Page 3 of 3 FirstFirst 123
Results 51 to 60 of 60

Thread: Evolving MLB Defense: 1901-1929, at 2B

  1. #51
    I just posted in the History Thread in a discussion of 3B and its ranking relative to other positions and the impact of that ranking on evaluation of third basemen overall as a ripple effect.

    As part of that discussion, Pie Traynor came up. I made my post in which I reviewed Traynor's career defensively coming up with a career rating slightly below MLB average at 3B when he played and amounting to -38 total defense runs over his entire career.

    Here, just keeping my hand in while I consider the heavy lifting in trying to get published, I thought it might be interesting to relate Traynor's seasonal ratings to the defensive leaders at 3B for each of his seasons.

    MLB "average" at 3B is approximately .959 [by my metric] over these seasons; so readers can get a sense of who was best and by how much. At .970, a 3B during these seasons would be playing at +3 DR > AVG, as a rate of defensive play. For full seasons, playing time would also have to be brought into play. For simplicity, here's how 3B looked during Traynor's playing days:

    1922:

    Pinelli .972
    Ellerbe .982
    Traynor .938

    1923:

    Kamm .976
    Lutzke .975
    Friberg .959
    Traynor .955

    1924:

    Kamm .967
    Lutzke .973
    Pinelli .965
    Traynor .955

    1925:

    Traynor .962
    Pinelli .973
    Dugan .964

    1926:

    Traynor .953
    Dressen .978
    McManus ,976
    Kamm .971

    1927:

    Trayner .948
    Dressen .968
    Friberg .974
    Bluege .971

    1928:

    Whitney .953
    Myer .969
    Bluege .974
    Traynor .951

    1929:

    Traynor .944
    Whitney .966
    Lindstrom .957
    McManus .957
    J. Sewell .970

    1930:

    Traynor .952
    Whitney .964
    McManus .951
    Bluege .950

    1931:

    Stripp .957
    Urbanski .965
    McManus .974
    Traynor .936

    1932:

    Stripp .967
    J. Sewell .972
    Kamm .957
    Scharein .972
    Traynor .933

    1933:

    Whitney .959
    Scharein .970
    Traynor .945

    1934:

    Traynor .931
    Bucky Walters .952
    Whitney .954
    Ryan .967
    Vergez .964
    Kamm .972
    Werber .977
    Travis .959 [rookie]

    Always looking for feedback, comments, questions.

  2. #52
    In the "History of the Game" forum on these board, poster Ben Grimm [post #67] raised the question of Gold Glove repetitive awards and their impact, if any, on Brooks Robinson's enduring reputation, his number of GG's and his HoF induction. The poster, as I understand it, was not challenging those things, merely asking if there were some larger context considerations that might have modified any of those.

    A few nights ago, as a participant in that same discussion, I had reviewed Brooks Robinson's entire [full-time 3B] career, year-by-year, out of my own curiosity. Here are those data with a few notes as to interpretation:

    1. My metric presents player defensive Ratings is a format that resembles fielding percentage. That makes it familiar and provides ease in translating Ratings into + or - Defense Runs to some model of comparison.

    2. In my book manuscript, my model of comparison is Player X, who is always about 8 defense runs below prevailing MLB defensive average. Numerically, at 3B, this number is very nearly .925.

    3. For the sake of this response, I am using MLB 3B average on defense at the position, which is closer to .955.

    4. Playing time is [should be] a critical consideration in making awards. A player rating .980 in 62 games should NOT, in a fair evaluation, win an award over a player rating .972 in 155 games. However, the bigger the gap, the more room there is for a player with fewer games winning an award.

    Here is the Brooks Robinson 3B career, season-by season, with his ratings and those of 3B competitors from both leagues. For context, G = games. Games are derived from innings played.

    Season..........Robinson Rating...................Other 3b with Ratings, Games .............

    1958......... .969, 124G...........K. Boyer .987, 142G; Mathews .975, 146G; Hoak .974, 109G; Carey .971, 88G; Malzone .981, 153G

    1959......... .980; 80G............Malzone .977, 152G. [Malzone won GG. Note games played by each].

    1960......... .972, 150G...........Kasko .985, 77G; C. Boyer .992, 87G [Robinson won GG, rightly so. Note games played].

    The 1960 GG started a skein of continuous, uninterrupted GG's for Robinson through 1975.

    1961......... .962, 162G...........K. Boyer .969, 150G; Tuttle .967, 74G; C. Boyer .999, 133G; Malzone .963, 149G

    1962......... .967, 161G...........C. Boyer .992, 155G; Santo .967, 152G

    1963......... .970, 157G...........Mathews .979, 116 G; C. Boyer .971, 138G

    1964......... .963, 162G...........Santo .975, 158G; Ward .970, 136G; Malzone .977, 136G; C. Boyer .978, 120G

    1965......... .964, 144G...........Ward .978, 130G; Santo .969, 164G, C. Smith .973, 123G; C. Boyer .993, 142G

    1966......... .957, 159G...........Santo .991, 150G; K. Boyer .982, 121G; Bailey .969, 90G; Mathews .963, 144 G; C. Smith .966, 102G;
    ..............................................C. Boyer .996, 82G

    1967......... .995, 157G...........NONE

    1968......... .971, 159G...........Santo .969, 160G; T. Taylor .976, 134G

    1969......... .984, 163G...........Kenney 1.003, 76G [Note: Kenney playing time less than half of Robinson's].

    1970......... .962, 154G...........Rader .980, 152G; Santo .966, 147G; Kenney .988, 115G; C. Boyer .965, 121G;
    ..............................................Aur. Rodriguez .969, 153G; Nettles .987, 146G

    1971......... .978, 152G...........Melton .992, 141G; Nettles 1.004, 156G

    1972......... .974, 147G...........Aur. Rodriguez .979, 148G

    1973......... .977, 149G...........Schmidt .983, 104G; Bell .983, 152G; Nettles .996, 152G

    1974......... 1.002, 151G..........NONE

    1975......... .983, 136G..........Nettles .983, 154G; Da. Evans .984, 151G
    Last edited by leewileyfan; 07-29-2012 at 11:53 AM.

