So, how much stock do you put in Baseball-Reference's Rfield stat?

[MyDogSparty]

I looked at Baseball-Reference and did a search on the All-Time career leaders in the Rfield stat hoping to discover the top fielders the game has seen. Below is a list of the top 20 according to that statistic....

Brooks Robinson, 294
Mark Belanger, 241
Andruw Jones, 239
Ozzie Smith, 239
Roberto Clemente, 205
Barry Bonds, 189
Willie Mays, 189
Carl Yastrzemski, 185
Cal Ripken, 181
Joe Tinker, 180
Buddy Bell, 177
Jim Piersall, 176
Paul Blair, 173
Brian Jordan, 163
Germany Smith, 159
Clete Boyer, 158
Robin Ventura, 156
Bid McPhee, 154
Ivan Rodriguez, 152
Al Kaline, 150
Joe Gordon, 150

...so my question is what are people's opinions on the Pros & Cons of that statistic?

[Matthew C.]

I looked at Baseball-Reference and did a search on the All-Time career leaders in the Rfield stat hoping to discover the top fielders the game has seen. Below is a list of the top 20 according to that statistic....

Brooks Robinson, 294
Mark Belanger, 241
Andruw Jones, 239
Ozzie Smith, 239
Roberto Clemente, 205
Barry Bonds, 189
Willie Mays, 189
Carl Yastrzemski, 185
Cal Ripken, 181
Joe Tinker, 180
Buddy Bell, 177
Jim Piersall, 176
Paul Blair, 173
Brian Jordan, 163
Germany Smith, 159
Clete Boyer, 158
Robin Ventura, 156
Bid McPhee, 154
Ivan Rodriguez, 152
Al Kaline, 150
Joe Gordon, 150

...so my question is what are people's opinions on the Pros & Cons of that statistic?

Aren't these just career Total Zone leaders? I think we have had some discussion on the merits if TZ recently, right?

[MyDogSparty]

Aren't these just career Total Zone leaders? I think we have had some discussion on the merits if TZ recently, right?

I don't know enough about either stat to tell you if they're the same. Looking at baseball-reference list for TZ, the values are a little different even for the players that are retired. Other than saying TZ comes from "play-by-play" data, I can't find a definition of what it means.

[Matthew C.]

I don't know enough about either stat to tell you if they're the same. Looking at baseball-reference list for TZ, the values are a little different even for the players that are retired. Other than saying TZ comes from "play-by-play" data, I can't find a definition of what it means.

http://www.baseball-reference.com/ab...tal_zone.shtml

[MyDogSparty]

http://www.baseball-reference.com/ab...tal_zone.shtml

Hmmm...the value between Total Zone and "rfield" (WAR runs fielding) aren't the same for all retired players. They're close but they're not the same. AND the "rfield" data exists in the DB for players prior to when play-by-play data exists.

[stevebogus]

Hmmm...the value between Total Zone and "rfield" (WAR runs fielding) aren't the same for all retired players. They're close but they're not the same. AND the "rfield" data exists in the DB for players prior to when play-by-play data exists.

I know that Total Zone was calculated for fielders back to the mid-1950s, which was the extent of Retrosheet boxscore data as of a few years ago. Since then Retrosheet has filled in seasons back to 1920. I really don't have any idea if Total Zone is being used for those earlier seasons or not. Maybe some other fielding metric is being substituted.

Total Zone works best when the hit locations are known. The problem with earlier seasons is many games are missing that information. If that information is missing an estimate is made based on the known tendencies of the hitter. Excellent information is available from the late 1980s to the present. Good information exists for retrosheet games when a game description exists. If only a boxscore is available then we don't know the hit locations unless perhaps a partial account exists somewhere (I'm speculating on that point).

For modern seasons I believe "rfield" is calculated by adding 3 fielding components together. Those are the Total Zone estimates for range, doubleplays, and arm (Rtz, Rdp, and Rof or Rctch). Infielders do not have an arm rating.

[stevebogus]

One possible reason for the discrepancies in career fielding runs could be rounding errors made when the seasonal totals are added up. The original data was presented in 1/10 run increments.

[Matthew C.]

Hmmm...the value between Total Zone and "rfield" (WAR runs fielding) aren't the same for all retired players. They're close but they're not the same. AND the "rfield" data exists in the DB for players prior to when play-by-play data exists.

I am not sure what the differences are, but I know it is TZ. Also, the defense numnbers are estimated prior to bpb data.

