guy who cannot play centerfield on a regular basis.

    Any 2nd baseman or 3rd basemen is a guy who cannot play shortstop on a regular basis.

    Any 1st basemen is a guy who is not a good enough fielder( I include arm strength in the fielding category) to even play right field.

    There is exceptions like A-ROD moving to 3rd because the Yanks had Jeter though.

    To me it is a travesty guys like Trammell,Larkin,Bernie Williams don't get more love.

29. Bastaducci Says:
    June 17th, 2011 at 12:37 pm

    @ post 26

    Because Carter was the superior player. he put up Dwight Evans like stats but did it at the much , much , much tougher postion. I loved Evans but to me he is not even in the same category as Carter.

30. John Autin Says:
    June 17th, 2011 at 12:51 pm

    JT @12: "does Steve's list not look like a pretty good collection of the best defenders of the past 60 years?"

    Johnny, I think I get your point. But from another perspective, couldn't someone say the same thing about a list of repeat Gold Glove winners? Yet we generally don't give a lot of credence to the Gold Gloves. (Correct me if I'm wrong.)

    If one of the premises of the whole enterprise of advanced defensive metrics is that reputation does not always match measurable reality [cough - Captain Clutch - cough], then isn't it a circular argument to say that the accuracy of the metrics is bolstered by the fact that they confirm the reputations?

31. Johnny Twisto Says:
    June 17th, 2011 at 12:57 pm

    how can we tell from the WAR data who has the most _total_ defensive value?

    Add Rfield (defensive performance) and Rpos (positional value). That will be the total runs of value the player added over a completely average defensive player in the middle of the defensive spectrum. The SS get bumped up, the corner OF move way down.

    The (possibly incomplete) top of the list (since 1950):

    O. Smith 386
    B. Robinson 362
    M. Belanger 354
    C. Ripken 324
    L. Aparicio 313
    I. Rodriguez 306

32. Hartvig Says:
    June 17th, 2011 at 12:59 pm

    Apparently I'm the only person who is amazed to see Sammy Sosa on this list. Got to be some sort of glitch in the formula there.

    And speaking of the Baltimore Orioles, check out their 1972 season when they allowed only 430 runs the entire season. Not only did they have Robinson, Belanger and Blair but also a young Bobby Grich.

    And while a lot of the formulas for measuring defensive value leave something to be desired I'm also reminded of Whitey Herzog commenting that it really didn't matter what Ozzie Smith hit because he saved the team 100 runs a year with his defense.

33. Lawrence Azrin Says:
    June 17th, 2011 at 1:08 pm

    @29/ Bastaducci Says: "@ post 26 - Because Carter was the superior player. he put up Dwight Evans like stats but did it at the much , much, much tougher postion. I loved Evans but to me he is not even in the same category as Carter"

    Bastaducci, I loved Dewey, but I've got to agree; he didn't become an outstanding hitter till 1981, nearly halfway through his career. Before that (1972-80), he was mostly a decent player with more potential than accomplishment.

    However - when they were both active, most everyone ranked Jim Rice ahead of Dewey. Nowadays, many people would reverse that ranking (not that this is enough to make him HOF-worthy).

34. John Autin Says:
    June 17th, 2011 at 1:12 pm

    On the Jim Palmer side thread (@9, etc.) --

    The bottom line on why Palmer didn't win 300 is simply the innings total. "Cakes" finished with 3,948 IP. The only 300-game winner with less than 4,135 IP was the amazing Lefty Grove (3,940). Only one other modern pitcher had as many wins as Palmer in fewer innings (Mike Mussina).

    It's also worth noting that, while Belanger and Brooksie were historically great defenders, Belanger was a very weak hitter, and Brooks was below average for much of Palmer's career.

    Palmer's first full year was 1966. Belanger hardly played then; the SS then was 32-year-old Luis Aparacio, starting to lose his range. B-Rob did haev a good offensive year in '66 (123 OPS+).