  3. #53
    About two nights ago I came upon a 2007 article on a another Internet baseball forum. The subject was the four "best" defensive shortstops in MLB history. It was an article in the sabermetric vein of ballplayer evaluation; it alluded to Win Shares and how they should be derived, interpreted and applied and it cited the work of yet another researcher.

    Bottom line was the conclusion that the best 4 defensive shortstops were:

    1. Rabbit Maranville
    2. Bill Dahlen
    3. Honus Wagner
    4. Ozzie Smith

    In the home stretch of editing my own book of defense and updating defense ratings through 2012, I wanted to see how I had these four rated. I plotted every season in the career of each, in which each played a considerable number of games at short. My findings are far different from those of the study.

    Each of the four players is listed below, with seasons at SS. What follows is my metric's rating for each season and the Defense Runs above/below MLB average for that season. The last column is Defense Runs +/- Average, adjusted precisely for actual playing time as part of the season schedule. [I love this stuff].

    Hope some find it interesting. I welcome comments, questions, other players to be compared ... whatever.

    Rabbit Maranville

    Season.................Rating............MLB AVG..............+/- DR to AVG; Actual Playing Time

    1913..................... .934............. .922....................... + 4.3
    1914..................... .964............. .928....................... +13.5
    1915..................... .944............. .927....................... + 5.5
    1916..................... .944............. .927....................... + 6.4
    1917..................... .942............. .927....................... + 5.2
    1918........................World War I
    1919..................... .969............. .928....................... +13.8
    1920..................... .953............. .929....................... + 8.4
    1921..................... .965............. .947....................... + 9.4
    1923..................... .988............. .947....................... +14.2
    1925..................... .981............. .948....................... + 5.8
    1926..................... .981............. .949....................... - 1.9
    1928..................... .973............. .949....................... + 8.0
    1929..................... .991............. .949....................... + 14.8
    1930..................... .965............. .942....................... + 7.8
    1931..................... .953............. .945....................... + 2.6

    Maranville: Career Defense Runs > AVG = +123.8, equivalent to +12.4 "Wins"

    Bill Dahlen: In all fairness to Dahlen, I can only evaluate his play beginning with the 1901 season, by which time he was 31 years of age. MLB play before 1901 had significant differences in regulations, distances, batted ball judgments, equipment [what there was of it]. My metric kicks in in 1901.

    Dahlen:

    Season...................Rating..............MLB AVG................+/- Defense Runs to AVG; Adj. for Playing Time

    *1901...................... .937............... .921....................... + 5.3*
    1902...................... .918............... .921....................... - 1.4
    1903...................... .929............... .922....................... + 3.0
    *1904...................... .935............... .922....................... + 4.6*
    *1905...................... .941............... .921....................... + 6.7*
    1906...................... .917............... .920....................... - 1.8
    1907...................... .907............... .921....................... - 5.2
    *1908...................... .967............... .922....................... + 15.7

    Dahlen's career Defense Runs = +26.9, equivalent to +2.69 "Wins" *NOTE: nice book-end and mid-span seasons, ages34, 35, and 38.



    Honus Wagner: Like Dahlen, Wagner's career began a few years before 1901; but his first season playing shortstop was 1901.

    Wagner:

    Season....................Rating..............MLB AVG.................+/- Defense Runs; Adj. for Playing Time

    1901....................... .948............... .922........................... + 4.4
    1903....................... .935............... .922........................... + 5.2
    1904....................... .912............... .922........................... - 2.9
    1905....................... .947............... .922........................... + 8.9
    1906....................... .942............... .922........................... + 6.7
    1907....................... .910............... .922........................... - 4.0
    1908....................... .917............... .922........................... - 1.9
    1909....................... .927............... .922........................... + 1.7
    1910....................... .916............... .922........................... - 2.0
    1911....................... .926............... .922........................... + 1.0
    1912....................... .945................ .922........................... + 8.1
    1913....................... .933................ .922........................... + 2.9
    1914....................... .928................ .922........................... + 1.9
    1915....................... .920................ .922........................... - 1.0
    1916....................... .905................ .922........................... - 3.8

    Wagner's Career Defense Runs = +24.1, Equivalent to + 2.41 "Wins"

    Ozzie Smith

    Season.......................Rating............... ...MLB AVG.............+/- DR; Adj. for Playing Time

    1978.......................... .996................... .947...................... 16.8
    1979.......................... .993................... .947...................... 14.2
    1980.......................... 1.019.................. .948..................... 26.9
    1981.......................... 1.007.................. .948...................... 22.4
    1982.......................... 1.015.................. .948...................... 22.7
    1983.......................... .985.................. .948...................... 12.7
    1984.......................... 1.009................... .948...................... 16.1
    1985.......................... .992.................. .948....................... 15.8
    1986.......................... .966.................. .948....................... 6.0
    1987.......................... .989.................. .948....................... 14.3
    1988.......................... .975.................. .948....................... 9.3
    1989.......................... .958.................. .948....................... 3.4
    1990.......................... .934.................. .948.......................- 4.3
    1991.......................... .939.................. .948......................- 2.9
    1992.......................... .973.................. .948...................... 7.4
    1993.......................... .997.................. .948...................... 14.4
    1994.......................... .973................... .948....................... 7.5
    1995.......................... .979................... .948....................... 3.1
    1996.......................... .976................... .948....................... 3.2

    Ozzie Smith's Career Defense Runs > AVG = +209, Equivalent to 20.9 "Wins"
    Last edited by leewileyfan; 08-15-2012 at 09:07 AM.