[leewileyfan]

If these are the top 20 defensive players ever to have played the game, I will humbly eat my hat. Inotice that the numbers cut off @ 150; so it suggests the numbers provided are accumulated over career or peak seasons.

Brooks Robinson, 294
Mark Belanger, 241
Andruw Jones, 239
Ozzie Smith, 239
Roberto Clemente, 205
Barry Bonds, 189
Willie Mays, 189
Carl Yastrzemski, 185
Cal Ripken, 181
Joe Tinker, 180
Buddy Bell, 177
Jim Piersall, 176
Paul Blair, 173
Brian Jordan, 163
Germany Smith, 159
Clete Boyer, 158
Robin Ventura, 156
Bid McPhee, 154
Ivan Rodriguez, 152
Al Kaline, 150
Joe Gordon, 150


Without too much effort, here's a list of defensive players whom my ratings on defense = 150 or above. Not many appear on the posted list. [I go back only to 1901 and after]. This only scratches the surface.

Speaker
Dom DiMaggio
Paul Waner
Kirby Puckett
Richie Ashburn
Max Carey
Sam West
Taylor Douthit
Jimmy Sheckard
Omar Moreno
Maddox
North
W. Wilson
Frisch
Gehringer
Mazeroski
Br. Robinson
O. Smith

.... haven't tallied catchers yet, but I-Rod and Al Lopez seem to be qualifiers.

[Matthew C.]

If these are the top 20 defensive players ever to have played the game, I will humbly eat my hat. Inotice that the numbers cut off @ 150; so it suggests the numbers provided are accumulated over career or peak seasons.

Brooks Robinson, 294
Mark Belanger, 241
Andruw Jones, 239
Ozzie Smith, 239
Roberto Clemente, 205
Barry Bonds, 189
Willie Mays, 189
Carl Yastrzemski, 185
Cal Ripken, 181
Joe Tinker, 180
Buddy Bell, 177
Jim Piersall, 176
Paul Blair, 173
Brian Jordan, 163
Germany Smith, 159
Clete Boyer, 158
Robin Ventura, 156
Bid McPhee, 154
Ivan Rodriguez, 152
Al Kaline, 150
Joe Gordon, 150


Without too much effort, here's a list of defensive players whom my ratings on defense = 150 or above. Not many appear on the posted list. [I go back only to 1901 and after]. This only scratches the surface.

Speaker
Dom DiMaggio
Paul Waner
Kirby Puckett
Richie Ashburn
Max Carey
Sam West
Taylor Douthit
Jimmy Sheckard
Omar Moreno
Maddox
North
W. Wilson
Frisch
Gehringer
Mazeroski
Br. Robinson
O. Smith

.... haven't tallied catchers yet, but I-Rod and Al Lopez seem to be qualifiers.

Well, those numbers do not account for position - so guys like Bonds drop a lot after that happens, obviously.

[leewileyfan]

Well, those numbers do not account for position - so guys like Bonds drop a lot after that happens, obviously.

IF the metric and its results do not account for such in incidental consideration as position, then why bother publishing extreme lists of best-evers until that little matter is resoved.

Or, is the argument made that 1B, C, 3B need not apply for consideration?

[Matthew C.]

IF the metric and its results do not account for such in incidental consideration as position, then why bother publishing extreme lists of best-evers until that little matter is resoved.

?

I don't know - why don't you ask BR?

[leewileyfan]

I don't know - why don't you ask BR?

Two reasons:

1. I was responding originally to the thread starter, who started with this : "So, how much stock do you put in Baseball-Reference's Rfield stat?"

2. You stated with a certain degree of conviction that "position" once considred would dilute the findings.

3. The question, as I put it, was obviously rhetorical and directed to whomever at B-R published the list in the first place. If you took it as a personal attack on you, then that's your perception, not mine.

As you like to say, I was addressing the question put to us on the board, as I saw it.

[Matthew C.]

Two reasons:

If you took it as a personal attack on you, then that's your perception, not mine.

.

No personal attack felt whatsoever. I don't get my feelings hurt about baseball - too many important things out there.
I was being serious...you can ask them if you want. But since you were being rhetorical...

[leewileyfan]

I looked at Baseball-Reference and did a search on the All-Time career leaders in the Rfield stat hoping to discover the top fielders the game has seen. Below is a list of the top 20 according to that statistic....





Robin Ventura, 156

...so my question is what are people's opinions on the Pros & Cons of that statistic?