    Palmer didn't top 100 IP again until 1969. His overlap with Belanger as the regular SS ran from 1969 through about 1980. During that period, Belanger had a combined OPS+ of 71.

    Palmer's overlap with Brooksie as the regular 3B, after '66, ran from 1969-76; Brooks had a combined 95 OPS+ in that span.

    I don't doubt that the total contributions from Belanger and Robinson were still a big net positive to Palmer. But then, Palmer did win over half his career starts, which is darn good.

35. Eric Says:
    June 17th, 2011 at 1:16 pm

    Clicked through to Mark Belager's player page and noticed that his nickname, "The Blade," is not listed.

36. Johnny Twisto Says:
    June 17th, 2011 at 1:19 pm

    JA/30, you have a point. I was thinking about the circular argument too. But if the stat has any validity, it has to have some correlation to the "truth," even if we can't be sure what is true and there is no one way to determine it. The old Bill James line was that a statistic which never surprises you adds no value, and a statistic which always surprises you is probably wrong. Certainly, only relying on our eyes and on biased defensive stats can lead to errors of judgement. A player with the penchant for the spectacular, but who is poorly positioned and gets bad jumps may look more impressive than he is. Still, if people had no idea what they were watching, as many people would think Howard Johnson was a great shortstop as did Ozzie Smith. I think that there is consensus about Smith indicates we can "know" things even in the absence of hard data.

    And I'm glad you brought up the Gold Gloves. While there is no question they have their problems -- a small, somewhat uninterested electorate, inertia in the voting, occasional complete WTF results -- *overall* they have meaning. With the notable exception of Captain Clutch, I think if you look at the list of guys who have won at least 4 times, they were all subjectively considered excellent fielders. I have checked the results of GG winners against Total Zone, UZR, and DRS, and at every position the average winner measures as a well above-average fielder. I would actually say the same thing about them as I did about the above list. If you have no information about a player except that he won multiple GGs, it's good bet that he was a good defender.

    I feel like this response is unsatisfactory. But I'll have to come back to it later.

37. John Autin Says:
    June 17th, 2011 at 1:20 pm

    @32, Hartvig, re: Sosa --

    Virtually all of his defensive value was amassed from age 22-28 -- 111 "runs", 10.9 dWAR). From age 29 on, those totals were just slightly negative, i.e., average.

    The irony, of course, is that the vast bulk of his offensive value was amassed from age 29 onward. His combined OPS+ was just 106 through age 28, but 145 thereafter.

    Which gets me thinking ... It's probably impossible to study, but it would be fascinating to know the effects of steroid use on defensive play.

38. Chuck Says:
    June 17th, 2011 at 1:21 pm

    "Apparently I'm the only person who is amazed to see Sammy Sosa on this list. Got to be some sort of glitch in the formula there."

    No, you're not, and Sosa's not the only one, either.

39. John Autin Says:
    June 17th, 2011 at 1:34 pm

    JT @36 -- I found your response quite satisfactory, actually. Thanks!

40. Lawrence Azrin Says:
    June 17th, 2011 at 1:38 pm

    @31/ Johnny Twisto Says: "how can we tell from the WAR data who has the most _total_ defensive value?
    Add Rfield (defensive performance) and Rpos (positional value). That will be the total runs of value the player added over a completely average defensive player in the middle of the defensive spectrum. The SS get bumped up, the corner OF move way down."

    Thanks, Johnny T. I did this for a few notoriously bad defensively players of different eras, and got:

    Babe Herman -95
    Dick Stuart -128
    Frank Howard -209
    Greg Luzinski -203
    Manny Ramirez - 223

    OTOH, Derek Jeter is respectable at +15

41. Jeff Says:
    June 17th, 2011 at 3:28 pm

    And people still want to exclude the great Barry Bonds with a defensive value at 191(6th all time)...Unbelievable.

    Best all around player ever and that proves it!