  4. #54
    For anyone who may be interested, there are many threads that begin with a one-on-one comparison between players. They may be contemporaries. They may have played generations apart. When I can, I contribute. However, when the threads get argumentative, or surly, or just toss several metric numbers around with no context, I just exit [stage right or left, whichever is handiest].

    Fans of WAR in all its nuanced guises may dismiss these considerations out of hand. I am no fan of WAR, since I cannot wrap my brain comfortably around a concept that distills to player talents at the lowest level in the game [MLB] destined to perform at a W-L rate of approximately 49-113 in a 162 game schedule. [That's .302 if my math is correct].

    Where I can see some practical use for Replacement Level is in the front office budget process. There, a GM might be guided in finding whatever "bargains" might be out there. It might also be a negotiations concept, with leverage in dealing with a player who vastly overestimates his own prospects. Outside of that, not much.

    That behind me, I have started a career season-by-season defensive analysis of 2B who have in common the fact that they are frequently mentioned in debates concerning "best," ""top 10" or "top 20" listings. I'll kick this off with Frankie Frisch.

    All players are first evaluated for defense at 2B, as compared to MLB 2B average during their playing years. Their ratings are then converted into Defense Runs better [+] or worse [-] than average. The +/- DR number for each is adjusted for actual playing time each season.

    FRISCH

    The defensive 2B AVG during Frisch's time was .947 - .949 by my metric.

    Season....................Rating.................. ..DR Adj. Playing Time

    1921....................... .999.......................... + 4.9 [part-time play]
    1922....................... .961.......................... + 2.5
    1923....................... .987.......................... + 12.7
    1924....................... 1.022.......................... + 18.3
    1925....................... .961.......................... + 6.7
    1926....................... .987.......................... +11.7
    1927....................... 1.027......................... +30.7
    1928....................... .966......................... + 6.7
    1929....................... .946......................... - 1.0
    1930....................... .997......................... +14.7
    1931....................... .966......................... + 5.6
    1932....................... .966......................... + 3.3
    1933....................... .931......................... - 5.5
    1934....................... .947......................... - 0.0
    1935....................... .930......................... - 3.9
    1936....................... .923......................... - 3.6

    Career Total DR........................................... +103.8 > MLB 2B AVG

    In addition, Frisch, batting is credited with 1460 RC/10099 PA, a RC rate pf .1446/PA in a league batting climate in which .1275 was average.

    I combine the DR + RC to get a performance snapshot of the player. No defense runs at any other position are considered. His total production is 1460+103.8 = 1563.8/10099 = .1548 in a .1275 + 0 average climate. This would give him a production value of 92.88 runs in a 600 PA season. Average, using the same parameters, would be 76.5 runs. Frisch, in a model season, would be +16.38 runs, or 1.64 Wins > AVG by this model.

    GEHRINGER

    For the sake of one-on-one comparison, here's Charlie Gehringer, whose career partially overlapped with Frisch.
    The 2B MLB 2B defensive AVG during Gehringer's career varied from roughly .945-.950; so for the most part the basic league figure is about .948.

    1927...................... .988....................... + 11.9
    1928...................... .965....................... + 12.0
    1929...................... .958....................... + 4.1
    1930...................... .959....................... + 4.1
    1931...................... .954....................... + 1.1 [limited playing time]
    1932...................... .961........................ + 4.9
    1933...................... .979........................ +11.6
    1934...................... .965........................ + 6.4
    1935...................... .966........................ + 6.6
    1936...................... .974........................ + 9.7
    1937...................... .976........................ + 9.7
    1938...................... .951........................ + 1.1
    1939...................... .939........................ - 2.2
    1940...................... .915........................ -11.1
    1941...................... .925........................ - 6.4

    Charlie Gehringer's career DR = +70.1.

    His batting RC, 1715/10244PA = .1674 in a batting climate with .1300 = AVG. Doing the same calculations:

    1715+70.1 = 1785.1/10244 = .1743/PA or 104.55 runs in a 600 PA model season. With AVG at 78, Gehringer rates 26.55 runs> AVG in productivity, or 2.66 "Wins."

    I hope to cover 20 2B in this effort over the next week or two. Comments/questions are welcome.

    J. ROBINSON [Same format and application of data]. Predominant 2B def. AVG = .946-.949 range.

    1948...................... .936.................... - 3.0
    1949...................... .936.................... - 4.1
    1950...................... .965.................... + 6.7
    1951...................... .980.................... +12.3
    1952...................... .941.................... - 1.0

    Jackie Robinson's DR at 2B come to +10.94. At bat, he is credited with 951 RC/5802 PA, a production rate of .1639. The batting climate for run creation while he was active was .1167. Applying the same data to a 600 PA season, we get Robinson at 10.94+951 = 961.94/5802 = .1658 production, or 99.48 runs. MLB average expectation [.1167 + 0 DR] = 70.02.

    The Robinson 2B snapshot has him at 29.46 runs 0r 2.94 "wins."


    EDDIE COLLINS [During Collins' career. prevailing 2B AVG was .920 - .925. After 1919 there was jump to .935-.940 and again to .945-.950 in mid decade, largely due to improved materials [end of WW I materiel restrictions] and a series of glove design improvements.

    1908.................... .898..................... - 3.0
    1909.................... .943..................... + 7.9
    1910.................... .943..................... + 7.9
    1911.................... .909..................... - 4.3
    1912.................... .926..................... + 1.5
    1913.................... .939..................... + 6.2
    1914.................... .914..................... - 1.6
    1915.................... .948..................... + 9.7
    1916.................... .931..................... + 3.4
    1917.................... .915..................... - 2.6
    1918.................... .950..................... + 7.9
    1919.................... .942..................... + 1.9
    1920.................... .960..................... + 8.6
    1921.................... .989..................... +13.9
    1922.................... .959..................... + 4.5
    1923.................... .968..................... + 7.2
    1924.................... .949..................... + 0.0
    1925.................... .949..................... + 0.0
    1926.................... .947..................... - 0.0
    1927.................... .943..................... - 1.0

    Collins' career defense runs = + 68.1 > AVG.