Rather than guess and suppose the context of the B-R metric being questioned, I looked it up and scribbled a few notes. I picked a single player's career numbers [Robin Ventura] to check out the math; and it is done this way:

1. They take the player into consideration defensively at whatever position he may have played and for whatever number of innings/games he appeared there; and they then apply defense runs as a + or - for that position at the time he played it against a standard that is not clearly explained. To this degree the metric is definitely position-specific with each data entry.

Ventura is credited at his primary position [3B] with +154 DR; so it is clearly career-based. He appears to get +8 DR credit for play at 1B and + 0 DR @ 2B, with -1 DR @ SS. Now, if we add up 154 + 8 + 0 - 1 = 161, which is not equal to the 156 credited to Ventura. That suggests that a positional penalty is decucted from his + 8 at 1B; OR bonus points impact his minimal numbers at 2B & SS. I believe he played all of seven inninings at each.

In specific response to the poster who started the thread, I do not appreciate the metric at all:

1. It purports to whittle down defensive performance to a best "20" with an added ever connotation in a method that is more complex that autopsying a jigsaw puzzle.

2. In seems to imbed positional bonus points and discounts, specific to the position played, regardless of the quality of defense rendered by the player. A solid metric should have such considerations in its essential weightings of input data.

3. It appears that the +/- runs are above average; but I hesitate to conclude that, because even the likes of Ozzie Smith and Brooks Robinson tail off in later career years before moving to other positions [if their bats are valuable].

4. This lack of clarity for the standard is particularly evidenced in players like Joe Tinker, Mark Belanger, Roberto Clemente and Brooks Robinson due to several issues of career duration, position played, performance by contemporary opponents, etc. These players have numbers that seem to be compared to a much lower standard than average.

5. It's a B-R interpretation of a metric with the interpretation suggesting a certain absoluteness of fact, with unexplained nuances and questionable results.

[leewileyfan]

I looked at Baseball-Reference and did a search on the All-Time career leaders in the Rfield stat hoping to discover the top fielders the game has seen. Below is a list of the top 20 according to that statistic....

...so my question is what are people's opinions on the Pros & Cons of that statistic?

Inclusion of the names of a number of players and what appeared to me as astounding numbers [Yastrzemski] got me digging a bit further, trying to get as close as possible to the original source for the TZ metric being applied.

The link provided below is to a 2008 article [part 2] by Sean. Note, for example, his challenges and adjustments made for some players, like Yaz.

Seems like B-R has applied numbers that are pre-adjustments.

http://www.hardballtimes.com/main/ar...o-1956-part-2/

[stevebogus]

Rather than guess and suppose the context of the B-R metric being questioned, I looked it up and scribbled a few notes. I picked a single player's career numbers [Robin Ventura] to check out the math; and it is done this way:

1. They take the player into consideration defensively at whatever position he may have played and for whatever number of innings/games he appeared there; and they then apply defense runs as a + or - for that position at the time he played it against a standard that is not clearly explained. To this degree the metric is definitely position-specific with each data entry.

Ventura is credited at his primary position [3B] with +154 DR; so it is clearly career-based. He appears to get +8 DR credit for play at 1B and + 0 DR @ 2B, with -1 DR @ SS. Now, if we add up 154 + 8 + 0 - 1 = 161, which is not equal to the 156 credited to Ventura. That suggests that a positional penalty is decucted from his + 8 at 1B; OR bonus points impact his minimal numbers at 2B & SS. I believe he played all of seven inninings at each.

In specific response to the poster who started the thread, I do not appreciate the metric at all:

1. It purports to whittle down defensive performance to a best "20" with an added ever connotation in a method that is more complex that autopsying a jigsaw puzzle.

2. In seems to imbed positional bonus points and discounts, specific to the position played, regardless of the quality of defense rendered by the player. A solid metric should have such considerations in its essential weightings of input data.

3. It appears that the +/- runs are above average; but I hesitate to conclude that, because even the likes of Ozzie Smith and Brooks Robinson tail off in later career years before moving to other positions [if their bats are valuable].

4. This lack of clarity for the standard is particularly evidenced in players like Joe Tinker, Mark Belanger, Roberto Clemente and Brooks Robinson due to several issues of career duration, position played, performance by contemporary opponents, etc. These players have numbers that seem to be compared to a much lower standard than average.

5. It's a B-R interpretation of a metric with the interpretation suggesting a certain absoluteness of fact, with unexplained nuances and questionable results.