42. Johnny Twisto Says:
    June 17th, 2011 at 4:10 pm

    Re worst defensive players, I get Frank Thomas at -259. Is there anyone lower? (And I'm of the opinion that the positional adjustment for DHs is too small) Thomas is the perfect storm for this -- a bad defensive player, becoming a DH, over a long career. He beats Edgar Martinez, Hal McRae, Harold Baines, David Ortiz. Other good candidates?

    Stuart might have the least defensive value on a per PA basis.

43. Lawrence Azrin Says:
    June 17th, 2011 at 4:51 pm

    @42/ Johnny Twisto Says: " Re worst defensive players, I get Frank Thomas at {-259}. Is there anyone lower?..."

    He doesn't quite surpass Thomas, and I didn't do a specific P-I search, but I have Dave Winfield at {-247}. A huge problem with this is that I don't believe Winfield is a worse fielder than Manny Ramirez or Greg Luzinski (see my #40), or even remotely as bad as those two.

    Explanations are welcomed.

    @31/ if I use the same method as Johnny Twisto did in #31, of {Rfield + Rpos} I get:

    Barry Bonds: +63
    Jim Edmonds: +102
    Kenny Lofton: +138
    Andruw Jones: +251

44. Lawrence Azrin Says:
    June 17th, 2011 at 4:54 pm

    @43/ OOPS, I was referring to post #41 at the bottom, not #31

45. RobM Says:
    June 17th, 2011 at 4:55 pm

    Robinson, Bell, Nettles, Schmidt...some mighty good defensive third basemen back in the 70s. The list doesn't even include Aurelio Rodriquez, who was a good glove man too, but not much with the bat.

46. Dan G Says:
    June 17th, 2011 at 5:07 pm

    I love the post. Can someone tell me how Richie Ashburn fares here?

47. Baseball: 100+ Career “Runs” From Fielding » Stathead » Blog Archive Says:
    June 17th, 2011 at 5:22 pm

    [...] 100+ Career “Runs” From Fielding: At the B-R Blog, Steve posted the leaders in TotalZone fielding runs above average, including everyone with at least 100 runs. [...]

48. RobM Says:
    June 17th, 2011 at 5:23 pm

    Boggs and Sosa are the two most surprising name to me on that list. Ripken being so high is also a surprise. Names like Robinson and Belanger at the top are certainly not a surprise.

    I need to understand RField a bit more. Is there a definition of how its calculated on the site? Is it cumulative, so someone who plays a very long time and is slightly plus will accumulate a higher score than someone who was a kick-ass great defender buy played for a shorter period of time. I might even be familiar with it already under another name.

49. Johnny Twisto Says:
    June 17th, 2011 at 5:47 pm

    Lawrence/43, I agree, I think the formula "misses" on Winfield for some reason. It gives me reason to think he wasn't as good as he was generally perceived, but I don't buy him being an all-time disaster. (He was my favorite player, so I am defensive about it too.)

    ***

    Can someone tell me how Richie Ashburn fares here?

    You can look up any player on his page. Ashburn rates 76 runs above average, including some terrible ratings at the end of his career which drag him down. And a positional adjustment of -19, giving him a career defensive value of +57.
    http://www.baseball-reference.com/players/a/ashburi01.shtml#batting_value::none

    ***

    I need to understand RField a bit more. Is there a definition of how its calculated on the site? Is it cumulative, so someone who plays a very long time and is slightly plus will accumulate a higher score than someone who was a kick-ass great defender buy played for a shorter period of time. I might even be familiar with it already under another name.

    Yes, it is cumulative (but can go up or down). It is originally known as "Total Zone." When it is incorporated into WAR, it gets called "Rfield" (fielding runs), I guess for consistency of nomenclature among all the components of WAR.