    At bat, Collins is credited with 1811 RC/12037 PA = .1505 rate. His career spans 3 batting climates, .1075, .1125 and .1150 netting out to .1120.
    Applying the same data in the same way, Collins [68.1 + 1811 = 1879.1/12037 = .1561. This equates to 93.67 in a 600 PA snapshot season, compared to [.1125+ 0 DR] * 600 = or 67.5 AVG. Collins rates at +26.17 runs or 2.62 "Wins."

    MORGAN

    MLB 2B AVG defense over the 20 year span of Joe Morgan's career averaged out to .948, with playing surfaces and enclosed stadia having more impact on batting than on defense, allowing for a transition period of one season. With all adjustments made:

    1965..................... .949................ + 0.0
    1966..................... .925................ - 6.4
    1967..................... .931................ - 5.3
    1969..................... .927................ - 4.3
    1970..................... .964................ + 4.8
    1971..................... .961................ + 4.2
    1972..................... .953................ + 1.4
    1973..................... .962................ + 4.5
    1974..................... .950................ + 0.0
    1975..................... .973................ + 7.4
    1976..................... .940................ - 2.6
    1977..................... .942................ - 2.2
    1978..................... .915................ - 8.8
    1979..................... .950................ + 0.0
    1980..................... .954................ + 1.3
    1981..................... .972................ + 4.4 [81 games]
    1982..................... .964................ + 4.0
    1983..................... .963................ + 3.3
    1984..................... .947................ - 0.0
    1985..................... .940................ - 1.4

    Morgans career DR = + 4.3.

    At bat he had 1804 RC/11329 PA = a .1592 rate. His career started in the lean-hitting '60s and spanned into the mid 1980s through which RC rates varied from .1060 to .1154. Overall, his batting climate is about .1085.

    Applying the data, with DR added in, Morgan has 4.3+1804 = 1808.3/11329 = 1.596, or 95.77 runs in a 600 season. That compares with AVG [.1085 + 0] * 600 = 65.1. Morgan is +30.67 runs or 3.07 "wins" in this scenario.


    NAP LAJOIE

    I allow myself to get involve in player metrics beginning in 1901. Before that time, there were just too many game differences for me to feel competent in evaluating defensive performance: fields and dimensions, the ball, pitching boxes and deliveries, counts ... and above all the equipment [or LACK of equipment] have me disqualifying myself from comps before 1901. Pre- 1901 Lajoie is not included here.

    The predominant AVG 2B rating I derive for the period 1901-1915 or so hover around .920 - .925.

    1901..................... .958.................... +13.5
    1902..................... .972.................... +14.2
    1903..................... .960.................... +12.1
    1904..................... .925.................... + 1.0
    1905..................... .952.................... + 4.3
    1906..................... .960.................... +12.1
    1907..................... .992.................... +22.2
    1908..................... .972.................... + 8.9
    1909..................... .947.................... + 7.3
    1910..................... .931.................... + 3.3
    1911..................... .909.................... - 1.2 [37 games]
    1912..................... .926.................... + 0.0
    1913..................... .939.................... + 5.3
    1914..................... .930.................... + 0.0 [80 games]
    1915..................... .949.................... + 7.3
    1916..................... .960.................... + 9.8

    Lajoie has +120.1 career defense runs at 2B > AVG. His RC at bat [1297/8256] = .1571, which compares to AVG batting RC at .1060. At at 600 PA season, that distills to 102.96 runs [Lajoie] vs 63.6 [AVG]. That's +39.36 runs, or 3.94 "Wins."

    BILL MAZEROSKI

    Bill Mazeroski is a departure here, because the 2B considered above had considerable impetus given their ratings by batting performance. Maz, in a career that spanned from 1956 through 1971, experienced 7 seasons in an average batting climate [1956-62], then the batting drought of the 1963-69 period. with a return to the norm in 1970.

    Mazeroski's value is essentially defense-related:

    1956.............. .966................. + 3.1
    1957.............. .960................. + 3.7
    1958.............. .974................. + 9.0
    1959.............. .947................. - 0.0
    1960.............. .975................. + 9.1
    1961.............. 1.001................. +18.0
    1962.............. .989................. +14.5
    1963.............. 1.020................. +22.5
    1964.............. .983................. +12.7
    1965.............. 1.004................. +16.6
    1966.............. 1.005................. +20.9
    1967.............. .975................. + 9.7
    1968.............. .984................. +11.2
    1969.............. .975................. + 3.7 [61 games]
    1970.............. .991................. + 9.6 [99 games]
    1971.............. .961................. + 1.9 [38 games]

    Mazeroski's total career defense runs > AVG comes to +164.6 runs.

    Mazeroski, with the bat, is credited with 822RC/8379PA = .0981, considerably below AVG, which comes to .1085.

    Applying the numbers, same format as with all others: 822RC + 164.6DR>AVG/8379 = 986.6/8379 = .1177. Distilled into a 600 PA season, this puts Mazeroski at 70.62 runs vs. [.1085 + 0 DR * 600] = 65.1 runs. This would put Mazeroski at +5.52 runs, or .552 "Wins."

    LOU WHITAKER

    Lou Whitaker's career spanned the years 1978-1955, a period in which the batting climate was pretty well recovered from the '60s slump. When he broke in, the DH was 5 years in place; and the average RC/PA rate over the span of his playing time was .1175. Many have questioned why he isn't in the HoF; and the only knock I can see is that he played in fewer than 135 games in 11 of his 17 seasons in which there was a full slate of games. [1981 excluded].