I have a bunch of comments here. Please do not take offense, as I am not picking on you personally.

Regarding Robin Ventura's 156 fielding runs, I cannot find where that number appears at Baseball-Reference. Also, the list presented by MyDogSparty has several minor discrepancies. I don't know how he generated that list, and the searches that I have done come up with slightly different numbers. Perhaps he managed to access an old list which has been superceded by newer tallies.

As for your list of issues with Total Zone I believe it is fair to point out that:

1. The top 20 list was created by MyDogSparty and Total Zone rates all players since 1954 (so far as I know). Ratings from earlier seasons use an unspecified method. The cutoff was the choice of the person who created the list. The Total Zone method may be complex, but that does not render it any less accurate (in general) than a number of other estimates.

2. Total Zone does not include bonus points or discounts. I believe you are confusing it with Runs/Wins Above Replacement (RAR and WAR), which does have those modifiers. RAR and WAR use Total Zone for fielding runs.

3. Any +/- rating/ranking implicitly assumes that the player is being compared to an average player unless otherwise specified (WAR is based at "replacement level" rather than average). I don't understand your comment regarding Brooks and Ozzie. Both players declined in fielding runs at the end of their careers. What is remarkable is how long they remained above average fielders, but then they are both in the Hall of Fame primarily due to their superb fielding skills.

4. Yes I agree there is a lack of clarity. But then you list four players renowned for their fielding skill and suggest that their numbers are inflated somehow. I guess Babe Ruth's home run totals are inflated, and Ty Cobb's batting averages. Great hitters separate themselves from the ordinary hitters by wide margins. Don't you suppose that great fielders do the same? There may be other influences at work here (such as park effects, and fielding behind a good pitching staff). Maybe some of the credit that Brooks and Belanger got rightfully belongs to Palmer, Cuellar, and McNally. But I don't see how that invalidates the method any more than hitting in Fenway invalidates Wade Boggs' batting averages. You (and I) do not have to accept the numbers at face value.

5. I agree that we are not dealing with "hard facts" here. These are estimates and should be treated as such. The question is whether or not Total Zone produces good and reasonable estimates. I think it can safely be assumed they are not perfect estimates.

Evaluating fielders from fielding stats is a tricky business because fielding stats only tell part of the story. Pitchers get dinged for the hit if a fielder cannot reach the ball or misplays it without making an obvious error. Some will never trust any fielding runs estimate. I can sympathize, but I'd rather have an estimate anyway. And if competing methods produce different results, then I'd like to figure out why.

[stevebogus]

Inclusion of the names of a number of players and what appeared to me as astounding numbers [Yastrzemski] got me digging a bit further, trying to get as close as possible to the original source for the TZ metric being applied.

The link provided below is to a 2008 article [part 2] by Sean. Note, for example, his challenges and adjustments made for some players, like Yaz.

Seems like B-R has applied numbers that are pre-adjustments.

http://www.hardballtimes.com/main/ar...o-1956-part-2/

Since Jimmy Wynn was mentioned I looked him up. Baseball-Reference is using the adjusted data. His 1969 fielding was adjusted from -27 to -9 runs. He gets a -14 that season because of a -5 run arm rating, which was not part of the original estimate.

[leewileyfan]

I have a bunch of comments here. Please do not take offense, as I am not picking on you personally.

Regarding Robin Ventura's 156 fielding runs, I cannot find where that number appears at Baseball-Reference. Also, the list presented by MyDogSparty has several minor discrepancies. I don't know how he generated that list, and the searches that I have done come up with slightly different numbers. Perhaps he managed to access an old list which has been superceded by newer tallies.

Here's the link I used to Robin Ventura: http://www.baseball-reference.com/pl...01-field.shtml

As for your list of issues with Total Zone I believe it is fair to point out that:

1. The top 20 list was created by MyDogSparty and Total Zone rates all players since 1954 (so far as I know). Ratings from earlier seasons use an unspecified method. The cutoff was the choice of the person who created the list. The Total Zone method may be complex, but that does not render it any less accurate (in general) than a number of other estimates.

I presumed that the poster compiled the list. I believed it was safe to believe that he perused a larger list and posted the top 20 from that. He did cite his source; and the number for Ventura seemed consistent with the content I found at the link. I wasn't going to challenge his effort in making a list. I was astounded by the content and the numbers.

2. Total Zone does not include bonus points or discounts. I believe you are confusing it with Runs/Wins Above Replacement (RAR and WAR), which does have those modifiers. RAR and WAR use Total Zone for fielding runs.