    A description is here: http://www.baseball-reference.com/about/total_zone.shtml . I don't think a comprehensive definition of the system is available in one place. You can probably find other pieces of it described around the Internet, in posts by "AROM" or "Rally" (Sean Smith) at http://apps.baseballprojection.com/blog/ or http://insidethebook.com/ee/ or http://www.baseballthinkfactory.org/ or http://thehardballtimes.com/

50. mccombe35 Says:
    June 17th, 2011 at 9:37 pm

    Ventura is underrated!

    & as for the WORST defensive player, the -15 or so a year for being a DH just doesn't make sense. You aren't hurting the team defense if you aren't out there.

51. Shping Says:
    June 17th, 2011 at 9:42 pm

    A lot of good stuff here, everyone, from one end to the other: #2: sorry, but ha ha ha ha. #48: i agree: Boggs?????

    #14 is a great list, but way too many players have not even been eligible yet or barely eligible and too soon to tell.

    We need to remember a new but not new name for the list: Gil Hodges is definitely one of the great HOF snubbs of all time. Very good hitter, great fielder, good postseason stats overall, longevity numbers, good manager...., including these stats:

    Assists as 1B
    1949 NL 80 (2nd)
    1950 NL 100 (2nd)
    1951 NL 126 (1st)
    1952 NL 116 (1st)
    1953 NL 99 (4th)
    1954 NL 132 (1st)
    1955 NL 105 (2nd)
    1956 NL 103 (2nd)
    1957 NL 115 (2nd)
    1958 NL 69 (5th)
    Career 1,281 (23rd)

52. Shping Says:
    June 17th, 2011 at 10:06 pm

    ...Of course, as Bill James once pointed out when comparing Garvey and Buckner, maybe those high assist totals are simply because Hodges preferred to flip the ball to the pitcher covering first, rather than taking the putout on his own. Nontheless, i think there's a lot of evidence showing he belongs in the hall.

    Has anyone ever run a report on best career WAR or other stats for snubbed HOFers?

53. Dan G Says:
    June 17th, 2011 at 10:24 pm

    I am asking again could someone tell me how Richie Ashburn rates here? It seems hard to believe that he did not make the list.

54. Johnny Twisto Says:
    June 17th, 2011 at 10:52 pm

    the -15 or so a year for being a DH just doesn't make sense. You aren't hurting the team defense if you aren't out there.

    It's not that you're hurting it. You're just not helping it. You are making no defensive contribution whatsoever. So compared to a theoretical completely average defensive player, the DH is deemed to be worth 15 fewer runs to his team.

    It's a tricky concept, because there is no ground floor for defense. A batter can bat .000 at the absolute worst, and we measure his contributions up from there. But there is no real parallel for fielding. Do you compare the contribution of a fielder to having no one out there at all? That's not realistic. So imagine the DH as having zero defensive value, and anyone who plays the field with some degree of competency rates higher, and the overall average is pegged at 15 runs higher than a DH.

    Or think of it this way: If you did not make the positional adjustment for DHs, then you are placing the defensive value of a DH as equal to an average 3Bman. The way WAR works, you have to evaluate DHs against a higher baseline, because it's a position anyone can play. By definition it requires no defensive ability. The number of players who can catch while also hitting enough to play in the majors is very small.

55. Chuck Says:
    June 17th, 2011 at 11:50 pm

    Dan,

    "I am asking again could someone tell me how Richie Ashburn rates here?"

    For his career, Ashburn was plus 76 for his career, although it must be noted he was minus 35 over his last four years.

    Otherwise, he would have made this list comfortably.

56. John Q Says:
    June 18th, 2011 at 12:12 am

    @52 Dan G,

    You can check any players "Rfield" by going to his player page and looking in the section labeled "Player Value".

    Ashburn has 76 career fielding runs or "Rfield".

    He would have made the list had he retired after the 1958 season because he had 113 fielding runs after his age 31 season. But it seems like his range and mobility started to affect him drastically during his age 32 season and he lost 37 fielding runs during his last 3 of 4 seasons.

    This downward trend toward negative fielding runs is somewhat common among good fielding or star center fielders as they age into their 30's. The better fielding CF tend to keep playing CF into their 30's when realistically they should be playing either LF or RF. As a result you see quite a few negative fielding run seasons for these players later in their careers.