    1978......................... 1.001................ +14.4
    1979......................... .992................ +10.7
    1980......................... .973................ + 7.1
    1981......................... .994................ +15.0
    1982......................... .989................ +14.2
    1983......................... .949................ + 0.0
    1984......................... .955................ + 2.0
    1985......................... .951................ + 1.0
    1986......................... .969................ + 6.0
    1987......................... .953................ + 1.5
    1988......................... .949................ + 0.0
    1989......................... .965................ + 4.8
    1990......................... .987................ + 9.6
    1991......................... .971................ + 5.8
    1992......................... .950................ + 1.0
    1993......................... .956................ + 2.4
    1994......................... .973................ + 5.4
    1995......................... .969................ + 2.5

    Whitaker comes in at +101.7 defense runs > AVG for his career.

    At bat, his 1395RC/9967 = .1400, which is 19% above MLB standard for his career. Running the numbers in the same fashion as for all other 2B rated here, we get:

    1395RC + 101.7 DR>AVG/9967 = 1497.7/9967 = .1503 * 600 = 90.16 runs in a compressed/distilled 600 PA season.
    MLB average, at .1175 + 0 DR = .1175 * 600 = 70.5 runs. Whitaker is 19.66 runs > AVG, or 1.97 "Wins."


    ROBERTO ALOMAR

    According to the records, Alomar "won" 10 Gold Gloves. He is in the HoF and Lou Whitaker is not. 'Nuff said. Same drill.

    Alomar spanned 1988 through 2004, from DH-less NL with RC/PA rates UNDER .1100 to an AL where RC/PA rated reached into the .1300++ range, similar to the 1930s. He wound up his career shuffling back and forth between leagues in a .1235 RC/PA atmosphere. The average applied here is .1260.

    Alomar created 1575 runs in 10400 PA = .1427. His defense record:

    1988........................... .952................. + 1.3
    1989........................... .947................. - 0.0
    1990........................... .938................. - 0.0
    1991........................... .934................. - 4.7
    1992........................... .925................. - 7.0
    1993........................... .953................. + 1.6
    1994........................... .953................. + 1.5
    1995........................... .957................. + 2.7
    1996........................... .983................. +10.2
    1997........................... .958................. + 2.2
    1998........................... .968................. + 5.9
    1999........................... .973................. + 7.8
    2000........................... .961................. + 4.0
    2001........................... .947................. - 0.0
    2002........................... .925................. - 6.9
    2003........................... .940................. - 2.2
    2004........................... .927................. - 1.5

    Alomar comes in at -91.0 defense runs below 2B AVG over the course of his career. Same drill with the data:

    1575 RC - 91.0 DR/10400 PA = .1427 runs/PA * 600 PA = 85.62 runs in a condensed season.
    AVG, at .1260 + 0 DR = .1260 * 600 PA = 75.6 runs. Alomar is +10.02 runs, or 1.00 Wins.


    BOBBY GRICH

    Bobby Grich broke in in 1973, splitting play between SS and 2B, logging only 376.9 innings at 2B. His debut season was the last year in the AL without the DH, a season in which AL batters produced runs at a .093 rate. To put that in context, the RC/PA in MLB has historically been .1167 [1876-Present], with some outlying lows being recorded in 1908 and 1968. Over the balance of his career the League RC/PA rose to an average of .1140.


    1972......................... .931...................... - 1.0 [42 games]
    1973......................... .988...................... +13.5
    1974......................... .960...................... + 4.1
    1975......................... .983...................... +11.3
    1976......................... .955...................... + 2.1
    1978......................... .952...................... + 1.2
    1979......................... .967...................... + 5.9
    1980......................... .983...................... +10.4
    1981......................... 1.011..................... +19.3
    1982......................... .989..................... +11.9
    1983......................... 1.001..................... +12.9 [114 games]
    1984......................... .974...................... + 4.8 [80 games]
    1985......................... 1.015..................... +15.3 [106 games]

    Defensively, Grich comes in at +111.70 DR > MLB 2B AVG. At the plate, Grich had a RC rate of 1127RC/8220PA = .1371, beating the MLB average of .1140.

    Doing the numbers, same as above for all, 1127RC + 111.7DR = 1238.7/8220PA = .1507. Distilled to a 600 PA season, that's 90.42 vs AVG [.1140 + 0 DR * 600] = 68.4, giving Grich an edge of +22.02, or 2.20 "Wins."


    ROGERS HORNSBY

    Hornsby played the better part of 4 seasons, starting as a teenager, playing SS and 3B, before starting his career at 2B in 1920. His batting record is legend; but his defensive skills are very much under-rated, especially by those who judge players by their personality or popularity, as "recollected" in media and collected histories.

    The defensive climate was at a transitional cusp in 1920, with a series of glove design technology refinements that impacted fielding ratings [my metric] from around .920-.925 in the deadball era, to about .937 in 1920 ... then up to .945-.950 after 1920.

    1920............................... .974............................. +13.5
    1921............................... .971............................. + 8.3
    1922............................... .963............................. + 6.0
    1923............................... .969............................. + 4.7 [88 games]
    1924............................... .988............................. +14.2
    1925............................... .951............................. + 1.3
    1926............................... .952............................. + 1.6
    1927............................... .993............................. +17.2
    1928............................... .961............................. + 4.8
    1929............................... .972............................. + 9.4
    1930............................... .939............................. - 0.0 [23 games]
    1931............................... .915............................. - 5.2 [67 games]

    Defensively, Hornsby is +75.8 DR above average over his career at 2B. At the plate, Hornsby created runs at [2049/9475] a .2163 rate.
    The batting RC climate AVG was .1250.