I am learning more about TZ than I ever really wanted to. I saw one or two players where the run totals seemed off by a handful; and, as I stated in my post, made an assumption that position bias might have had something to do with those minor discrepancies.
I am aware of those imbedded + and - weighting for positions. As for WAR, RAR and all variations, I pass and disregard them out-of-hand. Not being ornery here, just stating a simple fact that the entire replacement concept is something I can do without.

3. Any +/- rating/ranking implicitly assumes that the player is being compared to an average player unless otherwise specified (WAR is based at "replacement level" rather than average). I don't understand your comment regarding Brooks and Ozzie. Both players declined in fielding runs at the end of their careers. What is remarkable is how long they remained above average fielders, but then they are both in the Hall of Fame primarily due to their superb fielding skills.

For me, + or - convey no special attachment except that they make clear they are evaluating above or below a base line, usually presumed to be zero.

As to Brooks and Ozzie, I was fortunate to have seen both of them play, often. Both were great at their position. However, if the presumtion here is that +205 runs [Clemente]; +294 [Brooks Robinson]; +241 [Belanger]; +185 [Yaz]; +180 [Tinker] denote defensive superiority above average MLB performance at each position, career then the numbers are even more ludicrous than I had thought.

In the context of my own approach to measuring defense, I use a player rating that is formulated to resemble the old standby, fielding %. This is a conversion convention intended to be easily understood in forming a basis for derivin +/- defense runs against a standard.

So, a player with a .981 rating at 3B for his career, against a running MLB average at SS over the years of that career @ .935, will calculate to @ +11.4 defense runs above average in a typical season. To reach +294, Brooks Robinson would have to play at the .981 level against his .935 peers for 25.8 seasons of full-time play. No break in as a rookie, no injuries or setbacks, no decline period. That's why I use the word "astounding" [don't use it much]. Simply put: I ain't buying it.

4. Yes I agree there is a lack of clarity. But then you list four players renowned for their fielding skill and suggest that their numbers are inflated somehow. I guess Babe Ruth's home run totals are inflated, and Ty Cobb's batting averages.

Won't mix defense with batting averages. As a Red Sox fan, I am very familiar with Yaz [who replaced my boyhood idol, TSW] and his play. He was a fine defensive LF. He credited TSW, loud & clear, with being his mentor in learning to play LF at Fenway. Yaz had a long career; but much of it was at 1B.

I see a number like +185 defense runs for a guy who played mostly in a corner OF or 1B position; and I do not immediately thinK: Yaz.
[Al Simmons, maybe. Jimmy Sheckard, I'd have to check more closely. Rickey Henderson? Hmmm]. Again, +185 runs suggests a non-stop level of defensive play @ +15 defense runs above MLB average for some 12/3 seasons.

5. I agree that we are not dealing with "hard facts" here. These are estimates and should be treated as such. The question is whether or not Total Zone produces good and reasonable estimates. I think it can safely be assumed they are not perfect estimates.

Agreed. Not even close.

Evaluating fielders from fielding stats is a tricky business because fielding stats only tell part of the story. Pitchers get dinged for the hit if a fielder cannot reach the ball or misplays it without making an obvious error. Some will never trust any fielding runs estimate. I can sympathize, but I'd rather have an estimate anyway. And if competing methods produce different results, then I'd like to figure out why.

Me, too. Which is why I am laboring over a book a NYC agent had for almost 2 years before returning it unpublished. Fact is, he did me a favor, in that I emphasized defensive players far to much to have it sell in a mass market. I am reworking it [1901-2010] with most key roster players for all teams over that span, a tedious and often daunting task. Tomorrow I re-visit and update 1943, showing chronological progress. It's the added player inclusions, good, bad, indifferent that consume the time; but it's a far more rewarding study.

Key, for me, is recognizing each position as a separate department in a much larger operation. Each has its own history, equipment, area of responsibility, degree of game involvement, expectation of fielding opportunities and challenges. Each also has its unique negatives - damage done by plays not executed or chances muffed. Each also has unique input weightings for fielding opportunities as to degrees of difficulty or routine in execution.

[stevebogus]

leewileyfan:

I have to ask, are you claiming that it is implausible that great fielders can have that much impact in terms of runs?

Total Zone is based on knowing hit locations. When a fielder fails to make a play it is normally a hit or error (or sometimes a fielders choice). Often you cannot tell who to "charge" with the hit. Maybe it was a line drive that nobody had a real chance at, maybe it fell just in front of an outfielder, maybe it was a grounder hit between short and third. Total Zone, if it does not have the exact location and type of hit, makes estimates. I think that is a reasonable thing to do. And I think that Total Zone does a good job of estimating the impact of fielding.

If Brooks Robinson really did save 293 runs with his fielding skill that works out to be 14 runs per 1200 innings. It would take maybe 15-18 fielding plays to produce a 14-run swing, or about 1 play every 8 or 9 games that an ordinary thirdbasemen would not make. That doesn't seem unreasonable to me for a 16-time Gold Glove winner. In his "best" seasons Total Zone has him at about twice that rate. If Brooks, at his best, was taking away 1 hit every 4 games, well then that doesn't really seem outrageous to me. Brooks was widely regarded as the best thirdbaseman ever. Maybe he was both great and lucky in those seasons. Lucky in the sense that maybe there were more balls hit in his direction in those seasons than TZ estimates, making Brooks look a little better. Regarding the Baltimore Orioles of the late 1960s and 1970s: They were one of the greatest fielding teams ever. Year after year they turned a very high pergentage of balls in play into outs. This can be estimated by a Bill James invented stat dubbed DER (Defense Efficiency Rate, I think). There are a couple of ways to figure this, but essentially it is an estimate of the percentage of balls hit that are turned into outs. The Orioles of that era have several of the highest DER seasons in history. The Orioles had excellent pitching then, but a lot of the credit deservedly belongs to Brooks Robinson, Mark Belanger, Paul Blair, and the other Oriole fielders. Total Zone indicates that Robinson and Belanger were superb fielders. Their high ranking is consistent with the performance of their team. They were great fielders on a historically great fielding team.

Regarding Total Zone, I have noticed some things that don't necessarily make sense to me. That may mean I don't understand how a number was arrived at, or it may mean I *think* I know how it was computed and disagree with it.

Regarding Yaz, TZ credits him with 54 runs due to his arm (throwing out baserunners trying to advance or keeping baserunners from advancing). He is also credited with 38 runs in Fenway vs. 44 runs away from Fenway for his range. Yaz played centerfield one season, in 1964, and he nearly led the AL in outfield putouts, falling a dozen short of the leader. Yaz was a good centerfielder and overqualified to play left. I don't know that I trust the arm rating. It seems implausible to me that a leftfielder can have that big of an impact. But Yaz did lead the AL on outfield assists 7 times. I suspect that some (maybe most) of this may be due to his ability to read the bounce of the ball off the Green Monster and hold opposing hitters to singles or throw them out at second. Then again, leftfielders in general have weaker arms than the other outfield positions. Yaz is being compared to fielders who don't throw well.

Regarding the "baseline", or the zero-run level of TZ: It appears that for each season the AL and NL fielders are combined. For example, Brooks Robinson is not being compared to his AL opponents, he is being compared to all the thirdbasemen in both leagues. I believe there is potential for skewed results here. In that period the NL had higher batting averages and an overall higher rate of hits on balls in play. This might be due to the NL having poorer fielders, or better hitters, or weaker pitchers, larger outfields to cover, or a number of other factors. But when you rank the fielders it seems to me that the AL fielders of that time are generally going to look better than the NL fielders. I'm not sure this is justiified. I think it would have been better to keep the leagues separated.

Regarding home and road data: There are a few oddities I have noticed. It is normal for players to field better in home games. Unless they play in an unusual park players will hit for higher averages at home. When you analyze fielding stats on a home/away basis that translates into better fielding numbers at home. So it is pretty unusual for a fielder to gather more fielding runs in away games. On a single-season basis there is considerable "scatter" here, but when a trend is established over a period of years that gets my attention. I noticed that White Sox shortstops playing in old Comiskey Park often accumulated more fielding runs in away games. This happened to Luis Aparicio, Ron Hansen, and Bucky Dent. Aparicio actually accumulated more "Rtzrd" than any other shortsop, 86 compared to 65 for Ozzie Smith. But those two great shortstops had a huge difference in fielding runs at home. 148 for Ozzie Smith and 58 for Aparicio. The home/road split for Ozzie is essentially normal, for Aparicio it is the most unusual split I have come across. My guess is that TZ makes an adjustment for park effects and it just doesn't work for old Comiskey. We know very little of how ballparks affect fielding stats.

As you can tell, I am fascinated with fielding and the attempt to determine objectively just who the great fielders are and how big of an impact they had. Each system seems to have its own limitations and quirks. Rather than damn them I'm trying tio understand them.

[leewileyfan]

leewileyfan:

I have to ask, are you claiming that it is implausible that great fielders can have that much impact in terms of runs?

Yes, depending upon the position played. To be clear, I am not contradicting myself, having previously stated that I hate the concept =f +/- credits assigned to positions as defined by the position per se. The limitations my metric places on position values is based solely on such factors as opportunity; types of plays to be execuded; and relative damage done by misplays or range shortcomings. Such considerations are imbedded within my metric.

Total Zone is based on knowing hit locations. When a fielder fails to make a play it is normally a hit or error (or sometimes a fielders choice). Often you cannot tell who to "charge" with the hit. Maybe it was a line drive that nobody had a real chance at, maybe it fell just in front of an outfielder, maybe it was a grounder hit between short and third. Total Zone, if it does not have the exact location and type of hit, makes estimates. I think that is a reasonable thing to do. And I think that Total Zone does a good job of estimating the impact of fielding.

Here, at the risk of losing some readers, I do not concern myself with slicing the field into rays, curcles, slices or zones of batted ball distributions; nor do I impose my expectations of real estate to be covered by each player and failure of players to live up to those expectations. I believe the entire essence of defense at any position is hiding in full view right there in the box score.

If Brooks Robinson really did save 293 runs with his fielding skill that works out to be 14 runs per 1200 innings. It would take maybe 15-18 fielding plays to produce a 14-run swing, or about 1 play every 8 or 9 games that an ordinary thirdbasemen would not make. That doesn't seem unreasonable to me for a 16-time Gold Glove winner. In his "best" seasons Total Zone has him at about twice that rate.

Regarding Total Zone, I have noticed some things that don't necessarily make sense to me. That may mean I don't understand how a number was arrived at, or it may mean I *think* I know how it was computed and disagree with it.

I have no intention of debating against myself; so I selected only this portion of the extended question simply because it seems to defend TZ and all the subjective nuances you propose in Brooks' favor.

First, I have rated Brooks against average, not rplacement or bench. During Brooks' long career, some contemporary opposing 3B included Shannon, Santo, Clete Boyer, Graig Nettles, Aurelio Rodriguez, Rico Petrocelli, Melton [to name a few]. If +294 defense runs for Brooks Robinson is above average, I disagree.

My metric, designed to evaluate position players in a .900 - 1.000 range [for familiarity, since it resembles fielding %] is modulated to fit generations of play: turf and management; playing surfaces; ball; glove & mitt technology; batting climate, especially pitching K's vs. BBIP outs, etc. There are outlier to the desired range, with some few players coming in @ 1.011 and a few others under .900. That too is imbedded in the metric, because it gives me a heads-up in game defense dynamics. No argument: Brooks Robinson was a great HoF 3B.

Suppose my metric rates Brooks at .980 for his career at 3B and the running MLB average is .935. My weightings convert that superiority into +10.67 defense runs over a full season of play. If Brooks played 83% of his team's total innings in any given season at 3B, that +10.67 would be adjusted to 10.67*.83 = +8.86 defense runs above average.

Let's say Brooks played all 162 games possible @ 3B for 17 seasons. At +10.67 per season, that would come to 181.4 defense runs, career above average. If we are talking "replacement," them + 294 is credible. Otherwise, it is not.

I started out my research project with a player for each position identified as Player X. This concept permitted me a bit of subjective input, using rounded numbers below positional averages, but not drastically so. Poring over 110+ seasons of rosters and stats, I did not want the added burden of calculating averages for every player who played an inning at a position in a season. Player X gave me the positional picture during periods [generation] of defensive stability in the game; and when gaps approached or surpassed that 1.000 "cap," I had to dig in to find what changes had occurred to make defense so much more efficient. Player X applied to Brooks Robinson would give me .980 [Brooks] - .920 over 17 seasons = +14.2 DR > Player X = +241.4 DR, still a far shot from +294.

Regarding Yaz, TZ credits him with 54 runs due to his arm (throwing out baserunners trying to advance or keeping baserunners from advancing).

Quite simply, I address Yaz on defense by specific position played. Yes, he played some CF; but so did Hank Aaron; and Hank did it much better. Off the top of my head, I have Yaz somewhere @ +80 defense runs above MLB average. I will check that to see if I'm far off; but I don't believe so. Again, the balance of your question is lengthy and would put me in a position of defending the relative merits of TZ, with which I disagree. [Kinda like debating against yourself]. I'll take Al Simmons or Rickey Henderson, hands down on defense in LF + prime young Barry Bonds.

Regarding home and road data: There are a few oddities I have noticed.

Here again, I risk losing readers. I do not look at home-away stats; nor do I consider park factor [although I acknowledge it has some application]. Considering defense, park factor, IMO, has most impact when one considers HR's that admittedly might have been long outs in other parks or 'web gem" outs in others. However, wind at the same park; and underlying layers of air pressure lift or pressure can have the same effect as "another park."

Then, too, who played how many innings in a bad sun field? Which parks had best/worst/normal drainage during the decades 1901-1960? How high was the grass cut in the OF in Cincinnatti in 1949? What was the height of IF grass in Yankee Stadium during the Rizzuto years? How was the ground around 1B tamped, rolled and watered in St. Louis [Brock] or LA [Wills]?

I focus on in-play play; and at that, the plate is reasonably full. Arriving at the metric was a complex undrtaking. Beyond that, my approach is Keep It Simple!

As you can tell, I am fascinated with fielding and the attempt to determine objectively just who the great fielders are and how big of an impact they had. Each system seems to have its own limitations and quirks. Rather than damn them I'm trying tio understand them.

Me, too. Hope for some positive feedback.

[MyDogSparty]

Regarding Robin Ventura's 156 fielding runs, I cannot find where that number appears at Baseball-Reference. Also, the list presented by MyDogSparty has several minor discrepancies. I don't know how he generated that list, and the searches that I have done come up with slightly different numbers. Perhaps he managed to access an old list which has been superceded by newer tallies.

Here are the step by step directions that I used to generate the list from the Baseball-Reference website...

Click on Play Index
Click Player Batting
Select the following Query Choices:
*Find Totals for Combined Seasons or Careers
*Years from 1871 to 2010
*Sort by WAR Runs Fielding
Click the GET RESULTS button and it will yield the list I provided.

...I believe the values may fluctuate for the active players. Here are the top 20 after doing the query again today:

Player, Rfield
Brooks Robinson, 294
Andruw Jones, 241
Mark Belanger, 241
Ozzie Smith, 239
Roberto Clemente, 205
Barry Bonds, 189
Willie Mays, 189
Carl Yastrzemski, 185
Cal Ripken, 181
Joe Tinker, 180
Buddy Bell, 177
Jim Piersall, 176
Paul Blair, 173
Brian Jordan, 163
Germany Smith, 159
Clete Boyer, 158
Robin Ventura, 156
Bid McPhee, 154
Ivan Rodriguez, 152
Al Kaline, 150
Joe Gordon, 150

[leewileyfan]

I have a bunch of comments here. Please do not take offense, as I am not picking on you personally.

Regarding Robin Ventura's 156 fielding runs, I cannot find where that number appears at Baseball-Reference. Also, the list presented by MyDogSparty has several minor discrepancies. I don't know how he generated that list, and the searches that I have done come up with slightly different numbers. Perhaps he managed to access an old list which has been superceded by newer tallies.

To: MyDogSparty:

Please note that the quote you attributed to me was posted by steve bogus.

I thank you for the links, which I applied, and found your numbers to be correct where listed. Gaps in the top 10 seem to be a lure for [aid subscriptions.

Question: How much faith do you personally put in those numbers, which are indeed WAR based?

[MyDogSparty]

To: MyDogSparty:

Please note that the quote you attributed to me was posted by steve bogus.

I thank you for the links, which I applied, and found your numbers to be correct where listed. Gaps in the top 10 seem to be a lure for [aid subscriptions.

Question: How much faith do you personally put in those numbers, which are indeed WAR based?

Sorry about the quote. I fixed it for you. I must have overlooked the original in steve's post and saw it in your reply. The text appears in your post without any quotes hence I thought it came from you.

How much faith do I put in the numbers, you ask? Well, I don't know enough about them to say. When I first did the query it got me wondering if some of today's young infield talent (Longoria, Tulowitzki, Wright & Zimmerman) would someday crack that list. So far it looks as though Longoria & Tulowitzki have the best chance. It somewhat surprised me that Zimmerman was so far off pace though. That's what got me wondering about the numbers and what they're supposed to be telling me. I witnessed the majority of the players' careers on the list and I will say that I considered most of them as GREAT fielders. From what I've seen of Zimmerman thus far he seems like a better fielder than the numbers indicate.