    Ashburn's downward slide was drastic and sharp. Usually CF's skills deteriorate more slowly.

    The same thing happened to Griffey jr. when he reached his 30's in Cincinnati. Griffey should have switched to RF when he was 31 in 2001. There's definitely no way he should have still been playing CF in 2005-2006. But he was a big star so he was basically able to do what he wanted.

57. Evan Says:
    June 18th, 2011 at 1:13 am

    JT @ 41, 54

    I would by no means consider myself an expert on the subject, but the way you are using the positional adjustment is inapt as I understand the way it is computed as a component of WAR.

    My understanding is that it is an adjustments to account for the relative dearth or plethora of offense available at any given position, which is why it is a component of oWAR and not dWAR, even though it is based upon the player's defensive position. I don't believe it is in any manner intended to account for the differing contributions of various defensive positions to run prevention (which should vary depending upon handedness, velocity and GB/FB tendencies of the pitcher). Thus simply adding the number to the rField gives a measure of something, but I don't think there is any reason to believe that it is a measure of the player's total defensive contributions.

58. Evan Says:
    June 18th, 2011 at 1:14 am

    Pardon me, the previous should read JT @ 31, 54.

59. Johnny Twisto Says:
    June 18th, 2011 at 1:36 am

    Evan, it is an adjustment because of the different offense available at each position, but the reason why offense is different is because of the difficulty of playing the position. The harder the position is to play, the fewer players capable of playing it. While Derek Jeter is a whipping-boy for his defense, the fact that he *is* capable of playing SS, even if poorly, provides value to his team. Defensive value.

    You are correct that it's not directly an account of a position's contribution to run prevention. But implicitly I think it does that, as the adjustments are based on the relative defensive performance of players who appear at multiple positions

    As for why Rpos is included in oWAR, that is because people were complaining that the Rfield ratings were wrong. So Sean decided to break them off. If you want to plug in a guy's UZR instead, or your own mental rating, you can easily do that. I understand why he split it that way, though I disagree with the decision.

60. Johnny Twisto Says:
    June 18th, 2011 at 1:40 am

    Though, thinking about this more, I guess one could argue that Jeter's ability to play SS merely makes the value of his *offense* greater. Hmm. It may be somewhat semantics. I think merely being able to take the field and provide a minimum threshold of defense is defensive value. But I see your point. I think.

    And I guess one could argue that on a particular team, if a particular position saw very few chances, it shouldn't have the same implicit positional value as on an average team. Maybe that's accounted for because the reduced chances limits the player's ability to score high in Rfield? I need to ponder this.

61. Dvd Avins Says:
    June 18th, 2011 at 9:32 am

    Adding Rfield and Rpos is a big step in the right direction, but you still have the replacement-level problem. Someone who cam play 3B or CF (roughly average defensive positions) at an MLB-average level is in the majors in part because he can field. In other words, he is providing positive value with hiss fielding.

    By happy coincidence, someone who can't field and who is an average MLB hitter is just about on the borderline of who is good enough to play as a DH. Which means the negative Rpos from the bad side of the defensive spectrum should just about balance Rrep so that a first baseman or DH with zero Rbat should end up with somewhere around zero RAR.

    That means that the actual full defensive value can be approximated by adding Rrep as well as Rfield and Rpos, while the actual offensive value does not include Rrep. (Or at least not most of it. A more serious look at how bad someone who only hits has to be to lose his job might determine that some fraction, or even some small negative fraction, of Rrep belongs with offense.)

    As for the oWAR and dWAR split, I completely understand and even agree with why Sean wanted to display those numbers, but I strongly disagree with the labels he gave the numbers. That mislabeling has encouraged more muddled thinking about sabermetrics than anything I've seen since Palmer's implicit (but not entirely intended) use of MLB average as a replacement-level baseline.