    Same application of the numbers: 2049RC + 75.8DR = 2124.8/9475 = .2243 * 600 = 134.55 runs. With contemporary average at .1275 + 0 DR [.1250*600] = 75.0, Hornsby is [134.55 - 75.0] = +59.55 runs or +6.0 "Wins."


    JEFF KENT

    Kent's career years, 1992-2008, had a batting climate of RC/PA that averaged .1190. Kent's production was 1497/9537 = .1570.

    The 1996 season is not entered because Kent played only 65 innings at 2B that year. Prodominant rating average at 2B = .948.

    1992................................ .942.......................... - 0.0
    1993................................ .922.......................... - 6.6
    1994................................ .972.......................... + 7.6
    1995................................ .954.......................... + 1.7
    1997................................ .970.......................... + 6.4
    1998................................ .959.......................... + 3.1
    1999................................ .938.......................... - 2.7
    2000................................ .948.......................... + 0.0
    2001................................ .960.......................... + 3.3
    2002................................ .961.......................... + 3.9
    2003................................ .949.......................... + 0.0
    2004................................ .941.......................... - 2.0
    2005................................ .962.......................... + 4.0
    2006................................ .975.......................... + 5.7 [98 games]
    2007................................ .939.......................... - 2.3
    2008................................ .939.......................... - 1.9 [98 games]

    Kent's career 2B defense runs add to + 20.2. Running the numbers vs. AVG:

    1497RC + 20.2DR = 1517.2 runs/9537 = .1591, which amounts to 95.45 over a 600 PA season. This is 24.05 runs above MLB 2B AVG [.1190+0 DR*600]=71.4 runs. The 24.05 runs = 2.41 "Wins."

    RECAP: Using the "Wins" as the focal point, the 2B rated here so far:

    Frisch [1.64]; Gehringer [2.66]; Robinson [2.94]; Collins [2.62]; Morgan [3.07]; Lajoie [3.94]; Mazeroski [0.55]; Whitaker [1.97]; Alomar [1.53]; Grich [2.20]; Hornsby [6.0]; Kent [2.41]; Sandberg [2.54]; Doerr [2.05]. EDIT: Doerr. ADDED: Joe Gordon [1.62]; Schoendienst [0.47]. [Thread 2].


    RYNE SANDBERG

    Sandberg's career spans 1982 through 1997 and has a batting climate of PC/PA of .1135. Sandberg had RC/PA of 1342/9282 = .1446, so he was well above the average posted by his contemporaries. His play at 2B in 1982 [24 games] is not included. He did not play in 1995.

    1983...................................... 1.015........................... +22.5
    1984...................................... .997........................... +15.8
    1985...................................... .980........................... +10.2
    1986...................................... .966........................... + 5.9
    1987...................................... .957........................... + 2.4
    1988...................................... .981........................... +10.4
    1989...................................... .957........................... + 2.9
    1990...................................... .962........................... + 4.4
    1991...................................... .964........................... + 5.3
    1992...................................... .984........................... +11.8
    1993...................................... .968........................... + 4.7
    1994...................................... .996........................... + 8.0 [54 games]
    1996...................................... .956........................... + 2.3
    1997...................................... .928........................... - 4.7

    Career total DR > 2B MLB AVG = +104.2. Running the numbers:

    1342RC + 104.2DR/9282PA = .1558 runs * 600 PA = 93.48 runs > AVG at .1135 + 0 DR * 600 = 68.1 runs. Sandberg thus rates +25.38 runs, or 2.54 "Wins."


    BOBBY DOERR

    Doerr came up in 1937, playing 2B in 36 games. His career spanned through 1951, experiencing the hig PC/PA atmospheres of the '30s [.1300+]; the pre-War years [.1200+]; the War years [.1050+] and the post War period which gradually rose back above .1200. He did not play in 1945 due to military service.

    Doerr is credited with 1169RC/8028PA = .1456, considerably above the prevailing AVG over the span of his career [.1215].

    Not experiencing any notable decline phase, Doerr retired after the 1951 season in order not to aggravate a back minor injury that had become a concern. He didn't want to risk serious injury that might occur with continued play.

    1937.............................. .980............................ + 2.7 [36 games]
    1938.............................. .953............................ + 2.0
    1939.............................. .979............................ + 9.2
    1940.............................. .964............................ + 5.8
    1941.............................. .941............................ - 1.8
    1942.............................. .961............................ + 4.5
    1943.............................. .974............................ + 9.4
    1944.............................. .948............................ + 0.0
    1946.............................. .975............................ + 9.5
    1947.............................. .969............................ + 7.2
    1948.............................. .973............................ + 8.0
    1949.............................. .978............................ + 9.7
    1950.............................. .965............................ + 9.4
    1951.............................. .966............................ + 4.5 [105 games]

    Using the same routine, Doerr has +80.1 defense runs, which, added to his batting RC brings him to .1556 [93.36 runs], which is +20.46 runs better than contemporary MLB 2B AVG over a model season of 600 PA. This converts to +2.05 "Wins."
    Last edited by leewileyfan; 11-30-2012 at 11:53 AM.

  5. #55
    A Red Sox fan since 1939, I saw the brightest thing in the 2004 season as the wonderful defensive play of Pokey Reese at shortstop. Sure, it was great to see a World Series World's Champ pennant flying over Fenway Park after 65 seasons of disappointment; but even the pennant was a bit tarnished, with a "been there/ done that" feeling:

    1. The Sox actually had come in second to the Yankees, and they were three games behind. Not wanting to seem like a malcontent [think Peggy Lee singing "Is That All There Is?", but how many times had I seen that through my childhood and youth and young adulthood? For me, the Sox real-deal championship came only in 2007 when they actually won it all ... from soup to nuts.

    2. Going back to 2004 I wanted to see just how good Pokey Reese was and what kind of contribution he'd made, filling in after the departure of Nomar Garciaparra.

    Pokey Reese ... when the Sox hit the bargain mother lode [and WHY].