62. Dvd Avins Says:
    June 18th, 2011 at 10:10 am

    @57 That dearth you're talking about isn't really a dearth, but a prioritization. If a teacm thought it was profitable, it could put Jim Thome at shortstop. The shortage if of people who can BOTH hit and field. The Rpos is computed as you say, but you have to look at the market-like forces that make for the apparent dearths of players teams WANT to play at more demanding defensive positions. When you consider that, you see that the difference in Rpos has to correspond* with differences in the value of being an average defensive performer at the various positions.

    * The correspondence is only perfect if management's assessment of who should play where is perfect, but we can assume the correspondence is good enough to use for these purposes.

63. BSK Says:
    June 18th, 2011 at 10:42 am

    JT-

    How dare you besmirch the name of HoJo!

64. Johnny Twisto Says:
    June 18th, 2011 at 12:19 pm

    I guess one could argue that Jeter's ability to play SS merely makes the value of his *offense* greater.

    No, I'm retreating from this position. Jeter doesn't bat as a SS, he bats as a batter. If he creates a run offensively, it isn't worth any more than a run created by Mark Teixeira. Their respective offensive values have nothing to do with what position they play. The value of Jeter as a *player*, however, is increased because he is playing SS. That positional value may not directly translate into runs prevented, because there is no real "zero level" for defense (see #54). But it makes more sense to consider it as defensive value than offensive value.

65. Dvd Avins Says:
    June 18th, 2011 at 4:00 pm

    Hmm. Setting separate defensive and offensive replacement values is more complex than I'd thought. If you can't field a lick, you can stick around if you're an above average hitter. If you can't hit a lick (not even at what we consider replacement level) you're not going to play any position other than pitcher, no matter how well you field.

    There just aren't enough balls hit to any one position to justify putting, say, Fred Astaire at shortstop, even if he were as great a fielder as he was a dancer, if he was going to strike out every time he came up. And if a team did play him, his Rbat would be so bad that he would have negative RAR and WAR overall.

    So if you consider Mark Belanger, you have to say some of his value comes from hitting better than Fred Astaire, not just from his fielding.

    The problem is that we're introducing a comparison to replacement-level too early in the process. Fielding value is (as Johnny Twisto said) Rpos + Rfield. That's not value above replacement value, it's value above nothing. Offensive value is how much better the Rbat (and the other minor offensive components) is compared to a player who strikes out every time. Add those two together and you got total value over nothing.

    A replacement-level player provides substantial value over nothing. You have to subtract that (which is about the difference between league average hitting and Fred Astaire hitting) to get WAR. It's just instead of reporting Rrep as a single number to be added because it's in comparison to league average, you'd report two numbers, a larger one to be added that all strikeouts to the player in question, and another number to be subtracted that compares no value to a .325 winning percentage quality (which is the replacement level in use at bb-ref.)

    What portion of the subtracted number could be charged to a player's offense and what portion to the defense is then the question. I think the most sensible answer is to keep the proportions of the players contribution above zero.

66. Dvd Avins Says:
    June 18th, 2011 at 4:30 pm

    Running some stuff from the Play Index, I see that it takes just about 4 outs to equal -1 Rbat.

    So let's say Danny DeVito would play MLB baseball at a level indistinguishable from a corpse, except a little faster and before we get to RAR and WAR, we'll pass through RAD (AD is for "above DeVito") and WAD. Rdef is Rfield + Rpos. And Roff is Rbat + Rbaser + Roe + Rdp + PA/4. RAD is Rdef + Roff. A players true OffWAR would be his (Roff / RAD) * WAR. The true DefWAR would be (Rdef / RAD) * WAR.

    To actually get up to a replacement-level player, it seems (from looking at players like Willie Montanez) that you need to produce 2/9 of a RAD per plate appearance.

67. Evan Says:
    June 18th, 2011 at 4:46 pm

    I had understood the following:

    Rbat = relative contribution of the player as a batter to the average player

    Rfield = relative contribution of the player as a fielder to the average fielder at his position (pro-rated)

    Rrep = relative contribution of an average player to a replacement-level player (I had always thought this was offense, but now I'm thinking it might include defense even if this is a minimal contribution) - this seems to be a constant for the league, pro-rated for playing time

    Rpos = adjustment for position to account for the different expectations we have for offense from the various positions based upon our observations that certain positions produce more offense than average and others less therefore it is illogical to compare positions which produce above or below average offense to league average offense as a whole
    ---

    I had been thinking that Rrep was only offense and that the essential assumption that replacement level defense was roughly equivalent to average offense. I was comfortable with this idea because of my assumption that assuming basic competence at a position, offense was the overwhelming deciding factor in determining the MLB/AAA dividing line. Now I am thinking that any difference in fielding ability is wrapped into Rrep.

    I'm having trouble with the construction of "defensive value over nothing." It can't be comparing to playing defense without that player on the field (1B would be far more important because without him you would be reliant on the P or 2B getting to first to turn any groundball into an out). Against certain lefthanded hitters hitters many teams have decided that 3B (and in some cases SS) is less valuable than short RF.
    ---

    I'm not sure how one would go about determining the relative value that each defensive position provides. I'm fairly certain that the physical attributes needed to play certain positions competently are less conducive to making offensive contributions than those needed for less demanding positions. So there is a certain intertwining of offensive and defensive contributions.

68. Johnny Twisto Says:
    June 18th, 2011 at 5:42 pm

    Now I am thinking that any difference in fielding ability is wrapped into Rrep.

    Yes.

    Rrep is how much better an average player is than a replacement player. It's a playing time stat, and works out so that a replacement player is defined to be about 2 wins worse than average over a full season. And those two wins could come from anywhere -- hitting, fielding, whatever. A lot of (most?) actual replacement-type players probably fit the decent glove, marginal bat definition. But I've mentioned Shelley Duncan before, who I think is a replacement type with a decent bat but little defensive value.

    So DVD, WAR is not setting different offensive and defensive replacement values. It's just comparing every element of the player to average, and then making one adjustment for playing time.

    You guys have some other interesting points which I need to return to later.....

69. Dvd Avins Says:
    June 18th, 2011 at 5:55 pm

    But all value is fungible. If you're so much better in defense and exactly that much worse in offense, your overall value does not change. Rrep is computed from offense, but is actually an adjustment to overall value that cares not where the rest of the contribution comes from.

    One mistake I did make is in Rdef, I need to add back the Rpos value that is currently subtracted from a 1B/DH. If DH were not an available position (and despite being an AL fan, I wish it weren't) then zero-level defense would be a harder nut to crack. But in the present environment, I don't see the difficulty.

70. Dvd Avins Says:
    June 18th, 2011 at 5:59 pm

    BTW, a better Rrep would be computed from innings fielded, as well as PA, but I don't think that's they way it's done here.

71. Dvd Avins Says:
    June 18th, 2011 at 7:46 pm

    OK, I'm now prepared to admit at least temporary defeat in assigning wins above replacement to offense and defense. I believe my assignation of offensive runs above nothing and defensive runs above nothing is correct. But the scheme I was using, which assigns WAR in proportion with the runs-above-nothing components, would have a player with more offense (Bonds) assigned more defensive value than a player with similar defensive runs (Smith) merely because his total WAR was higher. That's clearly not reflective of any useful perspective.

72. Dvd Avins Says:
    June 18th, 2011 at 7:51 pm

    Sorry, I meant Bonds and Jones. Even Jones had more Rdef than Bonds, but close enough to provide a useful comparison. Smith is in a different category.

73. Dvd Avins Says:
    June 18th, 2011 at 8:23 pm

    OK, three quarters of Rrep seems a much better proxy for the difference between zero Rpos than all of Rpos does. So using 3/4*Rrep + Rfield + Rpos, these are the only player to have 400+Rdef in the DH era (1973-present):

    629 Cal Ripken
    619 Ozzie Smith
    559 Ivan Rodriguez
    558 Omar Vizquel
    510 Mark Belanger*
    435 Gary Carter
    434 Buddy Bell*
    425 Andruw Jones
    413 Alan Trammell

    *Career started before 1973 and is not counted in its entirety. (more relevant for Belanger than for Bell)

74. Michael Sullivan Says:
    June 18th, 2011 at 10:24 pm

    I agree with you on the last johnny, but this "defensive value" isn't going to quite jibe with our conceptions of greatest defensive players. And it's particularly bad when talking about the worst.

    Guys who play corners, 1B or a lot of DH will look worse the longer they play by this measure. The opposite is true of SS/C. In terms of "defensive value" this makes sense, but in terms of "who is the greatest/worst defender" some guy who plays average 1B or RF for 10 seasons and DH for another 10 shouldn't be compared to manny, or the guy I think is probably the worst fielder of all time:

    Here's a guy with -120 rfield, and he did it in only 977 games in the field. -202 by your measure, but with only 5482 PAs (consider Winfield's 12358 PAs to get to -257).

    Danny Tartabull is the Sandy Koufax of terrible defense. Not a lot of career value, but untouchable on rate stats.

75. nesnhab Says:
    June 19th, 2011 at 6:43 pm

    I don't remember Mark Belanger ever being called "The Blade" and I was living in Baltimore the last three years of his career. I do remember that nickname was given to Tom Hall. It fit Tom Hall, and it would have fit Mark. Honest to goodness I don't remember ever hearing Belanger called that.

76. nesnhab Says:
    June 19th, 2011 at 6:49 pm

    Since we're on the subject--Dave Johnson was never "Davey" until sometime after he was named Manager of the Mets. I have several Baseball Registers and Baseball Guides from his playing days that attest to that fact.

    Similarly, if Mark Belanger had been known as the "Blade,", the 1978 Baseball Register could have said so, and it didn't.

77. Johnny Twisto Says:
    June 19th, 2011 at 8:24 pm

    Catching up on a few comments here...

    So if you consider Mark Belanger, you have to say some of his value comes from hitting better than Fred Astaire, not just from his fielding.

    Well, I guess it depends how you are defining "value." If you think hitting better than Fred Astaire, or Danny DeVito, is value, then sure. But teams have options better than those. I'm sure the Orioles could have found plenty of players who would have hit better than Belanger, and tons more who hit better than Astaire. Belanger hit just enough to be worth playing for his glove.

    Fielding value is (as Johnny Twisto said) Rpos + Rfield. That's not value above replacement value, it's value above nothing.

    No, it's value above (or below) average.

    Dvd, I have to admit I am getting lost in #65/66. I'm not sure how looking at the comparisons to Astaire/DeVito is useful.

    ***

    I'm having trouble with the construction of "defensive value over nothing."

    I'm not sure if you are referring to my #54 or something else. I have trouble with it too, but I don't know if it's a concept which matters much except in trying to compare the defensive "contribution" of a DH to other positions. And I don't think there's any one right answer to that, except that other positions must be valued higher than DH.

    ***

    Rrep is computed from offense, but is actually an adjustment to overall value that cares not where the rest of the contribution comes from.

    You've lost me again. What do you mean it "is computed from offense"?

    a better Rrep would be computed from innings fielded, as well as PA

    I think you're right that it's only based on PA, and you're right this could cause an issue. Maybe not so much now, with smaller benches. But back in the day there were more guys who only pinch hit, or only appeared as defensive subs, and their PA and def innings have little connection.

    ***

    this "defensive value" isn't going to quite jibe with our conceptions of greatest defensive players

    I agree, and I think that's OK. "Greatest" doesn't have to be equivalent to "most valuable." I have no problem with someone calling Keith Hernandez one of the greatest defenders, even if he wasn't as valuable plying his trade at 1B.