    1. I am pretty sure I was not the only Sox fan to shake my head and sigh when all the sound and fury over Nomar Garciaparra's contract dispute had ended with Nomar's departure. Without going into how player salary shares were split among teams, the payroll general picture looked like this:

    Nomar Garciaparra, $11,500,000 ... now gone
    Orlando Cabrera, $6,000,000 ... the new SS
    Pokey Reese, $1,000,000 ...a new glove ... sub? late inning replacement???

    2. As it fell out in 2004, the BOSOX played 1,451 innings. The total investment at SS looked like $18,500,000 divided by whatever clauses prevailed for trades and acquistions, kinda looking like a possible upgrade in defense with a hit taken in the offense.

    3. Since I'm pretty much focused on defense, here's how the key BSX players at SS fared in 2004.

    Player......................Defense Rating...............Leage SS AVG...............+/-DR Adj for playing time

    Garciaparra.............. .898............................ .948.......................[21.4%] = - 3.98 [rate @ full season = -18.59]

    Cabrera................... .916............................ .948.......................[33.8%] = - 3.98 [rate @ full season = -11.81]

    Reese...................... .977............................. .948......................[35.0%] = + 3.80 [rate @ Full season = +10.98]

    Considering ONLY defensive contribution, in a "with you or without you approach," the BSX at short in 2004, without Reese [with others at SS for all 1,451 innings] would have shaped up like this, compared to League:

    Sox Rating [no Reese]: .913; LG AVG: .948; Defense Runs, Sox = -13.09 defense runs

    Sox Rating [all Reese]: .977; LG AVG: .948: Defense Runs, Sox = +10.85 defense runs, a swing of 23.94 runs for a full season.

    Strip all that "season - what if" away and just credit Pokey with +3.84 defense runs and the prevailing -4.58 play of others at short, that's a swing of +8.42 runs, for a bench player earning 5.4% of the budget slot for a key position.

    Great Return on Investment.

  6. #56
    A poster raised an interesting question regarding the 1990 defensive play at SS of Cal Ripken. Ripken played all but 30 or so innings for Baltimore that season, 1406.1 to be exact, and committed only 3 errors. I was atounded to find that I had Cal rated .944 at SS that season in an A.L where the average SS rated .958 Playing time considered, that put Cal about -4.9 to 5.1 DR below average. I wondered if my metric had gone off the tracks. No, I don’t believe so.

    I ran the entire AL once again, every inning of play at each position to be sure I hadn’t imbedded a glitch in the formulas. Nope. They were intact.
    The first thing that struck me was that only one team had fewer assists at SS than Baltimore, Kansas City, with 446 to Baltimore’s 447. Moreover, there was only one other AL team within 20 assists of Baltimore’s low rate. That was Oakland, with 460 assists.
    There are in-game events that can distort comparisons, like pitching staff strikeouts compared to League average. You get large numbers of K’s from the staff; and the fielders have fewer opportunities. Nope, the average staff K’d 906 batters. The Oriole’s staff fanned 776, 130 fewer K’s and 130 MORE batted balls for the fielders to share.

    Next, I looked at batted ball out distribution, especially OF to IF; and there was a possible hitch. The Oriole’s OF recorded the MOST PO in the AL, at 1,280 compared to a League average of 1,116. That can be taken as 164 fewer chances for the IF. However, how many OF PO are made on one club, when they might well have been executed by an IF on another club? So far fielders are getting 130 added opportunities on balls in play, while the OF may be seen to be benefitting from a fly ball inducing staff, to the possible cost to infielder opportunity.

    Next, taking ONLY the Hits portion of WHIP and subtracting HRs, I wanted to compared batted ball hits/ip compared to League. The pitching staff comes up at WHIP of .997 of League average per IP over 1,435.1 innings = 4.31 WHIP elements, about the equivalent of -1.51 runs allowed over the season. This pitching/defense collusion is almost perfectly average.

    I then applied my metric to each position for the AL in 1990, to get a defensive composite profile for each team.
    For brevity here, I’ll just post Baltimore position performance, all innings played vs. LG average at that position, and defense runs compared to LG.

    Pos. Rating…………….LG Rating……….DR

    C……… .928………………...... .936……… … - 7.6
    1B…… .975………………….. .967………… + 1.8
    2B…… .976………………….. .965………… + 3.8
    SS…… .946………………….. .958…………. - 4.5
    OF…… .991…………………. .959…………..31.0

    Team……………………………………………... +24.5*****

    To Summarize:
    1. The 1990 Baltimore Orioles finished with a losing record, 76-85.
    2. At bat, the O’s scored fewer runs than League average, 669 – 696, -27 BR.
    3. Pitching, seen from hits on BBIP, is virtually League neutral. + 1.6 Runs better than League average.
    4. The Orioles team defense is at +24.5 DR better than LG team average.
    5. The aggregate of the 3 inputs here [-27; +1.6; +24.5] = -0.9 runs, suggesting a team perhaps 1 loss below .500 [80-82]. The O’s ended up 76-85.

    I realize there are those who would cry “FOUL!” .. because one should NEVER expect such data sums to be precisely predictive. I acknowledge that; but metrics should be held to collective accountability to some degree. The reality suggest some resonance with the inputs.

    Cal Ripken was a fine defensive SS, who quite possibly should have won Gold Glove Awards in 1983, 1984, 1985, 1989, and 1991. His only season below average was 1990. If someone wants to put this through the shredder, I can anticipate such a reaction, with a few caveats:

    1. Errors do not define the defensive excellence of play at a position. At a very busy and much-challenged position, an error count may be counter-intuitive, discounting the more critical elements of range, effectively executed.

    2. Even “opportunity” considerations, like pitching staff K’s and batted ball outs distribution [OF vs. IF] contain grey elements, largely based on fielder range on particular types of batted balls.

    3. Some defense metrics ignore pop fly PO in the IF, which is to ignore #2 nuances.

    *****Footnote added. Metric suggests that defensive contribution came from BAL OF and 2B.
    Last edited by leewileyfan; 01-13-2013 at 08:46 PM.

  7. #57
    Rather than start an entire new thread, I figured to extend this one, for the following reasons:

    1. Have all but completed the essential editing process of a book manuscript that looks to be about 300or so pages on the history and evolution of MLB defense. In that defense is the major focus, many stars commonly featured prominently in most baseball centered publications, take a back seat to lesser known players, whose primary contributions came from their gloves rather than their bats.

    2. Although I have my own fairly firm opinions on the value of the metric in its purely original form, I acknowledge the need to be sensitive to Devil's Advocate positions that the metric has a pitching staff strikeout bias that favors position players on teams with fewer recorded K's than other staffs. The argument presented is that mathematical equivalence demands adjustment be made to "level the playing field" in defensive evaluations.

    I have addressed this single topic for pehaps 200 hours over the years; and I have arrived at a very specific formula that assigns a value/K and applies it across-the-board to each player's INPUT data. I can present Ratings and Defense Runs BEFORE and AFTER adjustment.

    3. However, and this is a bit mind-blowing, I have attempted to vet the metric for comporting with actual team W-L records at each season's end [1901-2012], as an added challenge. Some sabermetric papers say that this is a fool's errand in that baseball has too many elements and factors that naturally distort any tidy calculation of batting runs +/- defense runs +/- pitching runs producing a product that conforms to records. I do not dispute the general observation. I merely want to vet this metric in as critical a light as I can.

    Questions, constructive and critical suggestions MOST welcome:

    1. Should the final draft stick to the original premise: a year-by-year specific metric evaluation of sample players from each position/season [1901-2012], with Ratings and +/- Defense Runs compared to MLB average [either as a rate of performance OR actually adjusted to reflect position playing time]. Does the latter adjustment matter IF the metric is fully expalined up front?
    Reason for Asking: Too much information may kill interest. It also makes for more work for me].

    2. Should I let the entire manuscript stand alone and forget tying defense ratings into season end W-L percentages on a team level.
    Reason for Asking: In the several seasons for which I have done complete team defensive evaluations, I find raw ratings hold up as well as K-adjusted ratings. I have also found, to my surprise, that pitching run determination, vs. League, seems better predictive of performance if I relate ER and BFP rates to League, rather than FIP. Yeah, go figure.

    Several posters have made suggestions [noted] on bios, pictures and personal snippets on players, which have been introduced.

    This latest quandary may be no quandary at all. However, I did not want to prsume what potential readers might find unnecessary [too much information]. The [/B]metric[/B] is my baby. But I have to allow that a reader might demand more.

    Suggestions? Thanks.

  8. #58
    Join Date
    Jul 2007
    Posts
    6,830
    Blog Entries
    4
    Funny you should post Lee. I was wondering, why is a good question perhaps the book in my hand, about who the worst offensive players were that were substantial starters and whether they were actually good defensively. The players in mind were Ski Melillo circa 1930s Browns and George McBride circa dead ball Senators. I don't think my thought has anything to do with yours but you may have at least an opinion on the caliber of their defense.

  9. #59
    PVNICK: I am glad you happened to have the book in your hand. I am happy to get the question.

    I have a spreadsheet that I set up some years ago, on which I try to summarize total player values: Runs Created, Defense Runs ... against MLB average. A rough, down-and-dirty average for position player run production, batting is about .11 pr PA, or around 66 RC in 600 PA.

    George McBribe is at .0712 in that department, or 43 RC in 600 PA. He is #732 of 737 player on that list; so Yep, he was pretty poor on the batting end. Of note is that run creation at bat during McBride's career ran between 70% - 90% of MLB historic average [.08 - .099]; but he was low,

    However McBride was a solid fielding SS, above average at SS defensively except for 1909 when he was average; and 1914-15, a hair below average. He was among the top 3-5 SS defensively most other seasons; and, had there been Gold Glove Awards back then, he would have been a winner in 1913. It's intersting to note that he was in the "top 25" in MVP voting in every year from 1911through 1914.

    Ski Melillo is another matter, in that his PC/PA [.0937] during a period 110%-115% above the norm, was weak but no so much as McBride. Melillo had no real power; but he got his fair share of doubles and hustling triples ... with 22 career HRs. He was an excellent bunter and fanned only 40-45 times a year, so contact was decent, too.

    Fielding, Melillo was a gem. If Gold Gloves had existed, he would have won 2B honors, ceratinly in the consecutive seasons 1929 though 1935, among such competitors as Frisch, Critz, Lind, Cuccinello, Hale, Cissell, Myer, Kerr, Gehringer and Warstler. He got MVP votes [in the top 25] in 1926, 1931 and 1932, playing for the Browns.

    Thanks for the question.

  10. #60
    Late in October I responded [post #56 above] to a poster's questions about Cal Ripken's defensive season in 1990 and by implication other contribotors to the BAL defense overall.

    Despite nifty raw numbers, Cal came up in - DR territory agins=t league average; but the team fared well. I assumed then the big contribution can from the OF.

    Editing tonight, I dig into the 1990 season again and BAL OF collectively were at .993 defense rating [my metric] vs. LG average of .959, which, when figured as a collective of three positios, accounting for almost 40% of team "outs" was worth +39 defense runs.

    Interestingly, with the collective OF performance stands out, individuals do not stand out so well since so many players plied their trade in the BAL OF that season. However, leaders were:

    LF: Brady Anderson and Joe Orsulak

    CF: Mike Devereaux and Steve Finley

    RF: Joe Orsulak

    That's what helpd make defense metrics so challenging, and often, fun to explore.

Page 3 of 3 FirstFirst 123

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •