169 thoughts on “Circle of Greats 1974 Balloting Part 3”

1. Bob Eno (epm) February 18, 2019 at 8:31 am

   Here are charts for player stats this round. (I don’t see strong candidates to add among the newly eligible group.)

   For this round I’ve modified the charts, eliminating the WAR/Yr. category while changing WAR/9IP to WAR/162IP, and changing WAR/G to WAR/500PA. Those new categories essentially show WAR per standard “qualifying season” for pitchers and hitters. pWAR denotes bWAR for pitching only.

   These stats provide: one compilation measure (WAR, including, for hitters, fWAR, and for pitchers Total bWAR, reflecting hitting and defense); two measures of peak; two career rate stats; and one metric for comparative longevity. The goal is to provide a brief all-round image at a glance; different people assign different weights to these categories.

   Career length is relative to the briefest pitching and batting careers on the main ballot in terms of IP and PA. Hence: 1.0 = (Kevin Brown) 3256.1 IP / (Dick Allen) 7315 PA.

   Main Ballot Candidates
   Pitchers
   pWAR (Tot bWAR)…Peak5..Top5…WAR/162IP…..ERA+…Career length
   68.5 (68.3)……………37.0…..37.0….….3.4…………127……..1.0…………Kevin Brown
   67.2 (71.6)……………24.2…..29.0……..2.9…………118……..1.3…………Ted Lyons
   68.7 (67.4)……………22.5…..27.3….….2.1…………108……..1.6…………Don Sutton
   66.1 (66.7)……………28.7…..34.7……..3.1…………114……..1.2…………Luis Tiant

   Position Players
   WAR(fWAR)…Pk5……Top5….WAR/500PA….OPS+…Career length
   58.7 (61.3)…..31.5……36.7……….4.0…….……156………1.0………….…Dick Allen
   63.6 (57.6)..…31.6……32.7……….3.3………….111………1.3………….…Richie Ashburn
   62.8 (54.7)..…33.0……34.0……….3.8…….……116………1.2…………….Ken Boyer
   75.2 (77.5)..…22.6……29.8……….3.6………….110………1.4………….…Bill Dahlen
   64.4 (59.5)…..32.4……33.7……….3.0……….…119………1.5……….……Andre Dawson
   68.0 (65.7)…..28.7……32.2……….3.3………….110………1.4…………….Graig Nettles
   69.2 (66.3)..…28.7……29.9……….3.6…….……154………1.3…………….Manny Ramirez
   50.1 (54.2)..…23.3……26.4….……2.6………….118………1.4…………….Ted Simmons
   70.2 (62.4)..…28.6……31.3……….3.7…….……105………1.3…………….Bobby Wallace*
   *Wallace’s total bWAR (incl. pWAR) is 76.3.

   Secondary Ballot Candidates
   Pitchers
   pWAR (Tot bWAR)…Peak5..Top5……WAR/162IP….ERA+…Career length
   65.3 (60.3)……………40.2….40.2………..3.4………..127………0.9………Stan Coveleski
   23.0 (16.6)……………17.1….17.6………..1.8………..103………0.6………R.A. Dickey
   62.0 (62.4)……………27.9….29.6………..3.1………..116………1.0………Dennis Eckersley*
   60.9 (60.8)……………20.3….28.4………..3.0.……….117………1.0………Andy Pettitte
   68.2 (70.1)……………31.0….32.8………..3.1………..114………1.1………Rick Reuschel
   *Eckersley pitched 25% of his innings as a reliever, which makes his stats less comparable to others.

   Position Players
   WAR(fWAR)…Pk5……Top5….WAR/500PA…OPS+…Career length
   60.0 (59.6)……29.7……31.1………3.0…………128………..1.4….……….Bobby Abreu
   61.4 (55.0)……37.4……37.4………3.2…………133………..1.3…………..Todd Helton
   21.3 (20.8)……18.3……20.1………3.7…………125…….….0.4…………..Monte Irvin*
   50.5 (50.8)……28.4……30.2………3.3…………131………..1.1…………..Minnie Miñoso
   65.5 (62.1)……27.2……29.5………3.5…………104………..1.3…………..Willie Randolph
   64.6 (64.6)……26.9……29.0………4.0…………137………..1.1…………..Reggie Smith
   *Due to segregation, Irvin reached the Majors at age 30.

   Reply ↓
2. Bob Eno (epm) February 18, 2019 at 8:40 am

   As HHSers know, I’m an advocate of Bill Dahlen and Bobby Wallace for the CoG. It would be tedious for everyone if I made arguments for them evry round, but this round I’d like to make some, and this comment is more or less background, speaking to the issue of baseball in the period 1893-1900, which has been an issue in their cases.

   One of the main objections to Dahlen and Wallace is that although they qualify for the CoG because of their play after 1900, since they played a portion of their careers prior to 1901 (Dahlen 54%; Wallace 28%), those 1890s years should be counted for less or not at all because MLB baseball wasn’t yet the game we know. In this comment, I’d like to clear some ground for Dahlen and Wallace, explaining why MLB baseball from 1893-1900 should be considered the “game we know.” The core argument is a quick summary of more detailed research I posted on HHS last spring, with a few added issues not relevant to that series. This is just my view, but during the Redemption Round discussions it became clear that a few other HHS posters also see 1893-1900 as part of the same era as early 20th century ball, and Doug has raised the possibility of a rule change to reflect this view, so this comment relates to that issue as well.

   Why is 1893 a dividing line in the MLB game? In 1892, MLB consolidated into one National League with 12 teams, absorbing four franchises from the defunct American Association. The first pennant of this new era was won by Boston. The Boston manager, Frank Selee, elaborating on unsustained practices by forerunners, trained his team in a new style of play, emphasizing innovative batting strategies and fielding plays, instead of a more traditional, static approach. His “small ball” approach paid off. The same year, Ned Hanlon took over as manager of the last place Baltimore team and began rebuilding it into a team that could play small ball.

   In 1893, the diamond was reconfigured to modern dimensions: the pitcher’s mound was moved back to encourage more hitting. Selee’s Boston team continued its success, but more surprisingly, Hanlon’s Orioles rose to the middle of the standings. In 1894, Hanlon’s team won the pennant, a remarkable rise in a league with a history of stability. The success of Selee and Hanlon in 1892-94 spurred other teams to adopt a small ball approach – the trend was known as “scientific baseball,” and also “inside baseball” – and it remained the dominant style of play until the lively ball introduced in 1911 combined with Babe Ruth’s demonstrations of its potential in 1919-20. Taking the year the modern diamond was configured as a starting point, the years 1893-1918 comprise a single era in terms of the way the game itself was played. (The high-impact role of the Old Orioles in this story parallels in some ways the impact of the Yankees on the transition from the Dead Ball to Lively Ball eras.) The changes in the game itself from 1893 on were much noted by people in baseball and sportswriters at the time.

   Here are some challenges to that basic model:

   Doesn’t 1894 undermine this story? 1894 was probably the greatest hitting year in MLB history. The mound change of 1893 had increased hitting substantially, as intended, but ’94 saw another quantum leap, which seems to undermine the idea that small-ball “scientific baseball” was taking over the mode of play. However, upon examination of the line scores of every game in the years 1893-95 (available on Retrosheet), what becomes apparent is that the 1894 surge was not the continuation of a trend begun with the move of the mound that created a designed hitting surge of 1893, but that at the end of May 1894 there was an abrupt increase in scoring that lasted until late May 1895, when hitting returned to 1893 levels with equal suddenness. My argument is that the data is most elegantly explained – by far – by reasoning that the baseballs ordered for 1894 were designed to increase hitting (like Players League balls in 1890), and did so as soon as the old 1893 balls ran out, about Decoration Day. The 1895 balls returned hitting to the “new normal” for the altered diamond, from which point pitchers continued an adjustment throttling back hitting that had emerged in late 1893 and early 1894, before the late-May explosion. This lively ball effect does not seem to have changed the way the game was played; the trend towards “scientific baseball” was gaining momentum despite the change, as demonstrated by the Orioles’ championship using those techniques in ’94. Distortions due to changes in the ball occur periodically in the 20th and 21st centuries as well.

   Doesn’t 1899 disqualify the 1890s as truly Major League?> In 1899, the NL tolerated the ownership of multiple franchises by a single ownership group. The ownerships of Baltimore & Brooklyn, Louisville & Pittsburgh, and St. Louis & Cleveland were all examples of this “Syndicate Baseball.” The owners of the first two teams shifted talent to Brooklyn (which won the pennant), but because of resistance from some Baltimore stars who refused to move, Baltimore remained competitive, though no longer dominant. The owners of Louisville & Pittsburgh did not transfer players to improve either team; that only occurred when the NL contracted to 8 teams in 1900 and Louisville was terminated. However, the St. Louis ownership plundered its Cleveland team, resulting in the scandal of the 1899 20-134 Spiders, a team with no trace of MLB quality. Syndicate Baseball was a corruption of the NL – a bush league phenomenon – but it affected only two pairs of teams in a 12-team league and for only one season. A very similar form of corruption was tolerated by the AL during the period 1955-60, where because of the personal hold of Yankee owners over the KC Athletics ownership, the two teams effectively operated as a single syndicate (as was well known at the time), sustaining Yankee championships and KC mediocrity. The long-term effects on pennant races and standings over that period was more profound than the single year of Syndicate Baseball. (A somewhat similar situation obtained for the Yankees and Red Sox in the years 1920 and after.) It would not be appropriate to delegitimize the years 1893-1900 because of a one-year anomaly in 1899, affecting a minority of teams, that resembled 20th Century cases.

   Aren’t the levels of competition and play too low in the 1890s? There are several elements to this issue. One is about the standard of play. The quality of the MLB game has improved continuously since the game was stabilized in the early 1890s. The skills of players in the decade of 1901-10 were surely better than those of the 1890s, but the same can be said of any subsequent decades. Because managers like Selee and Hanlon served as models, emphasizing physical and skill training, and advanced tactics in hitting and fielding, the 1890s saw new styles of hitting and a dramatic improvement in fielding that marked the decade off from “Early Baseball” and tied it to future decades. In this respect, a line is drawn in the early ‘90s; no such line demarcates the quality of play in following decades; it’s a continuum.

   As for competition, there are two basic dimensions to this issue. The talent pool of the 1890s was not as broad as that of following decades, and here HHS’s Dr. Doom has pointed out that the addition of the American League in 1901 was accompanied by an increase in the geographical range of scouting and talent recruitment to include areas such as the Far West and Deep South that had been little represented in the 1890s. That is true. However, if you look at the 1890 census, you will find that the main recruiting regions of baseball (the Northeast, Border States, and Upper Midwest, including states on the west bank of the Mississippi) were home to 75% of the US population. Individual cases went beyond that range: Sam Crawford was recruited out of Nebraska, Joe Corbett out of California. After 1901 the percent of (white) young men scouted surely increased, but the additions were incremental, and, of course, there were also 33% more teams to stock. In thinking of the 1890s talent pool, it is also good to recall that in those days, apart from boxing, baseball was the only sport young athletes could aspire to make their profession. Today’s pool may draw more widely, but it is also competing with many other professional and Olympic sports, diluting its strength (at another time, it would be interesting also to discuss the impact of expanded talent from Central/South America and Asia, as well as of decreased black participation rates in the US).

   The second dimension of the issue concerns segregation. It is unquestionably true that the standard of competition in 1890s baseball was lowered by its restriction to white players. This, however, is also true of each succeeding period until the era of initial integration (1947- c.1960). In this respect, nothing divides the period 1893-1900 from 1901-1947. If we want to handicap players of the pre-integration age, we should do it uniformly; there is no reason to single out 1893-1900.

   Isn’t this comment way too long? Yes. I’ll sit down. I’ll be back for more specifically on Dahlen and Wallace.

   Reply ↓
3. Voomo Zanzibar February 18, 2019 at 10:18 am

   Vote:

   Richie Ashburn
   Ted Lyons
   Manny Ramirez

   Stan Coveleski
   Dennis Eckersley
   Willie Randolph

   Reply ↓
4. Gary Bateman February 18, 2019 at 10:21 am

   #4 Chet Lemon

   Reply ↓
5. Mike L February 18, 2019 at 11:38 am

   #6 Ruben Sierra in 1989 when he was 23. Odd career. Played 20 years, got 11 of his lifetime 16.8 BWAR in two seasons. Actually had negative WAR his last 16 seasons combined.

   Reply ↓
   1. Paul E February 18, 2019 at 11:58 am

      Mike L
      I believe Rube was an “investor” in after-hours clubs – either as a too-frequent patron or venture capitalist.

      Reply ↓
6. Dr. Doom February 18, 2019 at 12:37 pm

   Doug:

   1. I’m just checking for the dividing line: is there any change it’s Bruce Sutter? If it’s not, it must be someone earlier, because he definitely did it and retired much earlier than most of the “modern” relievers (1988).

   2. This is not an answer to the question, but as a point of interest, Lenny Harris has 13 seasons of 100 < PAs < 400. He does not qualify, as he does have two additional seasons of 400+ PAs.

   3. Nolan Ryan is the easy one, as he famously went 8-16 with a league-leading ERA in 1987. I also found Joe Magrane in 1985, with a 2.18 ERA and a 5-9 record. Dave Koslo led the NL in 1949 while posting a 14-19 record. Couldn't find the fourth… but honorable mention to Dolf Luque who had a .471 winning percentage in his second ERA-leading season.

   Reply ↓
   1. Doug February 18, 2019 at 2:51 pm

      Sutter is correct, as are the three NL pitchers you named.

      Reply ↓
7. Gary Bateman February 18, 2019 at 12:41 pm

   Main ballot: Dahlen, Ashburn, Boyer
   Secondary: Minoso, Eckersley, Helton

   Reply ↓
8. Richard Chester February 18, 2019 at 12:44 pm

   Answer to question #2: Jim Dwyer had 15 seasons of more than 99 PA and none with more than 399. His max PA total was 292.

   Reply ↓
9. koma February 18, 2019 at 1:19 pm

   #1 Bruce Sutter retired 1988 with 300 of his 512 GF as a save
   #2 Jim Dwyer had 15 seasons with PA>100 and no 400 PA season in his 18 year career
   #3 Joe Magrane in 1988, !!!Nolan Ryan 1987!!!, Stu Miller in 1958 and in Dave Koslo
   in 1949 (plus early Rube Waddell in 1900, Theodore Breitenstein 1893 and Henry Boyle in 1886)
   #5 Johnny Bench 1972 and 1973
   #6 Ruben Sierra in 1989 at age 23
   #7 Ross Grimsley with 92
   #8 Eric Hinske in 2007
   #9 Josh Hamilton in 2010

   main ballot:
   Manny Ramirez
   Bengie Molina
   Hideki Matsui

   secondary ballot:
   Andy Pettitte
   R.A. Dickey
   Dennis Eckersley

   Reply ↓
   1. Doug February 18, 2019 at 2:52 pm

      All correct, koma.

      Reply ↓
10. Bruce Gilbert February 18, 2019 at 1:20 pm

    This is difficult. There are four that I think are a little above the others, but there are only three votes available. So, Don Sutton, Bill Dahlen and Ted Simmons. In the secondary ballot I’ll go with Minnie Minoso, Rick Reuschel and Andy Pettitte. I will post a specific comment on Sutton later. Can’t right now. Thanks.

    Reply ↓
11. Richard Chester February 18, 2019 at 1:44 pm

    I’ll offer a semi-guess for question #11; Glenn Wright with 316 games.

    Reply ↓
    1. Doug February 18, 2019 at 2:52 pm

       Wright is the one.

       Reply ↓
       1. Richard Chester February 18, 2019 at 3:26 pm

          A comment I made not too long ago is what prompted me to guess Wright. With some help from you I mentioned the names of 4 players who played every inning of their team’s games in their debut season and Wright was one of them. I looked at his BR page and saw that he did it for his first 2 seasons, as did Al Simmons. And it looks like Simmons consecutive game streak at the start of his career reached 393 games.

          Reply ↓
12. CursedClevelander February 18, 2019 at 4:26 pm

    For the Matt Morris question, the town is Whittier, CA. Home to both Mark Kotsay and Nomar Garciaparra. Of those two, Garciaparra hit over .400 against Morris and Kotsay was closer to .150, so the answer seems to be Nomar.

    Reply ↓
    1. Doug Post authorFebruary 18, 2019 at 11:28 pm

       It is Nomar.

       Morris has good company; Nomar also batted .400 in 20+ AB against Dwight Gooden, Kenny Rogers, Jamie Moyer, Ken Hill and others.

       Reply ↓
13. Dr. Doom February 18, 2019 at 4:32 pm

    I want to talk about Monte Irvin. I’m not sure if I’m voting for him or not yet, so don’t read this post as advocacy for or against him; it’s just part of my thought process.

    As a basis for comparison, I went crude. I looked at WAR ages 30-35 among position players on either ballot who debuted in the Live Ball Era. I stopped at 35, which mildly shortchanges Irvin, because Dick Allen and Richie Ashburn didn’t play at age-36 or age-37. Here’s how they stack up (I’m putting WAR in their 20s in parentheses):

    Graig Nettles – 27.9 (29.0)
    Minnie Minoso – 26.9 (25.6)
    Todd Helton – 24.0 (35.0)
    Bobby Abreu – 22.3 (35.4)
    Reggie Smith – 22.2 (39.7)
    Willie Randolph – 21.9 (38.8)
    Monte Irvin – 18.5
    Richie Ashburn – 17.9 (46.0)
    Andre Dawson – 17.6 (43.4)
    Dick Allen – 15.8 (42.9)

    Again, I remind you: this post is not advocating for nor against Irvin; it’s just a question of what you need to assume about his value to make him like the others we’re considering for induction.

    My thoughts would be as follows. Taken as it is, for Irvin to belong in this group of players (average WAR through age-35 = 59.1 total WAR), his 20s would have had to have been worth 40+ WAR in order to get him there; to exceed everyone in this group in total age-35 WAR, he would need 45.5 WAR in his twenties, which would be second only to Ashburn among them.

    Is that fair? I don’t know. Would it be crazy to induct him for the performance we DO know about? No, not really. In the end, if we were doing a project with 100 more players (aka the size of the FULL Hall in Cooperstown), Irvin is such a no-brainer that it’s not even worth discussing – but, for me at least, so are the rest of these guys. Given the scope of the project at hand, though, I think he matches up nicely with these borderline guys to whom we can’t commit, yet whom we can’t quit. Still just not sure which side of the fence he lands on, though.

    Reply ↓
    1. Bob Eno (epm) February 18, 2019 at 9:04 pm

       Doom, First, I think it’s a good idea to subject Irvin’s candidacy to objective tests like this. I expect everyone is sympathetic to his case because of the effect on him of segregation, but when we give him a “bonus” we want the bonus to reflect real expectations, not empathy.

       I think, however, that there is a problem with your approach. Age 30 was Irvin’s rookie year: an extended cup of coffee (36 G, 93 PA) with lots of PH appearances. Like many rookies, he did not shine — the next year his talents were visible at the MLB level. Everyone else on the list was in (or past) their MLB primes at age 30; their rookie adjustments were far behind them. So I think the more accurate way to get a reading is to measure age 31-35 seasons. If you do, the list looks like this:

       Graig Nettles – 23.3 (33.6)
       Minnie Minoso – 20.7 (31.8)
       Reggie Smith – 19.6 (42.3)
       Monte Irvin – 18.3
       Willie Randolph – 17.9 (42.8)
       Todd Helton – 15.7 (43.3)
       Bobby Abreu – 15.7 (42.0)
       Andre Dawson – 15.4 (45.6)
       Richie Ashburn – 12.4 (51.5)
       Dick Allen – 7.2 (52.5)

       Irvin’s breakout year in the Negro Leagues was at age 21. So I think the question would be whether a player who compiles 18.3 WAR over his age 31-35 seasons can be expected to generate about 40.6 WAR ages 21-30.

       It might be good to note too that Irvin’s age 33 season was limited to 46 G because he broke his ankle sliding into third. Most players lose time to a major injury at some point, but in Irvin’s abbreviated career, the statistical impact is enlarged. If we look at overall career WAR rates (WAR/500PA), Irvin’s 3.7 is behind only Smith (4.0), Allen (4.0), and Boyer (3.8) among position players on both ballots, despite the fact that his prime ages, when his productivity might be expected to be highest, were lost to segregation.

       Reply ↓
       1. Dr. Doom February 19, 2019 at 12:08 pm

          OK…

          First, Bob, thank you for making that point about the age-30 season. I really didn’t know (or bother researching) why his games-played total was so low, and chalked it up to injury.

          So, given the new information, I came up with an idea. The average of all the position players (Ken Boyer included, since I somehow forgot him) averaged 16.9 WAR from ages 31-35 and averaged 42.5 WAR through age-30. That’s a factor of 2.5. We know that Irvin had 18.3 WAR from ages 31-35. At a factor of 2.5 before that, we’d get 64.3 WAR. We also know that Irvin totaled 2.8 WAR after age-35. That means we can make an estimate of 67.1 WAR. That’s more WAR than anyone on the Secondary Ballot, other than Rick Reuschel (at 68.3, so we’re not talking a world of difference here). I think that makes Irvin a “yes” on my secondary ballot.

          I’m not sure what I’m going to do with him if (once?) he gets on the main ballot. But we have at least one week to figure that out, now don’t we?

          Reply ↓
          1. Bob Eno (epm) February 20, 2019 at 8:09 am

             This seems a reasonable approach, Doom. Perhaps a little in Irvin’s favor. I did a search of all players who came up by age 21 and whose age 31-35 seasons were within one point of Irvin’s WAR total (eight players; I ruled out any with military service). Their age 21-30 WAR average was greater by a factor of 2.35.

             Reply ↓
             1. mosc February 26, 2019 at 11:20 am

                You guys are making perfect sense to me. I support Irvin as a candidate but I agree it’s very hard to project. Irvin played just enough to convince you of the talent but there’s so much that was taken away it’s hard to really see the career that would have been. I think the negro league stats we DO have though paint a picture of one of the all-time great hitters during his prime. Negro league stats are not as good but it’s not like there were lots of others with his production. We do have SOME record of that time from Irvin, even if it’s flawed, and it more than lines up with the kinds of projections you guys are doing. Just saying, it’s another piece of evidence to consider.
14. Bob Eno (epm) February 18, 2019 at 8:34 pm

    As threatened, I want to make some arguments in favor of Dahlen and Wallace. Because I’ve been doing this so long, I have enough points to put absolutely everyone to sleep, so I’m going to pick out only two for this comment, one very simplistic, the other more interesting (I hope). I’m relying on my previous post to serve as background. I’m not going to anticipate in this comment counter-arguments about whether or how to credit Dahlen and Wallace’s play prior to 1901, since I expect the other post will attract those.

    (1) The simplistic argument is about WAR. No one number can crunch all the facts about a player, but the closest we have is bWAR. Among the CoG candidates this round, Dahlen and Wallace are by far the leaders, especially if you count Wallace’s brief career as a successful starting pitcher (Dahlen 75.2; Wallace 70.2/76.3). By that measure, Dahlen is 6.0 WAR beyond the next position player behind Wallace (Ramirez); Wallace’s margin is narrower, but if you count pitching and hitting, his total is slightly higher than Dahlen’s. Only Ted Lyons, if you count his batting, breaks the 70 WAR barrier (71.6). (Wallace’s total falls to 62.4 if you use fWAR, but the only non-PED position player to exceed that total is Nettles, at 65.7. As for Dahlen, fWAR increases his lead over Manny to 11+ points, or almost 17%, with everyone else trailing.) Dahlen and Wallace also have among the highest bWAR rate stats among current candidates, in careers longer than the two who surpass them.

    I don’t think the WAR argument is sufficient at all, but I think it establishes that in Dahlen and Wallace we are dealing with players who are well above the CoG threshold, and I believe it places a burden on those who want to argue against making them priority selections.

    (2) The more interesting argument concerns fielding. Although Dahlen had a couple of years as a bona fide slugger early in his career (OPS+ seasons of 156 and 138), overall he was merely a pretty good hitter for a shortstop (OPS+ 110), and Wallace the same (OPS+ 105). Both players’ great strength came from fielding, in which they were far above average for their time. I think it is generally true that fielding accomplishments make less of an impression than hitting and pitching, especially when we must rely on stats and can’t actually view players’ fielding skills on video. We all understand why Brooks Robinson and Graig Nettles might be celebrated despite their ordinary hitting because we’ve seen their acrobatics. For Dahlen and Wallace it’s easier to bypass their cases. But Dahlen and Wallace weren’t just good fielders. As the premier shortstops of their day, they led a revolution in the game. The significance of their fielding accomplishments has to be understood in the context of the way the game was changing when they played.

    In the early 1890s the typical NL error rate (number of chances resulting in errors) was slightly over 7%. However, the move to “scientific baseball” that began at that time placed heavy focus on maximizing fielding performance, both through the development of new plays (e.g., throwing to a cut-off man) and intensive training to improve mechanics. By the close of the careers of Dahlen and Wallace, error rates had been cut nearly to 4%.

    On the other side of the coin, Defensive Efficiency (the percent of BiP that become outs) was on the rise. On the eve of the mound change, 1892, 67% of BiP resulted in outs. Unsurprisingly, the first year of the new mound distance, which we can see as a new baseline, it fell to 65%, and then again in 1894, a lively ball year (well, according to my research), it fell to 62% (a fall of 2% also occurred with the lively ball of 1911). From that point, it rises steadily through the remainder of the 1890s and the decade of the ‘00s, reaching a peak of almost 71% in 1908. (For comparison, Defensive Efficiency in 2018 was 69%.)

    In other words, the careers of Dahlen and Wallace as shortstops spanned a fifteen-year transformation of the fielding game from an error-ridden, inefficient state, to an efficient state, with about 35-40% fewer misplays.* Over the course of their careers, no aspect of the game saw as much refinement as fielding, and improved fielding was the area where greatest leverage for overall improvement could be found. Wallace and Dahlen are 11th and 12th all-time on the dWAR leaderboard (both are within one point of Omar Vizquel). The closest to them among players who broke in before 1900 is George Davis, in 24th place. They were the leaders in the fielding revolution that, along with the introduction of the approach of scientific baseball, rapidly transformed the Major League game into its modern form. (Wallace is particularly known for the invention of the modern infielder’s motion of fielding and throwing without stopping at a set position in between.) And they were above average hitters too.

    I don’t see other current CoG candidates who were transformative figures like Dahlen and Wallace, and, of course, none matches them on the gross quality measure of bWAR. I think that in addition to specific reservations about the 1890s, they are handicapped here simply by their remoteness in time. That’s what happened to them when the BBWAA voted, and since our task is to try to do the writers’ task better than they did, I think the cases for Dahlen and Wallace are very strong.

    *The improvement was not primarily caused by larger gloves. In 1894, the league set low limits for the sizes of gloves worn by position players other than catcher and first baseman, and at the end of Dahlen and Wallace’s careers, gloves looked much as they had fifteen years before.

    Reply ↓
    1. no statistician but February 19, 2019 at 2:38 pm

       Bob:

       I think you’re shortchanging Wallace a little as a hitter. I picked his 1898 season out to look at because he had 108 RBIs, and with a little more investigation I discovered that from age 23 through age 34, Wallace’s prime seasons as a position player, he led his teams in RBIs seven times and finished second five times.There’s no data on his position in the batting order available for those years, but whatever position he was in he was doing a very good job, considering the era and teams he played on, of getting runs across the plate. In eight of those years he was in the league’s top ten. Eight times in the league’s top ten in doubles, four times in triples.

       Dahlen is a little harder to figure because the majority of his better years at the plate came during the batting boom, which gives his career stats a boost that Wallace’s don’t share. In general, too, he played for teams with more offensive prowess than Wallace’s, another factor that clouds the issue a little. It’s safe to say, though, that he was fairly productive at the plate most of the time. Wallaces’s early prime seasons and Dahlen’s late prime seasons overlap, and there’s not much difference between them, which is to say that their managers had no reson to try to hide them in the lineup, even if they weren’t the biggest guns. Match that fact with their fielding abilities, and what do you have? Brooks Robinson moved thirty feet to the left.

       Reply ↓
       1. Bob Eno (epm) February 19, 2019 at 5:41 pm

          A very interesting point about Wallace, nsb. I hadn’t noted it because I pay too little attention to RBI, it being difficult to assess how the credit should be distributed. If you compare the primes of Wallace and Dahlen, what jumps out to me is that although they seem roughly comparable in OPS+ and XBH, Wallace is racking up RBI while Dahlen is racking up runs. (Though, oddly, only Dahlen actually wins an RBI title, and that comes much later, with a low total.) My thought is it’s likely a matter of batting order. Wallace likely had Jesse Burkett ahead of him, with his 200 hits and 60-70 BB year after year. But Wallace still had to deliver, and he did, so what you’ve spotted does improve his profile.

          Frankly, I’ve gone back and forth so many times trying to pick a priority between Wallace and Dahlen that I’ve given up. For every plus (e.g. Wallace’s RBIs) I always seem to find a balancing factor (e.g. Dahlen’s SB). If I had to pick just one today I’d go through the exercise again, with no idea in advance which one would edge out the other this time through. I fully agree with you that we’re seeing Brooks-like profiles here, and part of the problem these guys have may be precisely because there were two of them at the same time. If either one were the only Dahlen/Wallace, I suspect he’d have been in the CoG a while ago.

          Reply ↓
    2. bells February 22, 2019 at 7:10 pm

       That’s an intriguing angle, Bob, thanks for bringing it to light. I haven’t really thought about it in that way, but it seems to me a pretty good argument that in a decade where small ball and deadball tactics were developing, the two premier players at the key defensive position played a pivotal role in the development of baseball as a whole. I’ve already been voting for them (or at least one of them), but this is a good articulation of this point so I wanted to acknowledge that.

       Reply ↓
       1. Bob Eno (epm) February 23, 2019 at 12:25 am

          Thanks for the kind words, bells!

          Reply ↓
15. opal611 February 18, 2019 at 10:08 pm

    For the 1974 Part 3 election, I’m voting for:

    -Manny Ramirez
    -Don Sutton
    -Andre Dawson

    Other top candidates I considered highly (and/or will consider in future rounds):
    -Tiant
    -Brown
    -Boyer
    -Ashburn
    -Nettles
    -Allen
    -Wallace
    -Dahlen
    -Lyons

    Thanks!

    Reply ↓
16. opal611 February 18, 2019 at 10:25 pm

    For the Secondary Ballot, I’m voting for:
    -Dennis Eckersley
    -Todd Helton
    -Willie Randolph

    Thanks!

    Reply ↓
17. Dr. Doom February 19, 2019 at 12:15 pm

    On the main ballot, I’m going with:

    Kevin Brown
    Dick Allen
    Graig Nettles

    On the secondary, I’m going to say:
    Monte Irvin
    Rick Reuschel
    Todd Helton

    This is getting tough. There are five good/amazing candidates – A-Rod, Scott Rolen, Vlad Guerrero, David Ortiz, Tim Hudson, plus even Torii Hunter. Those are all decent to fabulous candidates, and we’re probably only looking at one, maybe two rounds next year. So I don’t know how much traction any of these guys even CAN get if they don’t get in now. So good luck to all the candidates right now!

    Reply ↓
    1. Michael Sullivan February 19, 2019 at 3:43 pm

       A-Rod seems a no-doubter, assuming we maintain our prior attitude towards players who use PEDs that players who blow away the COG standards get in no matter what.

       Rolen looks like an excellent COG candidate, better than anybody on our current ballot statistically and a pretty easy selection by my lights, but similar to guys who are in but took a bit of convincing back in the day, like Larry Walker and Lou Whitaker, Ron Santo and Bobby Grich. All *obvious* hall guys, but not so far above the COG borderline that everyone was excited.

       Vlad and Tim Hudson are reasonable candidates, but I think they are in line with those we’re considering weakly or who got some votes in the redemption round. They are both below my “Circle of Stats” line. Vlad will probably get some support, I doubt Hudson will.

       Ortiz and Hunter were great players, but I think they are pretty far from COG level. They don’t even make the Hall of Stats, and I’m not really sure they belong in the big Hall, let alone the COG.

       But it’s definitely a bigger year than 74. Hopefully we get more than two slots next year. And *really* hope we get the two, otherwise even Rolen will have to wait another year.

       It’s fun to look ahead to see who’s coming up!

       76, only Lance Berkman is really worth any discussion, and I don’t think he compares that well with our holdovers and next set of redemption candidates.

       77 will be another strong year with Roy Halladay, Carlos Beltran and Andruw Jones as pretty strong candidates, and Roy Oswalt who might get some love too. Halladay seems like the only maybe no-doubter.

       78 brings us Chase Utley and a bunch of fun to remember names for the Hall of Very good who might get a homer vote. Jimmy Rollins, Aramis Ramirez, Victor Martinez for the offense. And three pitchers who looked like hall of famers in their best years but weren’t good enough for long enough: Barry Zito, John Lackey and Cliff Lee.

       Reply ↓
       1. mosc February 26, 2019 at 11:23 am

          I’m not voting for Rodriguez. Steroid use was earlier in his career, clearly covered most of it, and most damning he effectively recruited players for his drug dealer. He is disqualified to me.

          Reply ↓
18. Paul E February 19, 2019 at 3:40 pm

    Don Newcombe has passed.

    Reply ↓
    1. no statistician but February 19, 2019 at 5:42 pm

       Newcombe:

       Some weird things about his career:

       He was remarkable steady as a pitcher: his home/road and first half-second/half stats are very close; vs. lefties and righties fairly close. He pitched well every month of the season; against 6 of the opposing NL teams he won 21, 21, 23, 22, 22, and 23 games lifetime. The Braves were his only weakness, as a 12-16 record attests.

       Minus his Braves record he finished an excellent 45-33 lifetime against winning teams (.577), belying to a degree, his reputation for funking it in big games. As to that, it was his poor showing in the 1955 and ’56 World Series that cast a backward shadow over the close games he started that the Dodgers lost in earlier seasons, usually through no fault of his. In 1956, after he racked up 27 wins for a team that barely won the pennant, Yogi Berra almost single-handedly burst the balloon with three clutch home runs against him in the Series, and Newk was never the same again, although his 1959 season for the Reds was very good, a kind of swan song.

       Would his career have had a different outcome had he not suffered a cruelly late call up for military service that forced him to miss his age 26 and 27 seasons? Probably.

       His SABR bio stresses his heavy drinking, starting at an early age, and reading between the lines, it’s almost impossible not to conclude that he was a very insecure and sensitive guy, a fact at odds with his size, tough expression, and general manner.

       The best hitting pitcher of his era by far.

       Reply ↓
    2. Bob Eno (epm) February 19, 2019 at 6:06 pm

       He was a mensch.

       Reply ↓
19. Bob Eno (epm) February 19, 2019 at 6:51 pm

    I was a supporter of Ted Lyons while he was on the Secondary Ballot — and, years ago, when he was a CoG candidate — but I’ve been rusty on the arguments in his favor. Because I’m now using my Main Ballot vote to support both Dahlen and Wallace, it leaves me frustratingly little room to try to make sure I’m adequately considering baseball after 1910 (!), and I’ve also thought that if I were to vote for Lyons alongside my turn-of-that-century guys, I’d appear to be nothing but a nostalgist (for eras I can’t even regret passing because I was born after them).

    Nevertheless, upon review, I do find the case for Lyons compelling. His pWAR figures are comparable to the other pitchers on the ballot, and he actually has an edge when hitting is included — all those are with margin of error. So what sets Lyons apart?

    What sets him apart are his three late-career War years. As many will recall, Lyons became a very successful pitcher in his late 30s when he was converted (against his will) to being a “Sunday pitcher,” starting only 20-23 times a season. Over the four seasons prior to the War (ages 38-41), Lyons pitched 20-22 starts and won 12-14 of them each year. His ERA+ varied, but averaged almost 150. His last pre-War year he won the ERA and ERA+ titles (2.10; 171) and completed all of his twenty starts (for 4.7 WAR).

    Then Lyons encounters three War years and rejoins his perpetual team, the White Sox, in 1946 at the age of 45. His career ends after only 5 starts, with a 1-4 W-L record; however, his record is not the reason his career ends. In fact, despite the W-L figures, Lyons was pitching very well. He completed all five starts and his ERA+ was precisely at the average of his four last pre-War years: 148 (his ERA was only 2.36). An example of his adventures that year was his initial game: a 1-2 loss. Lyons entered the 9th with a 2-hit shutout, but he lost on a two-out single, with the third potential out a ROE. Such was his short season, his losses were 1-2, 2-4, 1-4, 3-4 (six of those runs were unearned). The team was not all that bad: 74-80 over the year. But when Lyons quit, they were 10-20, and their improvement was largely due to the fact that Lyons quit in order to become their manager. His pitching W-L record had nothing to do with it.

    In fact, Lyons at 45 was pitching just like Lyons at 38-41, and the point is that the default expectation is that over the War years he would have continued at that rate, accumulating in the range of 3.5 WAR per year (actually more). If we imagine that “bonus”, Lyons’ WAR moves up to ~78, well past all other CoG candidates, and among position players only Dahlen is close. Moreover, had Lyons continued to pitch to form he would almost surely have compiled 300 wins and, despite his age, further lowered his career ERA and raised his ERA+.

    Lyons’ lifetime W-L record of 260-230 .531 is not overawing (I remember thinking it was mediocre when I was a kid). But he played for a team that went .460 during his career — he was way above his team in game results. His ERA+ is behind Kevin Brown’s (127 to 118), but there is that problem with Brown, and Lyons in a bit better than Luis and substantially above Sutton (114 and 108). His WAR accumulation rate is a shade behind Tiant, but, again, far ahead of Sutton (and Brown’s problems apply here as well, although his is higher).

    To be clear: I think Tiant, Sutton, and Brown were superior to Lyons in their pitching. This is because they all were 2-3 generations later than Lyon, and, like everything in baseball, pitching was far more advanced in their day. But our task is to judge the greatest players in light of the game within which they performed, and among the Main Ballot pitchers, I think Lyons was by a significant margin the best in that regard. It’s harder to compare Lyons with CoG position player candidates, but his Total bWAR exceeds all except Dahlen among those players, and with a suitable WAR bonus, I think there would need to be arguments to justify not voting for him.

    So call me a nostalgist or say I’m just an old fart trying to chase recent players off my beautifully mown diamond lawn, but I can’t see how it would be appropriate to leave Lyons off my ballot.

    Reply ↓
    1. Richard Chester February 19, 2019 at 7:08 pm

       Ted Lyons completed 356 of his 484 career starts for a .736 percentage. That’s the highest for all pitchers whose entire careers were after 1919 (minimum of 200 starts).

       Reply ↓
20. Bob Eno (epm) February 19, 2019 at 6:56 pm

    I’m ready to vote. I think I’ve written explanations of all six players in other comments, so I’ll keep this one short.

    Primary Ballot: Dahlen, Lyons, Wallace
    Secondary Ballot: Covaleski, Irvin, Smith

    I’ll keep attending to arguments that might persuade me to change a name or two before the Tuesday deadlne for that.

    Reply ↓
21. Jeff B February 19, 2019 at 9:27 pm

    Ted Lyons, Manny Ramirez, Don Sutton for main
    Dennis Eckersley, Todd Helton, Monte Irvin for Secondary

    Reply ↓
22. Bob Eno (epm) February 19, 2019 at 11:48 pm

    Eckersley has been strong in the Secondary Ballot early voting. I didn’t vote for him, but he’s clearly a very viable choice. The two things I recall most about Eck are his outstanding years with Oakland and his balanced career as a good starter and as, perhaps, the first true one-inning closer. But I was out of the country during his early peak years and have not had a really good sense of the shape of his career. I think no one has done a detailed advocacy post for Eck, so I thought I’d take a close look and pass along what I saw, in case others don’t have a clear memory of his total career profile (and because I’m still open to changing some of my votes). This isn’t really an advocacy post, though: the stats left me of two minds.

    Eck has one of the stranger career arcs among Hall-quality pitchers. As a starter, his career start (1975-79) was stronger than I’d been aware, but then went into a prolonged period of relative mediocrity (1980-86). His sensational years with Oakland (1987-92) were followed by an equal number of dismal ones (1993-98). So his career ups and downs are truly both plural.

    Because Eck’s career is half as a closer, cumulative WAR seems a completely inadequate measure for him, as is OPS+ (where he appears surprisingly weak for someone with 12 years as a closer). The closer role forecloses high season WAR totals, no matter how good a pitcher performs the role, and OPS+ tends to overstate a good closer’s performance because it compares him with pitchers who have to nurse their arms through many more innings. But it seems to me that a WAR rate-stat might come closer to placing someone like Eck in a comparative framework with pitchers like Coveleski, Pettitte, and Reuschel. In fact, their WAR/162IP stats look pretty comparable for most of the Secondary ballot pitchers (I’ll omit Dickey): Covey, 3.4; Eck, 3.1; Pettitte 3.0; Reuschel 3.1.

    But Eck reaches 3.1 (more precisely, 3.09) almost by being four different pitchers: a good starter, a mediocre starter, a terrific closer, a really mediocre closer.

    Years…………WAR……..IP……..WAR/162IP…….WAR rate as starter/closer
    1975-79……..27.9…..1148.1……3.94
    1980-86……..17.9…..1347.2……2.15………………2.97
    1987-92……..15.5…….475.1……5.28
    1993-98……….1.4…….314.1……0.72………………3.47

    In 1975-79 alone Eck is CoG quality as a starter, but his overall quality as a starter is not as clear; certainly, seven years (1980-86) at a 2.2 rate of WAR accumulation each “qualifying season” is a long time to be underwhelming (although he’s got two or three ok seasons among them, three are in WAA negative territory). He’s CoG Inner Circle for his six great Oakland years, but a below average pitcher for the following six.

    For those who prefer to judge by a player’s peak(s), Eck is a clear priority for the Circle. It’s gong to be tighter for those who count valleys too. Eck’s arm had worn out by 1986, and his conversation to the bullpen was originally an experiment in salvaging the short burst that arm was still capable of, a tactic that paid off in spades for six years — pennants, an MVP, a CYA, terrific hair. But had he closed as a starter (so to speak), it’s unlikely his WAR would have reached CoG levels, despite his fast start. His reduced IP and, consequently, his reduced total WAR (62.0) was a necessity for continuing his career (different from the somewhat similar case of John Smoltz, who, it turned out, could actually shine in either role).

    Reply ↓
    1. Paul E February 20, 2019 at 11:36 pm

       Bob epm,
       I seem to recall Eck’s wife leaving him for CF Rick Manning and this leading to a period of depression and alcohol consumption. Perhaps this led to the initial slump (1980-86). The second downturn might be just related to advanced years?

       Reply ↓
       1. Bob Eno (epm) February 21, 2019 at 12:54 am

          Paul, The alcoholism and later arm blowout are in his SABR bio (though not the adultery, which Wikipedia notes). I didn’t mean to suggest there was anything mysterious about his ups and downs, just that his roller coaster was steeper than most and with a rare double-hump.

          By the way, the “epm” in my handle is just to identify me as the guy who was “e pluribus munu” for about seven years on the site. Now that I have thrown caution to the winds and put my ancestors’ reputations at risk, I’ll answer to Bob solo.

          Reply ↓
          1. Michael Sullivan February 21, 2019 at 1:50 am

             Any relation to Brian?

             Reply ↓
             1. Bob Eno (epm) February 21, 2019 at 8:14 am

                Hmmm. Here too?

                My longwindedness fades into the background like ambient music, as generations of sleeping students (and perhaps HHS readers) can attest, but, no, otherwise no relation whatever.
             2. Michael Sullivan February 21, 2019 at 4:29 pm

                Didn’t mean to touch a nerve, just something about your mention of ancestors and family name that brought to mind that you share a last name with someone famous. I’m not even a fan, except that he did a couple collaborations with people I like a lot.

                Anyway, I take it you get that a lot, sorry.

                “Why should I change my name? He’s the one who SUCKS!”
             3. Bob Eno (epm) February 21, 2019 at 4:49 pm

                No worries, Michael, and no nerve touched. I do get it a lot, but I assume he’s tired of being asked if he’s related to me. (Not all assumptions are created equal.) I wish I liked his music more. I have only one word for a middle name; he’s got nine, so he’s clearly a very brilliant guy.
          2. Paul E February 21, 2019 at 1:58 pm

             Bob Eno,
             Supposedly, he and Manning were best friends. There were another 50+ wins as a starter in there if of sound mind and sober. I guess that kind of stuff builds character……eventually

             Reply ↓
             1. Bob Eno (epm) February 21, 2019 at 2:56 pm

                An interesting thought experiment, Paul. So that’s another route to 197 wins. But what are the odds that if you run history that second time Eck still winds up as the first closer?
             2. Paul E February 21, 2019 at 4:46 pm

                If he’s injury free, he’s another Sutton, 16-20 wins/year and he never closes and we’re spared the Tony LaRussa experiment – 6 inning starter, 7th inning guy, set-up man (Honeycutt), and closer (Eck). (One year in Oakland, Bob Welch won 27 games and a Cy Young with 2 CG and 6.8 IP/GS. He was 7th in WAR amongst the 7 vote getters and 2nd place Clemens had 3x as much WAR as Welch.)
                It’s quite possible that Eckersley might have had a Hall of Fame career as a starter. By the same token,
                his age 24 comps are a “who’s who” of the injured and disappointing:
                Gary Nolan, King Felix, Denny McLain, Vida Blue, Milt Pappas, Drysdale, Valenzuela, Larry Dierker, Pete Donahue. You would have thought some of these guys were unstoppable…..who knows? Maybe Eck is running a tool and die shop if his arm doesn’t comeback or Tony the Great doesn’t do his thing
    2. no statistician but February 21, 2019 at 12:30 am

       To me the only argument for Eckersley that might make sense is his role as a pioneer, or more specifically, a guy who made an impact on HOW the game was played. Four of his first five years as a starter were quite good, but not earthshaking, and then he dropped like a rock in terms of effectiveness, recovered somewhat for a couple of seasons with the Cubs but pitching fewer innings per start, and then, after faltering again in ’86, spent a year as an old-style reliever before becoming the transformed Closer, noted in my memory for two things: giving up few walks, and giving up a grand slam home run to a batter on crutches—slight exaggeration—a play that left his own team dead in the water while inspiring an inferior team to take charge. I can see him in the HOF, but to me there’s not enough depth to his career overall to rank him as an elite.

       And I like Eckersley.

       Reply ↓
       1. Bob Eno (epm) February 21, 2019 at 1:10 am

          It’s true that “depth” is a problem for Eckersley, but peaks are a compensating strength. 197 wins plus 390 saves (useless as that stat may be) is not a marginally substantial record. It’s impressive. Just not sure whether it’s impressive enough at this stage of the CoG.

          As for Gibson’s HR, Gibson couldn’t run (turned out he could trot), but he was an MVP hitter that year, and that’s what beat Eckersley. No shame in it, it’s to Gibby’s credit; it must have hurt like hell for Eck. After that, the other A’s beat themselves. Eck had nothing to do with that. He was fine in his other Series outing.

          Reply ↓
23. Andy February 20, 2019 at 2:47 am

    Primary:
    Kevin Brown
    Don Sutton
    Graig Nettles

    Secondary:
    Stan Coveleski
    Todd Helton
    Dennis Eckersley

    Reply ↓
24. JEV February 20, 2019 at 6:45 pm

    Primary: Brown, Ramirez, Simmons
    Secondary: Coveleski, Abreu, Eckersley

    Reply ↓
25. Hub Kid February 21, 2019 at 9:24 am

    Primary: Luis Tiant, Dick Allen, Manny Ramirez
    Secondary: Minnie Minoso, Reggie Smith, Rick Reuschel

    Somehow the CoG does not have anyone from Cuba in it, so I enjoyed topping both of my ballots with worthy Cubans. Although on our CoG borderline, Minoso has the additional impetus of being a total whiff by the BBWAA and Veterans Committees a la Ron Santo (if we’re lucky, Minoso will still get in someday, but they missed out on doing it during his lifetime, as they did with Santo).

    I’m starting to imagine how a CoG Old Timers Wing might look after reading Bob’s case(s) for 1893 cf. 1901 At the same time I still prefer modern players, et ceteris paribis. Although the various stats are not equal for the players on the ballot(s), they all seem equally difficult to rank, now we’re getting outside of the top 100 or so players of all time.

    Not that Lyons is affected by the 1893/1901 debate, but he is probably my next favourite on the main ballot, although Dahlen, Wallace and Nettles are close, too.

    Reply ↓
    1. mosc February 26, 2019 at 11:43 am

       Ron Santo explains the purpose of the VC’s, Don Baylor explains the frustration. Gah. I think somewhere on here there was a Ron Santo memorial that was full of outrage that he wasn’t in the hall at the time.

       Reply ↓
26. Josh Davis February 21, 2019 at 3:00 pm

    Has anyone else seen this article on Seamheads?
    http://seamheads.com/blog/2019/02/17/a-problem-with-war-defensive-value/

    It expresses some of my concerns with WAR, although I claim no great knowledge when it comes to the formula for WAR. I wondered how the HHS crowd would respond. If it helps frame the discussion, I’m thinking about it particularly in context of Ken Boyer vs. Graig Nettles. I’ve been revisiting my position (which has been that Boyer is the preferable third base candidate). I think Boyer is clearly the better hitter, which leaves me trying to decided how much Nettles’ defense is worth. He seems to be Boyer’s superior with the glove (though Ken at least had a good rep for defense with 5 gold gloves), but by how much, and how much should that be worth?

    Reply ↓
    1. no statistician but February 21, 2019 at 4:41 pm

       Josh:

       The only people who aren’t suspicious of dWAR figures are the people who gather them.

       What the Seamheads article questions is the huge boost given to Matt Chapman for his admittedly good fielding last year that no one else seems to find. When I looked at Chapman’s fielding data what I eventually discovered was that another third baseman, Kyle Seager, had fairly similar stats, not as good, more games played but fewer chances, but also fewer errors, more DPs, and a higher fielding %. Somehow those extra chances in the field gave Chapman 3.5 dWAR and condemned Seager to 0.0 dWAR. A little further investigation led to some other facts: that Chapman’s team, Oakland, allowed a BAbip of .271 vs Seager’s Mariners’ .296, with a groundout/ air out ratio of 1.05 to Seattle’s .99.

       Oakland’s pitching was better, and Oakland’s fielding was converting more ground balls into outs, in other words. Were all these conversions due to Chapman’s superior play? Not really, since his fellow A’s infielders all were well into positive territory dWAR-wise, and in any event, it’s hard to accept that, in those relatively few chances that Chapman converted over Seager, three and a half wins hung in the balance over a 162 game season. At least to me.

       My point, which I’ve made a couple of times in the past, though not recently, is that WAR works very well for hitting, and, I believe, that was its original instigation. Where I think its supporters goes wrong is in trying to enlarge the formula, whatever it is, to incorporate pitching and fielding, instead of keeping the three areas of play separate. Pitching is not the mirror image equivalent of offense, and neither are the equivalent of defense. The mystique of having an all-encompassing, perfect system that explains the universe is a strong incentive to denial, however.

       As to how much weight to give to defense vs offense, it’s good to keep in mind the fact that while great defense occasionally saves runs, very often no runs would have scored anyway, and while poor defense allows a greater chance of runs scoring, usually they don’t, because the runner who reached or advanced on an error never makes it all the way home—and even if he does, the percentage chance of that run being a deciding factor in the outcome of the game, while not minuscule, is probably in single digits. Factor in the opposite figure, the batting prowess of a good hitter producing runs to negate those occasional fielding lapses, and the picture becomes even cloudier.

       Reply ↓
       1. no statistician but February 21, 2019 at 8:44 pm

          Let’s clean up the subject verb agreement here: “. . . its supporters go wrong . . .” Right?

          Reply ↓
       2. Bob Eno (epm) February 21, 2019 at 9:25 pm

          B-R’s calculations for dWAR now rely on Baseball Info Solutions analyses of fielding performance. A while ago, I began to get hold of BIS’s more or less (I think less than more) annual publication, “The Fielding Bible,” to learn what those analyses were about. I’ve read four of their volumes, but not recently (vol. 5 doesn’t seem available yet). Here’s a summary of what I recall that may explain the non-intuitive dWAR figures.

          BIS, a company with which Bill James has been involved as an “associate” from the start, hires and trains a fleet of observers to watch every game using proprietary software that yields StatCast-like information on ball velocity, distance the fielder covers at what speed, etc. Every play is scored according to a comparison with norms for position fielders’ abilities to handle the type of ball hit at the relative location, data bank BIS maintains and updates (plays are sorted into typologies — B-R’s info page indicates there are over 80 positive and negative ones). BIS can afford to do this because their main market for all this is MLB teams; the publications you and I can buy are only issued after the year’s data has become somewhat stale. A player’s annual dWAR score is not derived from traditional figures or Range Factors: it is the non-transparent outcome of this play-by-play scoring process, which cumulatively generates each player’s Defensive Runs Scored. If you look on B-R’s WAR explanation page, under “Rdef. Fielding Runs”/”Defensive Runs Saved” there is a brief explanation and a link to “The Fielding Bible III” on Amazon.com — unfortunately, you can’t read it, you can only buy it (a used copy is available for about $5 total . . . there are two articles by James in vol. 3). Essentially, what BIS adds is a “degree of difficulty” factor to each play. (BIS is a division of SIS: Sports Info Solutions, which covers football too, and seems to have about 35 employees in addition to the video monitors. They are currently hiring people to move into Minor League monitoring, so they must believe MLB teams want more of what they do.)

          I’ve found BIS’s account of what it does very impressive, and I’m also impressed that they modify their process when possible weaknesses are noted (I’m thinking of some instance where someone suggested that their monitors might have a perceptual bias in some respect, and they agreed and compensated for it — sorry I don’t have details at hand, and poring through four large volumes isn’t happening tonight. BIS has also agonized over how to avoid overrating fielders as a function of the increasing use of shifts). I have far more faith that they are crunching real events carefully perceived than I have faith in traditional stats and range factors.

          B-R decided some years ago that BIS was more reliable than Total Zone Runs, which B-R used before BIS’s data became available. The two systems’ results for each player are listed as Rtot (Total Zone) and Rdrs (Def. Runs Saved) in the fielding section of each player’s B-R fielding chart. Chapman’s Rdrs number is 29; Seager’s is -5.

          Unfortunately, there is no way to know how those numbers were generated from the specific plays each fielder handled — the system is not transparent. (And if it were, I think none of us would have the time to replicate its methods to assess its accuracy.) I trust it to a pretty high degree because I’ve read the books and thought through their methods, but I’m no authority to rely on. Sean Foreman at B-R would be better, and Bill James, perhaps, better still. But the acid test seems to me to be the fact that MLB teams contract with BIS for their data and assessments.

          As for the Seamheads article, I think Professor Hoban made the same mistake I made a couple of threads ago (which Michael Sullivan caught): he seems to have forgotten that oWAR and dWAR each fully incorporate Rpos, which has to be subtracted from one when calculating WAR. That is why he claims the B-R stats “defy common sense.” (I’m hardly in a position to criticize him.) Apart from that, Hoban simply seems unaware that bWAR is based on a set of numbers measuring different things from Win Shares.

          As for nsb’s point about how counter-intuitive it is that over three wins would be determined by one fielder’s contribution, I have to agree that it’s counter-intuitive to me too. But I’ve read articles in “The Fielding Bible” that have taken me through this sort of thing with specific examples of individual player-seasons, and I’ve been persuaded upon reading. Basically, every play is assessed with regard to the predicted RE24 run values associated with its being made or muffed, as a product of the degree of difficulty.

          I’ve written about “The Fielding Bible” before in comments here, but I haven’t seen responses that indicate anyone else has look at the series. It would be great for me if that happened, because I have as little faith in my abilities to assess this stuff as — I’d guess — everyone else has when they hear me sound off. A second opinion would be very welcome.

          Reply ↓
          1. Bob Eno (epm) February 21, 2019 at 11:02 pm

             As so often, I regret failing to close the hyper link.

             Reply ↓
          2. Josh Davis February 23, 2019 at 1:37 am

             Bob, thanks for the reply. I think you are probably right as to the accuracy of the numbers (your argument that the clubs use the info is a compelling one), at least to say it is the best we can do for now. It’s just that the skeptic in me would like to see how the sausage is made. They might be making assumptions or value judgments with which I would not concur. That sounds horribly arrogant (and most certainly is), but I am loathe to depend on information when I can’t figure out where it comes from.

             Reply ↓
             1. Bob Eno (epm) February 23, 2019 at 12:43 pm

                Josh, I spotted this long after I’d seen and replied to what was really your response to Michael, below. Sorry. Bottom line: information that addresses your skepticism is out there, if the characterizations of it here are inadequate. Still, that will only tell you how the sausage is made, though in extreme detail. To see how it’s made you’d need to set aside weeks of time to observe the monitors and data crushers at work, and to be trained up to be able to assess what you’re seeing — not to mention get access, perhaps through high-placed friends or relatives. But for normal fan-level skepticism, used copies of the various Fielding Bibles are really cheap on Amazon: you can get all four volumes for about thirty bucks, and you don’t need all four. That and a few hours time should let you know whether your skepticism is going to stay put or relax.
       3. Michael Sullivan February 22, 2019 at 1:52 am

          Bob went into a lot of good detail about the modern fielding metrics that I won’t duplicate. Also, in this case, my point from a prior post about dWAR incorporating positional runs doesn’t affect the comparison between Seager and Chapman because they both played 3B and have almost the same rPos.

          But I want to talk a little about your assumption that the difference in chances couldn’t possibly amount to 3.5 WAR, and talk about why I see it very differently, and I think so do most sabermetricians.

          To get there, we have to go back to the whole reason people adopted modern fielding metrics. It turns out that Errors and fielding percentage and putouts, assists, etc. as they are traditionally determined are a really terrible way to assess fielding value. Generally, fielders are assessed errors when they get to a ball but fail to complete the play. In some cases, the better fielder will get more errors because they get to more and harder balls.

          Note that chances are not something due entirely to the luck of where the ball comes. You get more chances if you are able to reach more balls. Traditional scorers often rule something as no chance, not because a good/average fielder couldn’t make the play, because the actual player on the field was nowhere near it. While a better fielder might have gotten close enough to get a chance, and then get dinged with an error if they don’t make the play.

          Potentially every extra “chance” that didn’t turn into an error is an out that one player made and the other didn’t.

          Now, let’s talk about potential fielding value from these. It’s not some kind of mental judgement. It’s a fact of the outcome of the play. Good fielder got to a ball and made an out, poor fielder never got there and thus didn’t get a “chance”. The difference is that batter made an out in one case and got a hit in another. If our scorers who decide on chances and errors were perfect, and always judged correctly what plays an average or acceptable fielder would have made, then we could use fielding percentage and error rate, etc. as our primary defensive metric. But we know from experience that this isn’t even *close* to true. My alternate explanation (that extra chances are purely the result of better range/fielding) is closer to the truth.

          Finally, the value of a good fielding play is exactly as high as the value of a good batted ball, just in inverse! if a play would have been a hit or a double 99% of the time with an average fielder and you save it, that’s worth about as much as the hit/double would have been. And similar if your poor range or whatever turns an out into a hit, that’s also a comparably big deal.

          So what happens if we decide that every “chance” that isn’t an error was a hit that became an out because Chapman got there and Seager didn’t. If we normalize Seager’s chances to Chapman’s innings, I get 417 instead of 437. This means Chapman had 66 more chances than Seager over the same number of innings at 3B. Chapman did make 6 more errors out of those, but if we’re presuming these were balls Seager didn’t even have a chance at, an error is really the same as Seager’s “no chance” as far as the result goes.

          This means that 60 times over the course of the season, Chapman got to a ball, and fielded it for an out (putout or assist) that Seager would not have (assuming the spread of balls was equivalent). That’s potentially as many as 60 hits that became outs!!

          If you had two batters, and one had 140 hits and the other had 200 hits with same numbers of SO/Walks/BiP/HR, would you be surprised if the 200 hit batter had 3.5 more WAR? I wouldn’t. I’d actually be surprised if the difference was that *small*. I’d expect a difference of 5, maybe even closer to 6 WAR. The linear weight value of a single is .8 runs relative to an out and a few of those extra hits were probably doubles worth >1 run. 60 x .8 is 48 runs, and the 2018 AL context is 9.12 total runs/g, so we get 5.26 WAR difference from 60 more or fewer singles.

          Now, B-R is using RE24, so the situation matters, some singles won’t be worth .8, some may be worth more. Also, they are using this BIS data, as Bob Eno describes, which based on expert judgement of where the ball was hit and how well an average fielder would have done on it, so this is likely to be much more accurate than my first cut assumption that every extra chance is a saved hit. I suspect 3.5 wins is closer to the truth than 5.3 or 6 wins.

          But I’ve certainly got no good reason to think 3.5 WAR is far too much. 60 extra chances is a LOT, once you understand what extra “chances” really *mean*. It absolutely does NOT mean “60 more balls were hit his way”. It’s a lot closer to “he got close enough to 60 more balls to make a good fielding attempt” The real answer is somewhere in between, and exactly where in between is exactly what TZ, UZR and this BIS data are trying to determine/approximate.

          If he really made 60 more plays on essentially the same balls in his area, that’s worth much *more* than 3.5 Wins. The B-R data is saying that there wasn’t actually *that* much difference between them. *Some* of the extra plays Chapman made were because he got a few more balls in the right place than Seager did, otherwise the WAR difference would have been bigger.

          Note: even if you don’t normalize chances by IP fielded, there’s still a 40 made play difference between the two, enough for … wait for it … ~3.5 WAR.

          Reply ↓
          1. Bob Eno (epm) February 22, 2019 at 10:32 am

             Michael’s description is an improvement over mine (and not just because he didn’t trip over hyperlinks). His point that every good fielding play has precisely the same value magnitude as a good batted ball is precisely the obvious fact that has not been built into the intuitions of most of us, who grew up relying on batting statistics and largely ignoring fielding statistics.

             Michael’s discussion of errors is also very apt. In one of the “Fielding Bibles” (v.2, I got it off the shelf this morning) Bill James has an article on errors and the first sentence is, “The concept of an error is an error.” Instead, BIS uses what it calls the “plus/minus system,” which assigns for every play not judged routine a value and characterizes the nature of the value: 28 categories of “Good Fielding Plays” and 54 categories of “Defensive Misplays” (James describes many of these). What I’d forgotten is that the system was invented by James (the BIS director, John Dewan, writes about how James sold him on the idea). I don’t mean to suggest that James’s ideas have divine authority, but they are certainly never based on mechanical thinking.

             A tangential thought: Somewhere else, I read an article by James in which he had the idea that the concept of an error was not actually an error at the time it was developed: that the nature of late 19th century fielding, saturated by booted balls, was such that the concept was more productive. If I recall, when you look at very old box scores, fielding statistics (e.g., PO, A, E) are included next to batting stats. When the game was young, people were much more attuned to the value of fielding, in part because the differences in fielding quality between players tended to be more obvious (the quantity of misplays being so high), and in part because early baseball was conceived in terms of batters offering challenges to fielders, rather than pitchers offering challenges to batters, which would better describe our view today, where fielders are “support” for the pitcher.

             Reply ↓
             1. Paul E February 22, 2019 at 1:48 pm

                “When the game was young, people were much more attuned to the value of fielding, in part because the differences in fielding quality between players tended to be more obvious (the quantity of misplays being so high), and in part because early baseball was conceived in terms of batters offering challenges to fielders….”

                Bob Eno,
                Yes, perhaps to such an extent that even Babe Ruth, (at some point early in his career?) suggested that Hal Chase was the game’s greatest first baseman ever and would be part of his “all-time team”? This, despite a modest OPS+ for the position. What’s interesting about fielding (versus hitting) is that while the offensive production side of the game is probably a bell curve for a career arc, I have to believe the athleticism associated with fielding has to decline despite “experience”. For instance, Manny Machado may still improve slightly as a hitter at age 26 through 30 and level off at 32, his fielding will probably decline soon since there’s not as much positioning (and more pure athleticism) involved at 3B (compared to SS)? Same thing with Arenado….and Chapman? But, still, it seems there is a preponderance of guys on the all-time dWAR lists from the last few years. Perhaps the judgment calls by BIS are to blame or is it just a lac of data from way back? Is Andrelton Simmons really that much better than any other SS ever? Or, for that matter, is Jason Heyward really contributing THAT much when he’s hitting likea middle infielder from the 60’s?
             2. Michael Sullivan February 22, 2019 at 3:44 pm

                Paul, I think there’s a lot to the lack of data from before fairly recently. My guess is that the numbers we’re getting *now* are actually getting decently accurate but some guys from before were *far* better or worse than we realize, or have the data to determine now. And that doesn’t just mean that guys with high rField from totalzone were better and negative were worse.

                My guess is that there may be somebody that got scored as just very good by the limited data who was actually Belanger/Smith/Robinson level in the infield. And maybe some of the guys we revere weren’t quite as good as we think. Or maybe they were even further ahead of the pack! Basically we have to put big error bars around the data from before BIS, and we just will never know, because nobody took video of every single game back then that we can analyze.

                I mean, we have potential problems with *all* data, so we really should be thinking in terms of error bars around everything. The error bars around offensive data have gotten small enough that we tend to take the data for granted, but even there, you can see differences in RC+ vs. rBat etc. so there are some error bars there as well. Park ratings aren’t perfect and are based on rolling averages, and we don’t really understand all that clearly exactly why some parks are hitters/pitchers parks or why they change from year to year when nobody moved the walls.

                You still get some bigger error bars still today’s fielding numbers than offense IMO, but I think they are smaller now than for data from before UZR and BIS (2012 or so?). And IIRC, the farther back you go, the more TZ has to assume because the scoring was less rigorous and complete. So fielding numbers from the deadball and golden eras are going to be gigantic and more suspect than from the 60s, let alone the 90s/00s.

                But I think it’s foolish and completely the wrong answer to throw up our hands and think that we should just *ignore* the fielding numbers because they aren’t as good as we’d like them to be. The numbers we have represent *someone’s* (and usually someone very intelligent and expert at baseball quantitative analysis) central case estimate for what each player was worth.

                I sometimes see the assumption that because most of a player’s value was on offense, we can trust his numbers more than a player whose value is based on fielding. But that doesn’t make sense. The error bars around Manny Ramirez’s rField are *just as big* as around Brooks Robinson or Larry Walker! I mean, I think we can be fairly certain that Manny wasn’t a *good* fielder, but what if he were only moderately below average? Or what if he was actually a 60-70 runs *worse* than rField says? It makes the same potential difference for an “offense only” candidate as for a strong fielding candidate. Using our lack of confidence in fielding metrics to only induct offense only players among borderline candidates, or to always assume that absolute fielding numbers are likely to be overstated is a major logical error. (I’m not saying you’re doing this or suggesting it, only that I’ve seen it a lot here and elsewhere).

                Think of it like this. Back before sabermetrics was a thing, we had BA and SLG and runs, hits, HR RBIs and (sometimes) OBP. And pitcher W-L, K, BB and ERA. And that was pretty much it except for the “eye test” and “intangibles”. And people who knew baseball and had some understanding of what these metrics were and were not good for, largely understood that batting average was a mediocre metric, that RBIs were great to have, but that where you batted and who else was in your lineup had a tremendous effect on your ability to score and/or bat in runs. Everybody sane understood that none of these stats were the be-all and end-all of performance. That it was a lot easier to bat .300 or hit 30 homers in 1930 than in 1968 or 1907. That a .260 hitter with power who drew walks might be more valuable than a .300 slap hitter who never met a pitch he wouldn’t chase. But we didn’t have anything better, so we made the best judgements we could based on the data we had. What we didn’t do was say “I don’t need to know that guy’s batting average, or how many RBIs he had, because those are bad numbers, I’m going to just go by my eye (or whatever)” Because that would have been a far, far *worse* way to judge than using old school triple crown stats. We can say stuff like that *now* about BA and RBIs because we have much better stats to use instead that do a much better job of accounting for offensive performance.

                But ignoring modern fielding metrics because they aren’t yet really super awesome, would be like ignoring a batter’s triple crown stats back in the days when we didn’t have any advanced stats. I remember Tango saying once that most people who want to ignore WAR because it is not good enough, end up implicitly saying they prefer to rely on some other metric that they are more used to, but which we can conclusively demonstrate is not as good as WAR, whatever WAR’s limitations.

                This is why I’m friendly to an analysis that includes judgements based on gold-gloves or contemporaneous descriptions of how good somebody was in an effort to decide whether to shade fielding numbers a little one way or another, but I’m not at all receptive to the idea that we should throw out fielding numbers entirely because they aren’t perfect. They are *far* better than what we had before they were invented.

                Ok, I’m ranting now, and kinda way past what I was originally responding to, but it felt good, so I hope somebody enjoys reading it. 🙂
             3. no statistician but February 22, 2019 at 5:17 pm

                Michael:

                I just took another look at the Seamheads post and saw that you contributed your thoughts and analysis there as well, which I was hoping you might do. A further issue, not previously mentioned here but discussed from time to time in the past, is the great advance made in gloves and field maintenance over the years, though with gloves it’s really a twofold change: larger by far and better designed. The greatest of these changes occurred in the decades from the late nineteen-fifties through the mid ‘eighties, I would say, so they seem perhaps ancient history, but they underlie the potential for better fielding now.

                Another aspect of good fielding that ought to be considered, too, beyond agility and speed, is positioning. I draw your attention to Lou Boudreau, who fielded with a very primitive glove by later standards, but is awarded 23.4 dWAR by B-Ref in a career of only 1539 games in the field. Was Boudreau quick on his feet? Not at all. Having broken both ankles playing basketball in college, he was none too fast a runner and very slow getting started, factors, one would suppose, that would doom him as a shortstop. But Lou had the knack of being where the ball was hit anyway. How? By being in the right place at the right time.

                Why do outfielders shift toward right field when a lefty comes to the plate? Because a lefty pulls in that direction. Basic baseball. Well, Boudreau refined the notion of shifting where he was in response to individual opposing batters, given the pitcher and game situation, until it was close to being a science. My point: surely now, considering all the batting information available to every position player, coach and manager—not just close students like Boudreau—one would expect, one indeed ought to demand, fielding that is superior to that of the past.
          2. Josh Davis February 23, 2019 at 1:29 am

             Michael, thanks for the reply. I appreciate it. But I guess I’m still not understanding. You say that chances are not entirely due to the luck of wherever the ball comes. I get that a better fielder gets to more balls and thus has more chances. But I’m still confused as to how we know every extra chance Chapman got was due to his superior range. Is it not possible that he simply had more balls hit his direction? He got closer to 60 more balls than Seager did. You say that doesn’t mean 60 more balls were hit his way. But how do we know that? At one point you assume the “spread of balls was equivalent.” Why should we assume that? You say BIS is trying to approximate this, but that still leaves me fairly uncomfortable as to the accuracy of the system. Perhaps I’d feel better if such a disparity showed up over a period of time…

             Reply ↓
             1. Bob Eno (epm) February 23, 2019 at 10:54 am

                Josh, BIS monitors and records the details of every play of every game, including how far and at what speed players move to reach balls hit at what angles and speeds, all detected by software differing from StatCast basically in that the measurements are from video. Each play is also described according to their 82 classifications, descriptive data that classifies the actions of fielders on each play. They hire and train lots of people to do that and nothing else. They don’t regard their data as an approximation.

                This is the answer to your questions “How do we know that?” “What should we assume that?” If you don’t trust these answers, it’s easy to understand. How would we know that all this is done as BIS says, and is well done? But I think such justified natural scepticism really needs to survive reading BIS’s publicly available (although not online and not cost-free) materials. That’s where the detailed answers to your questions lie and there’s no real shortcut. You’ll find in their analyses many examples of players compared in the manner of Chapman and Seager, but over time. If going through all that material is too much time and trouble (and these are big volumes, so it well may be) the proxy for that would be reflecting that the source of all this, Bill James, is not a sloppy researcher, and that the primary consumers, MLB teams, don’t shell out money for garbage. (Well, we’ll see how Machado does . . .)
             2. Michael Sullivan February 23, 2019 at 11:39 am

                We don’t know that every extra chances were the result of superior range, and I don’t think that’s the case. As I said, if I assumed that were the case, the rField and dWAR difference would be much greater!

                But how do we know that at least *some*, and maybe even *most* of the extra chances were likely due to more range or better positioning? By watching the games. It’s really clear that fielders don’t ever get charged errors by failing to reach the ball except in egregious cases. They don’t get scored as a “chance” unless they either make the play or an error. And it’s just obvious watching the games, that plays rarely if ever get scored as errors because the fielder didn’t get to the ball, even when there’s just no way a superior fielder wouldn’t have. How much difference is not obvious, but that there *is* a difference has been obvious to any intelligent baseball fan, and certainly players and coaches since *long* before there was such a thing as advanced fielding metrics. In fact, it’s clear that coaches and players had their own, often informal metrics/judgements that were better than errors and fielding percentage for a long time before there were any standard advanced fielding stats.

                The whole idea of the new fielding metrics is to try to make educated guesses/decisions about where reality sits in between “chances are pure luck of the ball” and “chances are purely the result of better or worse play”.

                When I “assumed” that they were all due to better play, that was purely for the sake of argument, to note that if that were the case, the value of the better fielder would be *tremendous*, much more than the 3.5 dWAR that seemed so suspicious to you guys in the first place. I was speaking directly to mosc’s question about how could “so few” extra chances possibly add that much value.

                Note, that there’ve been two comparisons — here mosc brought up Kyle Seager, on the seamheads post, Hoban used Bregman as a comp (which isn’t quite as clean because Bregman had 217 innings at short which could be confusing things, but he did play a lot of and mostly 3B). Between Bregman and Chapman, the chance gap was *much* wider (over 100 normalized chances), even though the rField difference was about the same. So this suggests that BIS thought a bunch of Bregman’s missed chances were real no-way-no-how, and/or that Chapman did get a bunch more balls hit near him. If we think of an “opportunity” as a ball hit somewhere that a good fielder could reasonably get to, and better fielders will have higher chance/opportunity percentage, we could say that BIS thinks both Bregman and Seager had fewer opportunities than Chapman, but Seager got more than Bregman, and hence They have similar rField despite Seager having a quite a few more “chances” than Bregman. Chapman is getting credit for turning more of his opportunities into chances than they did, but he’s not getting credit for just getting more balls hit right to him that he fields correctly, or the WAR difference would be bigger, and *much* bigger than it was with Bregman.

                Does that make sense?
             3. Josh Davis February 23, 2019 at 5:56 pm

                Bob and Michael, thanks for the replies and bearing with my lack of knowledge on this front. I do think your recommendation to do some reading on the subject is a good one and might allay some of my concern. This explanation makes me feel *better* about dWAR, as I will readily admit it seems to be the best we can do at the moment. I suppose what is still a bit disappointing to me is that for all the criticism fielding percentage (rightfully) receives for being a scorekeeper’s opinion, dWAR is still relying on human opinions about what constitutes a chance. As Michael points out, this is improvement and less likely to be in error, but human judgment is still a factor. Perhaps that is as good as it gets for fielding, since there may be no objective way to say a play absolutely should have been made in certain cases. At any rate this was a helpful discussion for me and I appreciate the time you put into your explanations.

                And after all that….no one even tackled Boyer vs. Nettles!
             4. Bob Eno (epm) February 23, 2019 at 7:24 pm

                I ought to do Boyer vs. Nettles. I did Boyer last round and I haven’t revisited Nettles in ages.

                As for human judgment on fielding, remember balls and strikes — and they’re worse, because with technology now everyone can see how frequent the umps get it wrong.
             5. Bob Eno (epm) February 24, 2019 at 1:11 am

                In comparing Boyer and Nettles, I went to fielding data first. Since Rdrs stats are not available for their periods, I decided to look at Rtot, there being no leaderboards for dWAR. Obviously, Rtot figures are not as reliable as Rdrs, but I think I found some interesting stuff. This is going to be a ramble.

                Rtot is available only from 1953, but Boyer comes up in ’55 so it covers both Boyer and Nettles. When Boyer arrived, annual Rtot leaders at 3B had ranged from 4 to 10 in the brief period figures are available. Then, in 1956, Boyer’s figure is 17. He didn’t lead the league — a journeyman named Randy Jackson set a record of 18 (in only 80 games at third!), and Jackie Robinson, who shared third with Jackson that year (72 games at third) had 16 (!!). Something was in the water at Ebbets Field that year: 34 Rtot at 3B in a 154-game season; Manny Machado holds the individual single-season record of 34, set in 2013, and he played 156 games at third. . . . But getting back to Boyer, his 17 is, apart from Jackson (whose reputation for good fielding I recall), more than any MLB third baseman in the ’50s, and he posts good figures (in the context of the times) for a number of years, leading the league three times. In the NL, no one exceeds Boyer’s ’56 total (apart from Jackson) till Santo, in a completely atypical year, has 18 Rtot in ’67, and then no one again till 1974.

                Meanwhile, in the AL, Brooks Robinson becomes a regular in ’58 and almost immediately rises to the level of Boyer’s best: between 1960 and 1964, he reaches 17/18 Rtot three times. But in 1961, Boyer’s brother Clete blows away all previous standards with an Rtot of 30, followed by 28 the next year. It is Clete, not Brooks, who breaks out of the standards his brother Ken helped set (something Gold Glove voters did not register). Brooks does not break out of the teens until 1967-68, when he rises to Clete’s level and slightly beyond, with 33 and 32 Rtot those two years. That’s still pretty much a standard of rare excellence today. Since 1953, over 66 seasons, 30 Rtot has been achieved only seven times, and 25 Rtot only 16 times.

                So what about Nettles? Nettles becomes a regular in 1970 and debuts at 21 Rtot. The following year is his peak: 30 Rtot (so he’s one of the seven), a league leading figure, one of three times he led the AL. That’s the same number of times that Ken Boyer led the NL. Nettles is at much higher numbers (Boyer: 10, 13, 10; Nettles: 30, 20, 27). However, Boyer was, in the context if his era, pretty much parallel to Nettles, though his best two seasons (17 and 15 Rtot) he was a point off leading his league. First Clete Boyer and then Brooks Robinson had transformed the nature of 3B defensive play between the two eras.

                In the end, Nettles’ 3B Rtot/yr figure of 8 is not much high than Ken Boyer’s (5). (Compare Brooks: 14 and Clete: 16). All these players are excellent to exceptional at 3B — normal Rtot leaders at 3B have remained in the teens, and Ken Boyer ’56 would have led the NL in 2018 — but among them, only Boyer was a strong hitter as well. Nettles was clearly a better fielder than Boyer — he has exactly twice as much dWAR. But that’s largely a product of era. Third base caught up in quality with other fielding positions in two waves, first the 1950s, and Boyer was on top of that wave, and then in the 1960s and 70s, and Nettles was on top of that higher wave, but not a leader of it.

                Nettles’ career was longer than Boyer’s, and he amassed more total WAR. Boyer’s rate stat of WAR per qualifying season is far higher than Nettles’ (3.8 to 3.3). I think both players would be comfortable in the CoG, but I’d give priority to Boyer. On the other hand, it’s close enough that a vote for Nettles (whose acrobatics in the 1977 and ’78 Series made him look Brooksian) doesn’t seem wrongheaded to me.
27. Richard Chester February 21, 2019 at 5:36 pm

    My vote.

    Primary

    Dick Allen
    Ken Boyer
    Andre Dawson

    Secondary

    Bobby Abreu
    Todd Helton
    Ted Lyons

    Reply ↓
    1. Bob Eno (epm) February 21, 2019 at 7:54 pm

       Richard, Ted Lyons is now on the Primary Ballot.

       Reply ↓
       1. Richard Chester February 21, 2019 at 8:33 pm

          That’s the second time I fouled up.

          Again, primary

          Dick Allen
          Ted Lyons
          Andre Dawson

          Secondary

          Bobby Abreu
          Todd Helton
          Minnie Minoso

          Reply ↓
28. Bob Eno (epm) February 21, 2019 at 11:00 pm

    Interim vote report (as always, * signals a player on the bubble):

    Primary Ballot

    12 ballots submitted

    6 – Manny Ramirez
    ———————————————–50% (6)
    4 – Ted Lyons*, Don Sutton*
    3 – Dick Allen, Kevin Brown, Bill Dahlen
    ———————————————–25% (3)
    2 – Richie Ashburn*, Andre Dawson*, Graig Nettles, Ted Simmons*
    ———————————————– 10% (2)
    1 – Ken Boyer*, Hideki Matsui, Bengie Molina, Luis Tiant, Bobby Wallace

    Secondary Ballot

    12 ballots submitted

    7 – Dennis Eckersley*
    6 – Todd Helton
    ————————————————50% (6)
    4 – Stan Coveleski*, Minnie Minoso*
    3 – Monte Irvin*, Rick Reuschel
    ————————————————25% (3)
    2 – Bobby Abreu, Andy Pettitte, Willie Randolph, Reggie Smith*
    ————————————————10% (2)
    1 – R.A. Dickey*

    Voters: Voomo, Gary B, koma, Bruce G, opal611, Doom, epm, Jeff B, Andy, JEV, Hub Kid, Richard C

    Vote change deadline is Tuesday 11:59pm; vote deadline is next Thursday 11:59pm.

    Reply ↓
    1. Michael Sullivan February 22, 2019 at 12:54 am

       WTF? Seriously? Manny might make it but not Brown? Almost makes me willing to vote for Sutton just to give him some more competition.

       Reply ↓
       1. Bob Eno (epm) February 22, 2019 at 3:14 pm

          For those who believe the way to handle PED players is to try to come up with some assessment as to whether they would be CoGworthy if we could identify and discount the seasons that were tainted by PEDs, Manny represents a particularly difficult problem. Whereas Brown’s use is discussed in some detail in the Mitchell Report, allowing us to learn something about the time frames for which there is positive evidence, Manny is not mentioned there. Mitchell Report discussions often involve retrospective testimony or physical evidence (such as receipts for PED payments) that give us some insight into a player’s apparent history of PED use. In Manny’s case, all we have is two instances of failed drug tests, in 2009 and 2011, at the tail end of his career.

          I’m not much concerned with rule-breaking or character issues when it comes to PEDs, and as someone else reently mentioned, the CoG has no “character clause”: my concern is simply that the stats are tainted and we’re left doing detective work with crude tools to try to put PED-using and clean players on a level field for comparison (if we aren’t willing to ignore the issue). I think we can do that reasonably in Brown’s case, but in Manny’s it’s much more difficult. Minimally, only about 5% of his WAR was compiled from 2009 on (actually, a bit less), so his career stats largely seem to have integrity when used, say, to compare him with other CoG candidates. However, the fact that he failed tests twice and apparently used PEDs even knowing that testing would occur suggests: (1) That he had a pretty loose attitude towards taking PEDs; (2) That he had reason to think he could beat the tests. He turned out to be wrong on (2), but it’s hard to know whether that means he was an unusually slow learner, or whether his confidence was based on previous times when he did beat the tests (not to mention the pre-testing era).

          I don’t know the answer. Any approach we take could be fundamentally flawed — that’s not our fault: it’s Manny’s fault. For me, given that, as we all know, “Manny be Manny,” his recklessness in 2009 and 2011 makes it very hard for me to believe that PEDs do not color a much larger slice of his career than the tail end, and in practice I discount Manny’s stats much more (I’ll know how much when the CoG threshold drops to a level where I feel uncomfortable leaving him out — right now, he’s basically not competitive for me). On the other hand, I do bear in mind that Manny is not mentioned in the Mitchell Report, and that the more a player was involved in PEDs prior to the report the more likely it would be that someone would mention him to Mitchell. So it wouldn’t be responsible to assume that Manny was a heavy user during the heart of his career; as it stands, I actually discount Brown’s stats somewhat more.

          I realize how vague and subjective that all sounds, but, again, the fault lies with the PED users (and I’m happy to throw in their MLB enablers), which is also why I’m unwilling to walk away from the problem, which would effectively credit those at fault with their fraudulent statistics. If there’s a moral dimension to this for me it comes from resentment that those players handed fans who love baseball this problem.

          Reply ↓
          1. no statistician but February 22, 2019 at 5:38 pm

             Bob: I appreciate your position, but I think it’s OK to draw tentative conclusions based on performance records, so that’s what I’m going to do here.

             Ramirez’s OPS+ numbers 1995-98, his age 23-26 seasons: 147, 146, 144, 146.

             Ramirez’s subsequent OPS+ numbers starting 1999, his age 27 season: 174, 186, 162, 184, 159, 152, 153, 165.

             Prior to 1999 his highest OPS was .981. From 1999 through 2006 his lowest was .982, and every other year in the stretch was 1.009 or better, usually lots better.

             You see my conclusion.

             Reply ↓
             1. Bob Eno (epm) February 22, 2019 at 8:38 pm

                Yes, I do see. And I don’t disagree with your interpretation as a tentative conclusion, or the appropriateness of your not voting for Manny on that basis. (Of course, you generally don’t vote for anyone.) I’m just not prepared to rely on conclusions drawn from performance alone for CoG voting. I hope there are many cases where performance improvement is simply the result of disciplined effort; it happens (as with, for example, Jose Ramirez the past two seasons), and it may have happened with Manny.

                Some of those here who vote for PED-involved players like Manny and Brown have offered detailed reasons to explain their views, and out of respect for their care in argument (and for traditions of presumption of innocence), I try keep the threshold for evidence higher than I otherwise might, at least when providing my own reasons here.
          2. Michael Sullivan February 22, 2019 at 10:36 pm

             I don’t really like the practice of looking at a jump in performance and then assuming PEDs because players have always had performance spikes for various reasons (maybe it was greenies or tobacco juice?)

             What I do think, though, is that if Manny was willing to use PEDs when they were specifically banned and tested for to the point of two positive tests, am I really supposed to believe he wasn’t *also* using when they weren’t?

             And he’s borderline. More borderline than Brown. If we’re going to make *any* discount for PEDs at all in his case, he doesn’t belong. End of story.

             Plus, he was such a slacker. Yeah, I know Manny be Manny and all, but a lot of guys have gotten the “bad clubhouse” “no hustle” rap for stuff that’s nowhere *near* the kind of lackadaisical show he often put on. Sorry, if you’re going to jog to first base on damn near every groundball, and lollygag around in the field, you need to blow away the other candidates to get my support.

             Reply ↓
             1. Bob Eno (epm) February 23, 2019 at 12:13 am

                I don’t know, Michael. His namesake and fellow lollygagger just signed for $300m. Maybe there’s a secret sauce we’re missing.
             2. Michael Sullivan February 23, 2019 at 11:47 am

                Machado might jog out some weak grounders too, but he’s got 11.3 dWar in 7 years. 73 rfield at challenging infield positions. If he’s slacking in the field (which Manny R. *totally* did), goddamn, imagine how good he’d be if he weren’t.
             3. Bob Eno (epm) February 23, 2019 at 5:34 pm

                Yeah. I know. I wanted Manny to stay with Brookly — . . . I mean LA.
       2. Hub Kid February 23, 2019 at 8:10 am

          For argument’s sake, let’s say that Manny’s and Brown’s careers are about equally great (and obviously their COG and HOF cases suffer from the same primary weakness, although their other weaknesses vary). I think some of the difference between the votes for these two is how long they have been on the ballot. This is Manny’s 7th ballot, and Brown’s umpteenth (I used to keep track, but I’m not up to date anymore). I think strategy has some role here; Brown’s case is clearly a long game, but Manny’s might still have the potential to go somewhere within 10 rounds.

          For me, it’s the difference between my 3rd favourite player on the ballot, and something like my 7th favourite… Brown probably has been higher up, although I have only voted for him a few times. There’s not a lot of difference between them, really, and I can see Manny slipping further down my ballot (closer to Brown) if his case stalls in the same way.

          Reply ↓
          1. Bob Eno (epm) February 23, 2019 at 10:11 am

             Interesting strategy, Kid. Brown has been higher up, but Manny has too, having lost a CoG runoff by only two votes last year. You’ve written positively on Manny in the past, including your preference for him over Brown, but you haven’t been one of his regular voters. So it looks like Manny’s early support this round has mobilized you to try and get his election done.

             I think I’ve noticed a pattern in voting over the past couple of years: Brown and/or Ramirez pull(s) in early votes, but fade when the later votes come in. I have no idea why, or whether it will be the same this round. As vote totals have fallen it may be that voter resistant to PED players have fallen away as well. At this moment, Manny’s lead has sunk to one vote over Lyons and Sutton (and two over Allen). (Last round, Manny had very little support: he’s already doubled his total this round — CoG voting is strangely volatile in that way.)

             Here’s a question that’s been on my mind that might be good to share with you, rather than to ask of you, since you’re the exception. Allen and Ramirez have similar profiles, Allen’s total stats are lower because his career was shorter, but his rate and peak stats have the edge, and he is, of course, free of the PED issue. Although my ballot is stuffed with early guys whose cases I’ve been backing with an energy and futility that may someday match Doom’s efforts for Brown, I think Allen’s a very strong candidate, and if my ballot slots were more open I’d be happy to include him — he’s always in the running for the one slot Dahlen and Wallace leave open. I’d have expected anyone supporting Manny’s case to be a natural Allen voter as well. But in the last few rounds, you are the only one to put both Allen and Manny on your ballot, which you did this round. It’s a puzzle to me why voters would support Manny and not Allen.

             Reply ↓
             1. Josh Davis February 24, 2019 at 3:34 pm

                As someone who has voted for Ramirez and not Allen, I would say the length of career is definitely a big factor for me. Manny dwarfs Allen on sheer stats and that matters to me. It is a frequent question — who is more worthy, the player with the lengthy and very good career, or the player with a short but excellent career. I don’t know that I’m always consistent on this, but in this case (and in the case of Ted Simmons) I tend to value the longer career a bit more. Manny also earns points for me for postseason performance. I’m not sure I’ve heard anyone reference that here on HHS as criteria for COG, but it is a factor (though small) for me. Manny is a WS MVP who raked in the playoffs and won two rings — and was no small part of baseball history on the 2004 Red Sox. In short, I don’t feel like Allen’s peak (while a little higher than Ramirez’s) is enough to outweigh Manny’s longevity.
             2. Bob Eno (epm) February 24, 2019 at 4:32 pm

                Thanks for taking up the question, Josh. You didn’t address the PED issue, but I assume the difference in career longevity is more important (I do remember your position last year against singling out PEDs for penalties and not spitballs, etc.).
             3. Michael Sullivan February 24, 2019 at 4:32 pm

                If you’re looking at raw stats when you say “sheer stats”, it’s really tough to have an informed opinion given the massive scoring difference when the two of them played. That’s probably the biggest reason some of the not quite inner-circle guys from the 60s and 70s have struggled to get into the hall. Manny was in his prime at the peak of the steroid era and played all but the very end of his career in very high scoring environments. Allen, OTOH, played his entire career in relatively low scoring environments including some years in the 60s that would have fit right in during the deadball era. If you’re looking at numbers like H, RBI, HR, SLG, BA, OPS, Allen would have had to be Babe Ruth or Ted Williams to have comparable numbers in his time.

                Allen did have a much shorter career, but in terms of total career wins above average, he’s not far behind Manny at 32.9 to 35.7. If I look at their long-peaks, basically starting at their first good year, and stopping before a bad year that wasn’t followed by more good to excellent years, I get the following peaks:

                Ramirez 12 years: 7392 PAs WAR 58.7 WAA 33.2 WAR/seasn 4.9, WAR/650PA: 5.2

                Allen: 11 years: 6240 PAs WAR 58.3 WAA 35.9, WAR/season: 5.3, WAR/650PA: 6.1

                So over the course of their primes, they added very similar value, but it took Allen one less year and 85% of the PAs to do it, so all of his value/time numbers look better.

                So I think their base cases (ignoring PEDs) are pretty comparable — Allen has a slightly better peak, Ramirez adds more years at average to average- which beefs up his WAR, but he’s not really helping anybody win pennants at that point. In terms of the time when either one can realistically be a prime piece of a contending team, Allen comes off slightly better, I think.
             4. Josh Davis February 25, 2019 at 3:51 pm

                Years of 2+ WAR (starter value)
                Ramirez 15
                Allen 11

                Years of 5+ WAR (All-Star value)
                Allen 6
                Ramirez 5
29. Dr. Doom February 21, 2019 at 11:31 pm

    Well… I think I’m obligated to share this here. Guess who I shook hands with tonight? No guesses?
    …
    …
    Bernie Williams!
    Yeah, former Yankees CF Bernie Williams! He’s very active in a program called Turnaround Arts that encourages kids to get involved in arts and music. The local school district here in rural Minnesota flew him in from New York to do a presentation in our local school. He had a public Q&A/concert for about an hour at 6. Most of the kids had heard it all during the day, so they didn’t come back (and it was during conferences tonight). There were only about a dozen of us there, and he was very personable. He shared a ton of interesting stories. Here’s a sampling:
    -He said he’s on the road all the time now with his music career; he said he doesn’t really say that he “lives” in his New York home: he “crashes.”
    -He shared that his mother was a real force in his life. She was a teacher and very against him spending time with the “wrong crowd,” so she encouraged him to do everything: sports, guitar, and of course academics. School was always number one. But when he was about eight years old, his mother decided, “These kids (he and his brother) are watching too much TV.” It was at that point that he started both baseball and the guitar.
    -I asked about whether he was actively playing his guitar as stress relief, or whether he took it up more after his career. He said, “Both.” Basically, while playing, he goofed around every once in a while. But when he retired, he really tried to make a go of it. He applied to the music conservatory in one of the New York state colleges and got in. He was twice the age of the other students, and he said they were VERY skeptical of him. Certainly at first when he was just the weird old guy in their classes – and even moreso once they figured out who he was! He said they always had this attitude of wondering whether he really DESERVED it, or if he just got in because of his name. He said it was fun but very challenging to have to prove himself all over again.
    -One of the teachers asked what his transition to the majors had been like, and if any players took him under their wings. He named two: Don Mattingly and Willie Randolph. Mattingly, he said, sheltered him from a bunch of hazing. Williams stated that the “old guard” was still VERY into the rookie hazing stuff, and that Mattingly basically said, “Give the kid a break,” whenever he caught older players messing with Williams. In regard to Randolph, he told a particular anecdote. He said it was during the first season he felt like he’d arrived – 1994 or ’95; Radolph was the Yanks’ 3rd Base coach at the time – that he needed to remember that this wasn’t a fluke or dumb luck. He was good and he should be proud, and he could really feel like he belonged. It’s cool hearing a personal story of someone who’s pretty unanimously thought of as a good guy (Randolph) being so kind to a young player, and really “coaching,” almost the way a Little League coach might.
    -Not about Bernie himself, but… my wife told me that there was a kid there whose conference was actually last week. He just had to come tonight because his dad is from Puerto Rico (our little town of about 5000 is about 35% Hispanic; the school over 50%), and he just had to meet Bernie Williams, so they came back a week later, anyway. I watched them and the kid, far too young to really know/care about Bernie Williams, was just so kind to his father as he watched his dad watching Bernie. It was a very cute part of the night for me.

    Anyway, I stayed after for a minute or two to shake hands with him, to tell him that he was in the field for the second-ever MLB game I ever attended (in 1993 at County Stadium), and that, as the son of two music teachers and a huge baseball fan, he had basically lived my dream life. He couldn’t have been kinder, and this was as a man who is going to spend the next two weeks among Florida, Arizona, and Nevada, and all he could do was speak positively of the experience he had today with the kids in a tiny town in rural Minnesota, where the temperature topped out at, I believe, 19 degrees today. It was lovely, and is truly one of those times I can say I met one of my all-time favorite players. I even admitted to him that I was a Yankee-hater as a kid, but that he was one of my personal favorites. I just loved someone who had such a good all-around game – power, average, and he certainly looked good to me in the field. I’m glad to have met him. He really was as friendly and kind as he’s come across in media; very humble and willing to talk with anyone about anything. It was an awesome night, and my wife (a teacher) and the superintendent told me that he couldn’t have been better with the kids. One last story: he said that, at the end of the day, a first-grader just RAN up to him and hugged his leg and wouldn’t let go. He said that was the highlight of the trip.

    In short, it was a really cool chance to meet a childhood hero of mine (and yes, I know it will make many of you feel very old the Bernie Williams played during my childhood). I know I’ve shared my personal story of meeting Harmon Killebrew before, and you all have enjoyed hearing that. This one was not as good from an anecdote standpoint, but the player was actually someone I personally watched, so it may be more important to me. Anyway, thanks for tolerating this rambly post about a player who will never get COG-reconsideration, but whom I very much admire as a player and a man.

    Reply ↓
    1. Dr. Doom February 21, 2019 at 11:36 pm

       One more thing:
       I nearly forgot what I thought was the funniest anecdote. Williams explained that he’s still involved with the Yankees organization – charity work, “special instructor” status (he’s heading to spring training soon), and – of course – the Old Timers’ game. He said something I thought was really funny: that, once you’re retired, you’re immediately an Old Timer. He retired at 37. He said it was a very humbling (and awkwardly funny) experience to suddenly be playing with 80-year-old guys who had to walk hunched over, and to be considered the “same” as them, just because you were retired.
       Oh – he also shared that he thought the team was joking when they informed him that his number was being retired, which was (I thought) a kind of cute reaction.

       Reply ↓
       1. Bob Eno (epm) February 22, 2019 at 12:43 am

          Thanks, Doom. An exclusive report for HHSers on a really appealing player, and very warmly done, despite the freezing weather. (I too was a Yankee hater who liked Williams as a good guy — though middle aged then, and feeling very old right now!).

          Stay warm. (I guess you folks know how to do that.)

          Reply ↓
    2. Michael Sullivan February 22, 2019 at 12:53 am

       Bernie always came across as a really good guy in interviews, and he was a lot of fun to watch. He had a really fine bat for a CF, and you always knew something could happen when he came to the plate. And he really was just a couple good years short of a good HOF case. Kind of a shame that he got dumped in that dog’s breakfast of a ballot in 2013. Although Lofton (who made the COG) was the most egregious dump. 8 (probably 9 or 10) future HOFers (and 11 COG members) on that ballot and they don’t elect anybody.

       Anyway, this is a great story. I’m so glad to hear he really is as good a guy as he always seemed, and sounds like his post-baseball life is interesting and rewarding. I was an adult when he came into MLB (we share our birth year), but I was a relatively new baseball fan, and he will always be one of the primary faces of the Yankees for me.

       Thanks for this story!

       Reply ↓
30. Dr. Doom February 22, 2019 at 1:25 pm

    Michael Sullivan made the point below that Kevin Brown might not make it, but that Manny Ramirez (somehow) might. I understand that some folks will never vote for either of them. But if you’re going to vote for one… WHY would it be Manny?

    1. They are both PED-associated players. But one of them continued using PEDs into the testing era… I’ll give you a guess as to which one. (And yes, this is only because Brown was retired, and Ramirez was not… but it goes to Manny’s flakiness.)

    2. Black Ink and Gray Ink are almost the same for Brown and Ramirez: 19/21 and 166/154. Ramirez is better in HOF Monitor and Standards, but I’m not too concerned.

    3. Look at Ramirez’s OPS+: it’s 154. Brown’s ERA+ is 127. Those don’t really seem comparable. Big edge to Ramirez, right? OK… well here’s the thing. I’ve introduced in a Don Sutton post earlier in COG balloting a way to look at pitcher ERA relative to ERA+ and ERA.
    Brown allowed an ERA of 3.28 with an average expected ERA of 4.17. That implies a replacement ERA of 5.00. That’s 1.72 R/9. That’s a total 622 Runs Saved Above Replacement in 3256.1 Innings Pitched. Dividing by the overall run environment (8.34), you get 74.6 WAR. OK?
    Ramirez, as an offensive force, is even greater. By basic Runs Created, he created 1968 runs in 9774 PAs. Given a 154 OPS+ and an assumption of 38 PA/G (which is pretty close), implies a 4.97 run/game expectation, with Ramirez producing a remarkable 7.66. Anyway, the replacement expectation is 3.97 R/G. That’s 3.69 R/G better than replacement, times 257 “games” is 948 Runs Above Replacement, in a 9.94 Run Environment is 95.4 WAR.
    So how could Brown possibly be better than Ramirez with a 20-WAR gap? Defense. We haven’t accounted for the defense. If we assume Baseball-Reference is right in their positional adjustment, position alone drops him by (thankfully) exactly 10 WAR. That narrows the gap to 85.4-74.6 We’re 10 WAR apart now. So the real question is, did Ramirez cost his teams 10 wins with his defensive play, or not? That’s really the only question you have to answer. Baseball-Reference estimates it to have been about 12 games (a shade over, actually). I don’t think that’s an unreasonable estimate; quiet reasonable, actually. UZR is even less kind to him than TZ or DRS, so I think it’s safe to say that he’s 12 games worse at best. To me, you also have to look at peak. So on total value, I have Brown a shade ahead. (BTW, both played 19 years, so this is a pretty perfect way to compare them.)

    4. Let’s say you want to talk peak. I would agree with that, because, frankly, you have a better chance of winning World Series titles with more peak, and it’s about winning titles, isn’t it?
    Well, Kevin Brown was, for five years, as good as anyone. There were eight active COG starting pitchers at their peak(-ish) during Brown’s 5-year peak of 1996-2000. Here they all are, listed by WAR (their VERY best 5-year peak in their career is listed in parentheses):
    Pedro Martinez – 41.8 (42.9)
    Kevin Brown – 36.7 (36.7)
    Roger Clemens – 35.2 (41.6)
    Randy Johnson – 32.1 (43.8) – Johnson was hurt in 1996; it may have cost him 5-7 WAR
    Greg Maddux – 31.3 (41.4)
    Curt Schilling – 27.4 (36.5)
    Tom Glavine – 25.3 (25.3)
    Mike Mussina – 24.2 (28.3)
    John Smoltz – 19.9 (24.0)Smoltz missed all of 2000; it probably cost him 4-5 WAR
    Brown is not at the bottom of the group; he was second in the group (third if you want to give Johnson injury credit), and he is in the middle overall in peak. The MIDDLE; not the bottom. You’d want him MORE than Glavine, Mussina, and Smoltz. They are in; he is out.
    Was Ramirez ever comparable as a player?
    Manny had 28.8 WAR in his five-year peak (1999-2003). There were 15 COG players active at the time*
    Barry Bonds – 44.3 (51.2) [42.4 pre-steroids] Bonds was hurt in 1999, costing him 2-3 WAR
    Manny Ramirez – 28.8 (28.8)
    Chipper Jones – 28.7 (31.3)
    Jim Thome – 26.9 (26.9)
    Jeff Bagwell – 26.5 (34.5)
    Larry Walker – 25.1 (30.1)
    Derek Jeter – 25.0 (30.3)
    Ivan Rodriguez – 23.7 (30.3)
    Edgar Martinez – 21.3 (30.2)
    Roberto Alomar – 20.8 (27.5)
    Mike Piazza – 18.9 (30.9)
    Kenny Lofton – 18.1 (31.0)
    Frank Thomas – 14.6 (31.8) Thomas missed nearly all of 2001, costing him anywhere from 2-5 WAR
    Ken Griffey, Jr. – 12.9 (37.8) Injured 2001-03; your guess is as good as mine as to his lost value; peak includes injured 1995
    Craig Biggio – 12.8 (32.8)
    Barry Larkin – 8.7 (27.4)
    In other words, Ramirez was ALSO the second-best among COG members in the league in his five-year peak… however A.) we omitted non-COG players – Sosa (29.3/33.0), Rolen (25.9/30.4), A-Rod (40.6/43.5), Garciaparra (27.4/28.3), Helton (32.1/37.5), Palmeiro (20.9/26.3) – which would knock him down to fifth, AND he has among the worst 5-year peaks of any of these players. Only the oft-injured Larkin, Garciaparra (who basically is forced to include his missed 2001 in any 5-year peak), and Thome and Alomar, who struggled MIGHTILY to get in, as well as producing more career value than Ramirez. Everyone else goes over 30 WAR in their five-year peak. AND we didn’t get into contemporaries who didn’t overlap with that particular 1999-2003 period, but did overlap with him substantially: Ichiro (29.9), Pujols (44.6; Pujols manages 20.8 WAR 1999-2003 even though he wasn’t active in 1999 or 2000), Ripken (35.5), Gwynn (30.1), McGwire (29.8). To me, Ramirez doesn’t really fit into the COG, nor does he measure up to what Brown produces among the best pitchers of all-time. And, again, this is even if we DON’T account for the fact that he continued to blatantly use PEDs even in the testing era. If you take all the COG guys excepting Bonds, you get an average 5-year peak of 31.0 (30.96, actually) WAR. Ramirez is closer to the Fred McGriff-Rafael Palmeiro kind of player.
    So that’s my thought. Kevin Brown – dominant and right up there with the greatest pitchers of all-time. Manny Ramirez – a great player who has obvious Hall of Fame credentials, but doesn’t match the COG’s demanding standards. VOTE BROWN 2019!!!

    Reply ↓
    1. Mike L February 22, 2019 at 3:47 pm

       Doom, I’m never going to vote for either Manny or Brown, but I would note that Manny’s HOF monitor was 226, 34th among batters, with the average set at 100. Brown’s was 93, 117th among pitchers, with an average of 100
       ‘m not sure how I would discount their stats for PED use, but that’s a big difference.

       Reply ↓
       1. Dr. Doom February 22, 2019 at 6:11 pm

          I really don’t worry too much about HOFs and HOFm. The reason I don’t is fourfold:

          1.) Those two standards were configured to be commensurate with Hall of Famers in the 1980s. The standards were never reconfigured for players of the Selig years. The argument isn’t that Kevin Brown has a better ERA than Three Finger Brown, nor than Ramirez had fewer HRs than Eddie Collins. With those numbers never being adjusted, when looking at players of the ’90s and ’00s, hitters are systematically overrated and pitchers are systematically underrated. It’s just a fact of how those things work.
          2.) On some level, and related to the first comment, comparing pitchers and hitters is apples and oranges.
          3.) HOFm and HOFs see Jack Morris and Bert Blyleven as being relatively comparable pitchers, with Morris coming out ahead on HOFm. Since I learned that, I have a hard time taking those measures as seriously. What can I say?
          4.) The measures weren’t designed (and this relates to point 3 above) to measure better players; they were meant to measure how closely people conform to the players in the HOF 35 years ago, not how successful they were as players. It’s a fine distinction, but a necessary one for an exercise like the COG.

          Reply ↓
          1. Bob Eno (epm) February 22, 2019 at 8:52 pm

             On #4, I think the distinction is a wide one, not fine at all. James was initially very clear that his devices were meant to predict HoF electability, not HoF worthiness, and we know what James thought about HoF election standards. Still, if I found my assessments radically out of line with those measures, I’d want to be able to explain why. I think Doom’s first point does that well as a first pass — I’ve never thought about the point he makes (since I pay almost no attention to those measures, or to JAWS, for that matter — we have so many stats that let us form our own judgments).

             Reply ↓
       2. Michael Sullivan February 22, 2019 at 10:27 pm

          Hall monitor and standards were based on Bill James predicting how the BBWAA voters would vote based on their voting history in the 60s-80s (maybe 90s?). It wasn’t about judging players, but about predicting what hall voters were looking for. I *thought* that the collective at HHS had established that we weren’t at all happy with the BBWAA’s way of judging candidates, and that’s the whole purpose of these alternate halls. COG in particular is *expressly* trying to relitigate hall of fame voting in a certain way, because we think that the BBWAA f’d up a lot.

          So I don’t see how James’s hall monitor or standards are the least bit relevant to evaluation of players for the COG. If all we’re going to do is take our cues from old school BBWAA voters, why are we even bothering to do this?

          Reply ↓
          1. Mike L February 22, 2019 at 10:54 pm

             Michael, and Doom, I’m going to bow out here. I don’t vote for juicers anyway, and I probably shouldn’t have intruded.

             Reply ↓
             1. Bob Eno (epm) February 23, 2019 at 12:11 am

                Not an intrusion, Mike — no intrusions here — a legitimate observation. Several people have recently invoked the Hall indexes at the bottom of B-R pages and it was useful to have a chance to discuss them.

                (Actually, I do recall an intrusion here once. In the midst of some long ago CoG round, some guy popped up and said, “If you don’t have Gil Hodges in your circle this whole thing is a joke.” Or words to that effect. It added color.)
31. Paul E February 22, 2019 at 7:01 pm

    Vote:
    Allen, Lyons, Sutton
    Coveleski, Randolph, Reggie Smith

    Reply ↓
32. Doug February 24, 2019 at 1:23 am

    Main: Tiant, Wallace, Dawson
    Secondary: Coveleski, Eckersley, Reuschel

    Reply ↓
33. no statistician but February 24, 2019 at 6:00 pm

    Somebody’s has to be elected in this thing, so I’m going to make one of my usual inflammatory remarks: Ramirez is a ridiculous choice, but he’s now in the lead. Why? The troika effect? Dunno, but the only players on the ballot whose records approximate those of current members of the COG are Dahlen and Wallace, who, despite Bob Eno’s electioneering, or maybe because of it, are getting short shrift. At any rate, time still remains to change your votes and keep injustice from being done.

    Reply ↓
    1. Voomo Zanzibar February 24, 2019 at 6:43 pm

       Ridiculous is a strong adjective.
       We did vote for Harmon Killebrew.

       Manny is eighth all time in ops.
       With an ops+ in the ballpark of Frank Robinson, Willie Mays, and Hank Aaron.

       The contrary arguments are plentiful and valid, yes.

       Reply ↓
       1. Voomo Zanzibar February 24, 2019 at 8:02 pm

          Screw it, let’s have a 5-way runoff:

          Vote change:

          Richie Ashburn
          Ted Lyons
          Manny Ramirez

          to

          Ashburn
          Kevin Brown
          Lyons

          Reply ↓
          1. Bob Eno (epm) February 24, 2019 at 8:26 pm

             Well, Voomo, that change does move us closer to that outcome. Ramirez drops to 5 votes out of 14 cast so far, tied with Lyons and Sutton. Allen and Brown are now just one vote behind. And there are still four full days to go till the Thursday night voting deadline.

             I want to add that although I sometimes wonder whether HHS voters casting ballots for Brown or Manny have engaged with discussions we’ve had about voting for players with PED profiles, I remember that in the past you made strong arguments for your willingness to do so. The fact that posters like you, Doom, Josh Davis and others are comfortable with your reasoning on that side of the debate, while others, like me, feel the same way in reverse, means that this problem is not likely to go away. I really regret it because our business should just be debating on the basis if stats that have high heat, and not arguing elusive meta-issues about character, ethics, drug effects, forensic evidence, or seat-of-the-pants, low-heat statistical adjustments, all of which prevent us from having straightforward votes based on a shared obsession with baseball statistics.

             At points like this, I sometimes write something like, “It’s not our fault; it’s the fault of players who used PEDs,” but, for the sake of HHS unity I’m going to switch to: “It’s not our fault; it’s Bud Selig’s fault!”

             Reply ↓
             1. Voomo Zanzibar February 25, 2019 at 12:39 am

                It is precisely because I want to focus on what I love about baseball that I choose to accept (most of) the numbers from that era.

                We failed as a society, and are still failing, to deal with the performance enhancing drug issue.
                Im not going to vilify individuals. That’s nothing but a distraction.
                The folks in charge of every industry/government tell us to argue with each other over nonsense so that we can never collectively have a useful big-picture discussion.

                We’re supposed to discredit individuals for doing something that wasnt even against the rules.
                Kafka would crumple the idea up as a failed first draft.
             2. Bob Eno (epm) February 25, 2019 at 1:36 am

                I understand the argument, Voomo, though it’s not my view. (And I want nothing to do with vilification; I’m interested only in statistical comparability.) I suspect we’re in agreement on Selig.
             3. Voomo Zanzibar February 25, 2019 at 11:16 am

                It may also be that I often feel obliged to be a contrarian.
                Ideas that have been absorbed into popular consensus need pariahs to question them so that they remain living systems. I was born without the disability of needing to be liked.

                Besides which, this issue has the problem of mass hypocrisy.
                We all knew what we were looking at in 1998.
                And we all chose suspension of disbelief because we were enjoying the story.

                Chicks dig the longball.
             4. Michael Sullivan February 25, 2019 at 12:40 pm

                “Chicks dig the longball.”

                LOL, You beat me to it. I’ve been meaning to insert that quote somewhere in this discussion, ’cause it seems like not just chicks, but some HHS voters as well.
             5. Hub Kid February 25, 2019 at 3:57 pm

                Voomo, a double hat-tip to you: 1) for turning this (rather suddenly) into a pretty exciting CoG voting round. I think I will have to watch the deadlines and think about changing my vote.

                2) for clear and unapologetic reasoning in the great PED debate (apologies to HHS-ers on the other side of the debate, I recognise that your arguments are valid and made in good faith, although I don’t happen to agree with you).
             6. mosc February 26, 2019 at 11:53 am

                Voomo, keep being you. I am not going to vote for Manny but I respect your rationale and appreciate you explaining it.
    2. Bob Eno (epm) February 24, 2019 at 7:15 pm

       ” . . . despite Bob Eno’s electioneering, or maybe because of it . . . ”

       Ouch. But, yeah.

       Reply ↓
34. Chris C February 24, 2019 at 8:08 pm

    Main Ballot:
    Ramirez, Sutton, Simmons

    Secondary Ballot:
    Randolph, Eck, Minoso

    Reply ↓
35. Bob Eno (epm) February 24, 2019 at 11:25 pm

    Ted Lyons and Don Sutton are among the leaders in votes so far. I voted for Lyons and wrote a comment explaining why. In thinking about Lyons vs. Sutton, I want to pick up a point I made about Lyons and compare him to Sutton. Both Lyons and Sutton were pitchers with long careers (Lyons 21 seasons plus 3 War years; Sutton 23 seasons), and high win totals.

    Lyon’s W-L record was 260-230 (he lost three years to the War late in his career, when he was enjoying a string of very good years that would almost surely have brought him to 300 victories, and he retired not because he was done, but because he became manager, and at 45 felt he could not do both). Sutton’s W-L record is stronger: 324-256.

    However, here’s another way to see this comparison:

    ….W-L………..Pct……Team W-L Pct.*….Pct, Differential
    260-230…… .531 …… .460 ……………………..+.071…………Ted Lyons
    324-256…… .559 …… .533 ……………………..+.026…………Don Sutton

    *The team records are only for years with 100+ IP, so I didn’t count Lyons 1923 and 1946 (3-5, 65.1 IP) or Sutton 1988 (3-6, 87 IP). I also counted Sutton ’82 as all-Houston and ’85 as all-Oakland, although Sutton pitched a total of 86.1 IP for other teams those years. Lyons’ team W-L Pct. in those seasons was .462, just about his average; Sutton’s second teams in ’82 and ’85 had better records than his main team, and his team’s W-L Pct. in ’88 was .584, so the figures here are slightly skewed in Sutton’s favor.

    Reply ↓
    1. Mike L February 24, 2019 at 11:55 pm

       Bob, I just wanted to add something here….I’d put a lot greater weight on W-P percentage differentials the farther we go back in time, simply because starters pitched more games and more innings per game. I’d even make the argument (and expose myself to scorn and other forms of hysterical laughter) by saying that a pitcher who starts with the expectation that he’s going nine is a lot more likely to have to pitch strategically (and pitch to the score?) than one who is satisfied when he does six. Blake Snell had a spectacular season in 2018, and was surely deserving of high honors. But he averaged less that 6 IP a start. Lyons, by contrast, pitched 3907 IP in 484 starts.

       Reply ↓
       1. no statistician but February 25, 2019 at 12:48 am

          Mike L:

          I’m not laughing. I’d extend your argument further to say that most starting pitchers into the late 1980s had the hope, if not always the expectation, of throwing a complete game in every outing. It’s quite clear, looking at the leader boards, that this is the case. Relievers were used with more and more frequency in the decades leading up to the sea change, true, but starters weren’t let in on the secret that they weren’t expected to go the distance, I don’t think.

          It’s interesting that two things happened together in 1988: Eckersley became the first modern closer and pitch count became a statistic. The jump in strikeouts recorded is another manifestation of this redirection.

          At any rate, the view now that assigning wins and losses to specific pitchers is irrelevant fails to take into account the fact that pitching by committee is a development of recent times.

          Reply ↓
          1. Bob Eno (epm) February 25, 2019 at 1:32 am

             I agree with some of what you’ve written, Mike and nsb, but since I was doing a comparison with Sutton, the difference is a bit blunted. Sutton didn’t pitch as many CGs as some of his contemporaries, but he was still of a generation of starters that probably entered most games expecting to go the distance. So Mike’s basic point about W-L may work a bit in Lyons’ favor (he was, after all, the all-time leader in percent of starts completed, as Richard has taught us — 74% vs. Sutton’s 24% in a very different era), but it wouldn’t be a deal-breaker for Sutton if his W-L record were as far above his teams’ as Lyons’ is. (Sutton seems to have averaged almost 7 IP per start, which is pretty good. His contemporary, Seaver, stands at almost 7.5 IP. When they pitched, the “committee” was a small one.)

             I’d also add that some post-1980s pitchers continued to hold old-time attitudes (Clemens, Schilling, Halladay), even if their CG percentages were much lower than the old timers’. Today I wonder if there’s a pitcher left who starts a game expecting to finish it. How could they, when league leaders last year each compiled just two CGs?

             I guess I’d say too that while I think the W-L Pct. differential for Lyons and Sutton is significant, I don’t think it’s the most important statistical measure we should be using — for me, WAR rate over roughly comparable career spans is probably the one I’d rely on most, though no one measure is adequate. But in the case of Lyons and Sutton, I think the W-L Pct. differential gap is a useful confirmation if the equally large WAR rate gap (Lyons 2.9/162IP vs. Sutton 2.1/162IP). After all, WAR is not a fully transparent measure — unless we put in a lot of time, and even then we just know how the formula works: we can’t visualize it as events on the diamond. The W-L Pct. differential is a simple and transparent data item. (It would be even clearer if I’d subtracted Lyons and Sutton from their team’s records, but I didn’t keep the team-record data from my earlier Lyons calculation and I was too lazy to generate it again.)

             Reply ↓
          2. Michael Sullivan February 25, 2019 at 11:06 am

             “At any rate, the view now that assigning wins and losses to specific pitchers is irrelevant fails to take into account the fact that pitching by committee is a development of recent times.”

             I’m not sure anyone thinks that it’s “irrelevant” so much as that it’s a poor way to capture a pitcher’s performance and paying too much attention to it allows luck to bias your picture significantly.

             I suppose you could make an argument that pitcher W-L is *marginally* more relevant in an era of 50-70% CGs than in an era where 6-7 IP is effectively treated as a complete game unless there are bullpen issues.

             But the pitcher doesn’t control everything on the defensive side of the ball, and they have basically no effect (unless you’re Ferrell or the like) on the offensive side. they may be the most important player who has the most effect on the outcome, but their performance controls less than 1/2 of their team’s performance at best, and I think it’s more like 1/4 even in a CG. Consider that a pitcher’s affect on batted balls is relatively small compared to TTO, and TTO are about a third of plays in today’s game (perhaps was more like a quarter years ago). Now 1/4 is a LOT, given that there are 9 players playing (and 9.5 in the AL since 1973).

             so you can assign W-L to a pitcher, but it’s a lot like assigning wins to a quarterback in football. Yes, they are the most important player, and a great quarterback can be the difference between an average team and a super bowl team, just like, in any given game, a dominant starting pitcher gives all but the weakest teams a shot to beat anybody. But just because you put Greg Maddux or Roger Clemens out there doesn’t mean you’ll go .550 with nothing but replacement players, and assigning a win to a pitcher effectively implies that should be the case. Win differential is worth something, but it’s still got some massive error bars around what it implies about the pitcher.

             To think people distrust modern fielding metrics and yet care about pitcher wins!

             Reply ↓
             1. Bob Eno (epm) February 25, 2019 at 4:03 pm

                Michael, There’s much true in what you write. However, you are addressing pitcher W-L Pct, and noting the role of team quality. The issue here is pitcher W-L Pct. beyond or below team quality, Over the course of a long career, I think that can be a meaningful figure, though not as solid as a number of advanced stats.
             2. Michael Sullivan February 26, 2019 at 1:08 am

                There are too many other problems with pitcher wins, for W-L differential to be a very valuable way of lookign at anything. IMO, the only reason to look at W-L differential is if you’re fixating on pitcher wins and trying hard to do something with them that’s not completely useless because you’ve got W-L on the brain.

                Start with WAR. If that’s not definitive, look at year by year, look at FIP, ERA+ GB/FB etc. But “pitching to the score” is a steaming pill of bullshit, and pitcher W-L is on of the most useless stats ever invented.

                Ok, I’ll admit it, I think they’re irrelevant.
             3. Bob Eno (epm) February 26, 2019 at 1:57 am

                Michael, I disagree with your total dismissal of W-L Pct. differential (stating your opinion and calling one aspect of the discussion bullshit aren’t really arguments), and I think you may have missed something I said. W-L Pct. differential is of value when it aligns with advanced stats, like WAR rate, because it can be intuitively grasped as an event on the diamond, where WAR is basically a figure most of us accept on authority.

                I think citing W-L Pct. differential is a valid argument strategy when it aligns with advanced stats. As a Lyons backer, I wondered why the large WAR/162IP difference between Lyons and Sutton seemed to be getting inadequate traction, so I used a parallel stat that can be easily visualized. Moreover, since I argued in a comment I just posted that I think you seriously underestimate the pitcher’s contribution to defense, I believe the correlation between WAR/162IP and W-L Pct. differential is generally meaningful.

                By the way, “pitching to the score” is closely related to “pacing yourself.” If you read accounts of how past starting pitchers describe their efforts, the importance of pacing themselves, or reserving enough arm strength for critical ABs if the game is close is often noted. The fact that pitchers used to do this is really not arguable. I suppose a young Bob Feller had no use for the idea, but pitchers were paid for wins, not Ks or ERA: their focus was on the W. That’s no longer true; views of the game have changed. (I remember Dick Donovan, who won an ERA title with a 10-10 season one year and won 20 games with a much higher ERA with a better team the next, being asked which was the better outcome. His answer was that the wins were all that really mattered, and he’d have been happy to trade his ERA title for more of them.)
             4. Michael Sullivan February 27, 2019 at 8:00 pm

                What is it that you’re hoping to capture with W-L differential that WAR and ERA+ don’t already cover?

                The only thing I can imagine is the idea of “pitching to the score”, somehow eking out more wins with the same runs allowed as some other pitcher.

                I feel like the whole idea of pitching to the score has been definitively shown to be somewhere between bunk and a minuscule effect. In every athletic contest that can be measured, epople have been trying to demonstrate or falsify an ability to win games/matches that is independent of simple better offensive or defensive efficiency (equivalent of runs allowed for pitchers or linear weights for batters), and every time, they fail to find evidence for a repeatable ability to do anything “to the score” that goes beyond simply scoring more or making your opponents score less.

                There are some perfect information games where this matters (go-chess) but there’s too much variance in outcomes due to chance in athletic contests. The best way to ensure you win a game is invariably to continue to play better and score more points/runs than the other team.

                If you’ve got something else you think you can get out of W-L differential that’s not captured in WAR, let’s talk about that. I might change my mind.

                Note: I agree that W-L differential is a much better measure than W-L. And if I don’t have better information, it’s well worth looking at. Back in 1970, when we didn’t have WAR or fielding metrics, I would have been looking hard at W-L differential! But I just don’t think it’s worth it now.
             5. Bob Eno (epm) February 26, 2019 at 1:27 am

                Michael, I’m troubled by your notion that the pitcher is “in control of” only 1/4 – 1/2 of the defensive game. I understand why it would make sense when the only considerations are TTO and BiP. But the game includes far more than those alternatives, especially in terms of the pitchers role in defense. I’m going to run through a very rough thought experiment of an alternative approach to assessing how much pitchers are in control total team defense.

                A pitcher, in his role as pitcher, is “in control of” only one thing: pitches. For a pitcher, the unit of play is the pitch, not the out or safe base. In a routine game let’s say a pitcher (or all team pitchers), as a pitcher, makes about 100 pitches. Call each one a “quantum” of play for which the pitcher is entirely responsible (I’m going to duck the issue of catcher/manager pitch call and catcher framing). If we designate quanta of play as ‘Q’, that means the pitcher, as a pitcher, is responsible for 100 Q. Whether the pitches are strikes, balls, good pitches or bad is not relevant to this issue of being “in control of” play.

                On offense, batters are the exact complement of pitchers, and, in a 100-pitch game, would collectively account for 100 Q. In this respect, a taken pitch, be it a ball or strike, is just as much one unit of batter “control of play” as a HR or DP grounder. Like pitchers, batters are not “in control of” outcomes, only their inputs. (Although, of course, the product of pitcher and batter inputs is the TTO or BiP outcome. But that product in “in the control” of no one player.)

                It’s a little different with fielders in two respects: (1) we can call every “chance,” whether assist or putout, 1Q, but there are times when players fail to field easily fieldable balls (and commit no scoreable [sp?] error); (2) tough chances made add a premium degree of difficulty to “in control of,” and tough chances missed require some discount in responsibility because the player had diminished control/responsibility; (3) there are some defensive plays that are not chances — fielding a clean signal; receiving an outfielder’s throw; etc. — but these are still defensive contributions . . . perhaps “fractional Q.” Nevertheless, for simplicity, let’s call each chance a 1 Q opportunity for each fielder involved, understanding that there are vast ranges in degrees of control (which may even out). (This is all in contrast to the pitcher and batter: the pitcher is always in full control of his pitch (gripping the ball and throwing it to the catcher — well, balks are an exception), although the quality of his effort may vary; the batter is in control of his at-bat: standing in, assessing pitches/responding with a swing or not.)

                I believe there are, on average, 38 PAs per team-game. Making up the rest of the game stats, aiming for intuitively normal levels and avoiding issues like fieldable foul pops, SB/CSs, DPs, SFs, non-ball related fielder moves (like holding a runner, backing up a play, etc.), and so forth (because this post would become endless and its author even more confused), lets assume, in a 100-pitch game, the 38 PAs generate 6 strikeouts, 3 walks, 1 HR, and 28 BiPs — of which 8 are hits (with two minor fielder involvements), 7 are air outs, and 14 are ground outs, requiring two fielder involvements. Basically, the pitcher has control of 100 Q, fielders on the order of 35 Q in assist/put-out contexts and perhaps 20-25 additional fractional Q.

                This simple model suggests that even in a pretty low TTO game, the pitcher has considerably more Q than the rest of the defense.

                Now, we can throw away all the details and focus on the main point: every pitch is a play, and all pitcher plays are part of the total defensive effort. I think your approach basically throws out every pitch that does not end a PA, but those pitches are precisely equivalent in form, effort, and skill to pitches that do end PAs — and they have concrete outcome consequences, as differential batting averages on various pitch counts demonstrate. (Beyond, that, pitch sequencing itself can disrupt or fail to disrupt batter performance, and many apparent failures — balls — are actually successes, pitches that set up an ensuing strike. Poor sequencing can also lead to an apparent success — a strike — leading to failure, if it makes the following pitch guessable, or seem more hittable.)

                So I think that if you see the only elements of a game as PA-ending TTO or BiP outcomes, your estimate may work, but that this (a little like fWAR) writes off more than half the game from pitchers’ and batters’ perspectives.
             6. Dr. Doom February 26, 2019 at 11:45 am

                It’s an interesting thought experiment, Bob, but I think your method is far too speculative.

                The point of baseball is to win games. Winning games means scoring runs. While you might say that a pitcher has control of “100Q,” not all of those 100Q are created equal. Each third strike, fourth ball, or BIP has anywhere from, say five to fifty times as much impact on the final score of the game. A single ball thrown in the first inning to the second batter on a 1-1 count might, yes, be part of the larger drama of the game; it may set up what comes next. But ultimately, the result of the game is not different, nor likely to be different, if that ball is, in fact, fouled back over the net and into the stands. Yes, a 2-1 count and a 1-2 count may yield something different, and could change the outcome of the game. That’s true; but we’re not in the business of speculating about might’ve-beens. We’re attempting to assess value. The result of a single pitch, less it ends the PA, is essentially a non-event. You might argue otherwise, and I expect you will. But, beware that, to maintain intellectual consistency, I have NEVER read you write anything about pitch-by-pitch analysis; I have never read a word you’ve written on what sorts of counts these pitchers worked in; I have never read one word about effectiveness in various situations. Like virtually all analysis for HOF cases (and COG cases and the like), your focus, Michael’s focus, and mine – as well as basically everyone else in history, from Bill James on – has been interested in the pitches that end the PA.

                Not all pitches are created equal. This is not the same as the leverage index, although perhaps one could make that argument. What makes a pitch have any value, positive or negative, is its effect on run-scoring. Fouled-back pitches, called/swinging (non-third) strikes, and (non-fourth) balls are non-events. The pitcher is “controlling” non-events. So… fine. You can argue that. But you can’t possibly expect someone to take seriously the claim that any given pitch is equally important as any given fielding opportunity.
             7. Bob Eno (epm) February 26, 2019 at 4:46 pm

                Well, Doom, first, I hope my own past comments don’t limit the ideas I’m allowed to have. The ideas in my post on Q’s followed on my thoughts about fWAR assigning 100% of control to pitchers on TTO outcomes, which also grew out of my belief that “chance” plays no role in BiPs. But certainly this is the first time I’ve written along these explicit lines and of course I know that this isn’t the way these issues are usually framed. That’s what a thought experiment is about, no?

                When you say, “Not all pitches are created equal” I think what you mean to say is that not all pitches lead to equally impactful outcomes. That’s certainly true. But I’d argue the pitches are indeed equal and the pitcher’s contribution to the defense is identical in magnitude in each case. With every pitch, the pitcher puts the team defense on the line, and since the outcome cannot be known at the point of release, and every pitch risks one or more runs, from the pitcher’s point of view he needs to mobilize the same level of skills for every pitch (within the framework of game context, and in that sense, it’s true that a late inning pitch in a 10-0 game differs from one in a 2-2 tie). A 5.0 WAR-type pitcher needs to aim (realistically, since he has the skill) for a 5.0 WAR-type pitcher’s pitch with every pitch, or he’s not going to be a 5.0 WAR pitcher for long.

                The model we normally use counts a pitcher’s defensive effort only as a consequence of something successful or unsuccessful the batter does (fourth balls that are not close enough to create a judgment problem for batters would be an exception . . . and we could get into HBP too), and if what the batter does isn’t a PA-ending act, we view the pitcher’s effort as a non-event, just as you describe it. But it isn’t a non-event. It’s a defensive event that we ignore when we reduce the game to stats after the fact.

                If the creators of the first box scores had included a column for total pitched balls and strikes (something MLB Game Day now tracks in real time), we’d probably view pitcher games very differently from the way we do now (and statheads would long since have refined ball and strike counts into more meaningful categories). We don’t look at games that way, but when Michael argued that pitchers account for only about 1/4 of team defense, having been thinking about the pitcher/batter interplay in terms of my complaints about fWAR, I noticed that we could think about games this way, and perhaps we should. These defensive events — non-PA ending pitches — really do occur and really do have game consequences: the pitcher faces risk with each one. I’m not sure how interesting it would be to extend this idea to other facets of the game, but in the case of measuring the pitcher’s share of team defensive contributions — the topic of Michael’s comment — I think we have to take it into account. (Emphasis on “I think”: it’s still in the thought experiment stage for me — I realize you felt the experiment was done and the results were negative.)
             8. Michael Sullivan February 27, 2019 at 7:47 pm

                So Bob, first of all, I was referring to *both* sides of the ball. So to say that the pitcher is responsible for 1/4 is to say that he is responsible for roughly half the defense.

                I do also agree with Doom about using pitches vs. PA results.

                I think you’re right in some sense to say that each pitch in a given PA is of roughly equal importance in some fashion, but the end result happens on a particular pitch, that is either a fourth ball, third strike, HR or BIP. You could break down the component pitches for any given PA, but the total value of the pitchers has to come down to the total value of the final result. If it’s a walk, all those pitches need to add up to the negative run value of a walk (or that particular walk in that particular situation (RA9 or RE24).

                If it’s a TTO, then I think you clearly say that the defense for that PA is *all* on the pitcher (unless/until we can measure catcher framing/calling).

                But if it’s a BIP, I think it’s *crazy* to give the pitcher full credit. I don’t think it’s even realistic to give pitchers half credit for defense on BIP, as very few pitchers have consistently good or bad BIP results, and the swing is pretty small over the long term for those who do.

                If we split the credit evenly between pitchers and fielders (on balls in play only), then pitchers get roughly 2/3 of the credit for defense. If it’s 1/3 pitcher, 2/3 fielders, the pitchers get credit for 5/9 of the defense or a bit over half. 1/4 of the game gives them a little less than that, which is the rough stance I took in my previous comment.

                If you want to say it’s really somewhere closer to 1/3 of the game (they should get half credit for BIP), I disagree, but I don’t think your position is unreasonable — I’d have to do more work on the spread of good vs. bad fielding numbers vs. the range of pitcher BiP results and try to pin it down better. But I think it’s crazy to think the pitcher is more than 1/2 responsible on BiP. Some sabermetricians have studied this a lot and are pretty sure it’s a lot *less* than the 1/4-1/3 on BIP I’m implicitly giving credit for by saying they are responsible for 1/4 of the game result.
             9. Bob Eno (epm) February 27, 2019 at 9:20 pm

                Michael, I think the clarifications you make here actually get us close to agreement. Now that I’m re-reading your initial message, I see I did not connect the dots that would make the 1/4 figure cover both sides of the game. I’d grant pitchers a little over 1/3 on BiP (most defensive plays involve 1-2 fielder performances), and with my argument about non-PA-ending pitches I’d give them somewhat more than 2/3 of the total defense. That would yield the 1/3 of total team performance, reached by a different basis from the one you anticipated I’d argue. We still disagree, but I’d think the difference is much smaller than I anticipated.

                Now when you address differential W-L Pct. using my 1/3 figure (just ’cause this is my comment), what I’m saying is that if you look at interchangeable 1/3-impact units (that is, pitchers), and find that one unit consistently results in considerably more wins and fewer losses than the other units, the correlation is significant, so long as it aligns with advanced stats that would presumably flag outlier cases of pitching that is consistently stronger or weaker than game results suggest because of long-odds results from other elements of the defense or offense.

                When Lyons was that 1/3-impact unit, across a stretch of well over 3000 games, the White Sox won about 53% of their games; when other pitchers were that 1/3-impact unit they won under 46%. Check out Lyons’ WAR and ERA+ and they place his value in a similar ballpark. To me, that indicates that it’s valid to illustrate his quality impact by calculating his W-L Pct. differential.
             10. Michael Sullivan February 28, 2019 at 2:36 pm

                 I’m not going to say it’s “invalid” to look at W-L differential. It’s a stat that gets kept, and you can look at it, and it’s certainly going to give a better picture than pure W_L.

                 But that said, here’s where the rubber meets the road for me — What do I think if W-L differential and WAR disagree? What if, instead of roughly matching what WAR and ERA+ say, W-L differential suggested that Lyons was real inner circle material, a no-doubter COG selection like Maddux or Seaver. Or the other way — what if W-L differential made him look like a pitcher who shouldn’t even be in the COG conversation, more David Wells than David Cone (with whom Lyons is roughly comparable by the WAR based JAWS and HR)?

                 In either case, I’d be more likely to assume that some of “all else” wasn’t really equal. That perhaps he had often lined up versus better or worse teams or better or worse pitchers than the other pitchers on his team. Or that the offensive players or fielders on his team had underperformed or overperformend on average when he was pitching. To believe that W-L differential is telling you something meaningful, you have to believe it’s more likely that the raw pitcher W-L is saying something meaningful than that one of those things happened. And people have examined various pitchers and shows that even among same team, one pitcher can get significantly better performance even over a very long multiple year period from their offense versus another pitcher on the same team.

                 This is why I consider it dubious. At *best*, seeing a big gap in W-L differential vs. WAR might lead me to look in detail at what might have been different about the appearances of that pitcher vs. his teammate starters, and if I was able to look at everything and couldn’t identify *any* difference in offensive performance or fielding performance or who the opponents were, etc., maybe I would *start* to question whether W-L differential was telling me something. But my prior against it being particularly valuable info — given that we already know WAR and ERA+ — is fairly strong at this point.

                 In a world where we didn’t *have* WAR or ERA or RA for pitchers before a certain time, yeah I’d look at W-L differential, but IMO those other stats cover what it’s trying to cover *better*.
    2. Doug February 25, 2019 at 1:45 am

       To your point about penciling in Lyons for 300 wins if he doesn’t lose three years to the war, certainly it would have been possible, but he might have come up just short, as he was only starting 20-23 games each year, so the 13 or 14 wins that he would have needed would be about the most that could be expected.

       Lyons certainly thrived on that reduced workload, with a 141 ERA+ starting at age 37, compared to 112 prior. Only Wilhelm (153) and Grove (142) have better ERA+ scores in 500+ IP aged 37+. Lyons’ two best ERA+ seasons came at age 38 and 41, in the latter season becoming the oldest to lead his league in ERA, until Roger Clemens led the NL in 2005 at age 42.

       Reply ↓
36. Mike L February 25, 2019 at 9:15 am

    Ugh, this is a difficult set of choices. Allen, Wallace, and, for the first time for me, Ashburn.
    Once again, I won’t be voting in the Secondary ballot.

    Reply ↓
37. mosc February 26, 2019 at 12:09 pm

    Jeter got in so I guess I’m down to playing long shots (Nettles, Dawson, Simmons, Eck, and Irvin). I have to say I really don’t think Manny is a good candidate. I also compensate for segregation (docking Lyons, for example) and military service (favoring Lyons, for example) so I don’t think Steroids are a particularly unique thing we have to deal with. I guess at this point I’m leaning towards voting AGAINST Manny with a swoop-in vote for somebody like Lyons. I’ll have to see if I can pull that off. Maybe I should stay out of it and vote for the three I think deserve and let you all fight it out. Y’all did put Killebrew in after all.

    I want to go back to the redemption round and get Munson. Did you guys read Bob Eno (epm)’s vote? in the redemption round? I feel bad we forget details of baseball history like Munson and don’t have proper ways to balance out the intangibles that clearly impact the stats. Wouldn’t you feel better voting Thurman Munson instead of Manny Ramirez?

    Reply ↓
    1. mosc February 26, 2019 at 12:10 pm

       That post really requires an edit button. For shame!

       Reply ↓
    2. mosc February 26, 2019 at 12:35 pm

       I’m also slightly willing to consider some of Hideki Matsui’s Japanese league stats:

       .304/.413/.582/.996 over 5504 PA’s is pretty prolific.

       His age 30+ MLB stats compare less flatteringly with guys like Sheffield, Ramirez, etc. I can sorta buy the argument but I think we saw his best in 2004 and it frankly wasn’t that historic.

       Reply ↓
       1. Mike L February 26, 2019 at 5:23 pm

          Kind of an interesting point to raise, mosc, given that we look at years lost to service time and the color barrier.

          Reply ↓
       2. Michael Sullivan February 27, 2019 at 7:12 pm

          Yeah, I definitely took a look at Matsui, he was a very nice player for a bit and didn’t get to MLB until age 29. But I agree, his age 30+ profile just doesn’t look like that of a COG candidate, except maybe a few with monster early peaks and problems/injuries leading to mediocre late careers like Junior. Most of the guys who look like him age 30+ are HoVG or borderline big hall at best.

          I’d be more inclined to give him serious consideration if his 30+ was about the same or better than our borderliners. I think I’m giving his Japanese stats their due by doing that. It’s not like he was Barry Bonds over there and had a known injury some something.

          Reply ↓
          1. Bob Eno (epm) February 27, 2019 at 8:53 pm

             During our discussion last year of Ichiro’s candidacy (which I supported), there was a lot of talk about taking his years in Japan into consideration in some way. My feeling was that that approach was not valid in a case like Ichiro’s because his long stay in Japan was by choice, not an enforced limitation.

             Military service and segregation-based exclusion are examples of things for which we generally allow some sort of supplementary credit. On this string, we see that thinking in the cases of Lyons and Irvin. Those cases are based on institutional limitations, either completely beyond a player’s control or, in the case of the military volunteers, acts of self-sacrifice in a player’s citizen role. (An interesting case of this type is George Davis, who lost a year because he was a casualty of the AL-NL “war” of 1901-2. He received consideration for that because he was qualified and ready to play, but institutional forces prevented it.)

             There are cases which are less clear: Coveleski, who found himself trapped in the PCL for a few years because of management decisions, would be an example. Coveleski lost years of his MLB career because of forces beyond his control, but the reasons were individual, rather than institutional. Thurman Munson would be a different kind of example.

             But when players like Suzuki or Matsui remain in Japan after they have demonstrated exceptional “minor league” type credentials which open the option of an MLB contract, the loss of MLB career years is a choice that they make, not one enforced upon them. Their great careers in Japan explain why they came to the Majors late and bloomed immediately, but my feeling is that they should neither get extra credit for their successes in Japan nor compensation for shortened MLB careers. They chose their trade-off.

             One type of extra consideration falls outside these general categories: the “catcher’s bonus” is based on different reasoning. In general, catcher’s choose their position, but we give them a credit boost nevertheless. However, that model is clearly not relevant to cases like Matsui’s. Apart from that one exception, I think the general rule should be that “extra credit” compensation of any kind should apply only to circumstances outside a player’s control.

             Reply ↓
             1. Michael Sullivan February 28, 2019 at 3:32 pm

                If you’re really not taking Japan into consideration at all, I don’t see how you support Ichiro for the COG, He’s 59 WAR and 23 WAA with a hall rating of 108. That’s not really a COG resume. close maybe, but he’s behind a slew of guys with limited support, and a couple who made it that many of us think were poor selections. There’s nobody even at the outside border of the COG conversation that he’s significantly better than by stats in MLB.

                But I think he gets in, because most of us see he got to MLB at 27 as a potential MVP level performer, and had monster stats in Japan for years before that, such that it’s apparent he would have had a few solid years in the majors had he come to the attention of scouts at 18-20 and decided to pursue a major league career. And that’s enough to make him a pretty clear COG candidate. To my mind you’re implicitly giving *some* credit to japan by supporting Ichiro Suzuki.

                That said, I tend to agree with you vis a vis Negro Leagues and wartime. I’m willing to basically assume every benefit of the doubt for people who were barred from the league for no good reason, and a conservative projection of performance for war years. I’m not sure I’m wiling to go that far for Japanese league players, but I think the ones who came here and performed at a COG level from age whatever on, probably should get *some* credit. Since Ichiro doesn’t need much, I think he’s a fine selection.

                Matsui, OTOH, needs the kind of credit we are possibly giving to Monte Irvin and *more*. Irvin came into the league a year later, at age 30, with age 31 his first full MLB season. He accumulated roughly the same WAR as Matsui, in less than 60% of the PAs, while generating more than 10 WAA vs. Matsui’s 3.3 (10.2 v. 7.7 if you ignore -WAA seasons). That’s a HUGE difference. And you still basically have to fill in his prime to get Irvin to COG level. And he was a year older, 50+ years earlier.

                I realize we’re both saying no on Matsui, but I was really trying to say — even if you gave him the same level of BOD that I want to give Irvin, he’s still not really there. And I agree with you that he shouldn’t get the same BOD as Negro League players.
             2. mosc February 28, 2019 at 3:51 pm

                Totally agree with you Mr. Sullivan
    3. Hub Kid February 26, 2019 at 1:35 pm

       mosc, if you are looking at the current contenders to win this ballot and are not interested in voting for Manny, let me push for Allen: as Doom rather eloquently puts it above, they are very similar candidates. And of course, Allen is not a recorded PED user.

       I think Allen also had to deal with quite a bit of racism (not to mention the reserve clause), and I think it all makes up for the shorter career.

       I love complete games (and career longevity, and “leaving something in the tank”), so I am sorely tempted to change my vote to include Lyons, but I’m not happy with the drop-off in Tiant’s votes so I am hesitant to take away my vote for Tiant.

       Reply ↓
       1. Hub Kid February 26, 2019 at 3:31 pm

          Apologies to eminent HHS-ers Bob Eno and Dr. Doom – I have confused my Manny Ramirez vs. X arguments. Doom’s case is for Kevin Brown over Manny, and Bob’s question is why are there so few Manny AND Dick Allen voters (making the argument for the similarity of their hitting abilities)…

          Reply ↓
          1. Bob Eno (epm) February 26, 2019 at 4:50 pm

             Yeah. I saw that and I thought: I’m moving up in the world.

             Reply ↓
    4. Michael Sullivan February 27, 2019 at 9:34 pm

       BTW, I’ve got your back on dumping Manny from the top or creating a runoff if it’s possible, just not planning to vote until tomorrow evening.

       Reply ↓
38. Paul E February 26, 2019 at 12:59 pm

    Regarding the Manny debate, his 154 OPS+ is impressive with a .312/.411/.585 39/129 w/94 BBs per 162 Games. Honestly, I believe his OPS+ might have been even higher if:
    1) He had juiced his whole career (starting in 1993) and had an even longer peak.
    2) The whole GD league wasn’t juicing. His raw numbers are out of sight but, relative to the league, not insane. In the ballpark of his 154 OPS+ we find Ott, Aaron, DiMaggio, Mays, Allen. I have to believe, as Caminiti (“50%” ) and Canseco (“80%”) suggested, more than half of the league was under the influence of PEDs.
    But, for voting’s sake:
    .302/.390/.552 career converted up to 4.25 Runs per game environment (ALLEN)
    .296/.393/.555 career converted down to 4.25 Runs / game environment (MANNY)
    129 RC/162G ALLEN (4.25 Runs)
    123 RC/162G MANNY (4.25 Runs)
    7.946 RC/27 ALLEN (Career RC/27/AIR)
    8.077 RC/27 MANNY (Career RC/27/AIR)
    8.534 RC/27 ALLEN (Peak RC/27/AIR)
    8.385 RC/27 MANNY (Peak RC/27/AIR)

    Pretty much splitting hairs except the one guy wasn’t on steroids and attempted to play a significant defensive positions for about 4 years.
    Looking back….was Bagwell juicing?

    Reply ↓
39. Josh Davis February 26, 2019 at 3:42 pm

    VOTE:
    Ted Simmons
    Ken Boyer
    Kevin Brown

    Secondary:
    Todd Helton
    Minnie Minoso
    Dennis Eckersley

    I’ve appreciated the conversation about Manny Ramirez and have chosen not to vote for him this round due to some of that, although I very well might in the future and think he would fit fine in the COG at this point. I still don’t see the fairness in docking Manny for PED’s when proven cheaters who didn’t serve penalties for their crimes (like Manny did) are already elected and others remain on the ballot apparently untarnished. I realize some of you have a more nuanced view, and can respect that, but dealing in hypotheticals is not something I’m excited about.

    Reply ↓
    1. mosc February 26, 2019 at 4:27 pm

       I just don’t see how we avoid hypotheticals. I understand and agree it’s not concrete but we have to deal with military service and strikes and segregation as well as more traditional differences in ballparks, run scoring enviornments, continuously increasing relief pitcher utilization, crappy defensive metrics, etc.

       It’s unavoidable.

       Reply ↓
       1. Josh Davis February 28, 2019 at 4:01 pm

          You are right, or course, to some extent this is unavoidable. But I’d still like to put more stock in what actually happened on the field rather than resorting to “what ifs.” Which is probably why I tend to support longer career guys.

          Reply ↓
40. mosc February 28, 2019 at 9:15 am

    As I look at the sheet, I see Sutton has tied Ramirez. I don’t like Sutton much either, his WAA is atrocious for a candidate and perhaps more than any other player in history, his appeal is pure innings. Looking at the guys a single vote behind, I get Allen, Brown, and Lyons. I don’t think I’d pick Ramirez over Brown but I don’t think I have to pick either. Allen and Lyons had fairly different careers. To be honest I think I’m overvaluing Lyons 1946 competence over 42 innings in projecting his 42-45 missed time but he returned as an above average pitcher at 45 and that’s not nothing.

    Lyons was also about 5 WAR better with the bat than Sutton, for what that’s worth. It was a measurable difference between the two.

    Michael Sullivan, if you’re out there I haven’t decided for you how this is going to go so GL to you sir.

    Primay: Lyons, Dawson, Nettles
    Secondary: Eckersly, Irvin, Randolph

    Reply ↓
41. Paul E February 28, 2019 at 12:12 pm

    Somewhat on topic, Kenny Lofton sounds off on the HoF and steroids:

    https://sports.yahoo.com/kenny-lofton-blames-alex-rodriguez-ped-users-omission-hall-fame-020906075.html

    Cliffs’ Notes: It appears he’s anti-steroids, dislikes A-Rod, and is further pissed about MLB’s stance on past known users.

    Reply ↓
    1. no statistician but February 28, 2019 at 1:54 pm

       It’s a point of view that doesn’t get much notice, but let’s say you were in a grad school class at a top flight research U, and half the other students had rumbled an easy way to bamboozle the lazy, tenured prof by buying research on line or a similar scam, while you did your own work, sweated the research and the grunt work, and, true, you got the A grade, but the cheaters got the A+. I’d guess you’d feel pretty resentful.

       From my own experience I know that , being a student back in the antediluvian past when such things mattered, what drove me crazy was a simple thing like students not being penalized for turning in work late. It was an insult to the rest of us who followed the rules.

       We’ve discussed this ad nauseum, I realize, but those who don’t care about the external issues surrounding the PED problem in Lofton’s era are basically saying that guys like Lofton were dummys for not jumping on the bandwagon, and the ones who participated have clean consciences. Their OPS+ jacked up, didn’t it, and hey, that’s what it’s really all about.

       Reply ↓
       1. Paul E February 28, 2019 at 2:01 pm

          ” Their OPS+ jacked up, didn’t it, and hey, that’s what it’s really all about ”
          Yes, and getting paid. Sosa and Bonds both landed sweet extensions from the Cubs and Giants….something on the order of 4 years/$80,000,000 – which is still, to this day, a lot of money for a free agent .

          Reply ↓
       2. Mike L February 28, 2019 at 3:58 pm

          NSB, I think Lofton’s overall case might have been hurt by the number of teams he played for. After his last full season in Cleveland in 2001 at age 34, he played for an incredible 9 different teams in 6 years–always useful.

          Reply ↓
    2. mosc February 28, 2019 at 3:31 pm

       If you describe a player who’s going to get undervalued more than anybody else, you’re going to describe Kenny Lofton. He’s like perfectly built for people who look at triple slash lines to not understand. He’s not a shortstop where position alone gives you a reputation (independent of reality) yet he was incredibly valuable on the defensive side. He didn’t demolish stolen base records but he certainly added value with the legs. Even hitting from the left side with his approach and wheels helped him stay out of the double play and advance runners where others would create more outs.

       A speedy natural lefty with a great glove, excellent pitch recognition, contact skills hitting the other way, and a decent arm has got to check every box of “you just don’t understand how good they are”

       I kinda agree with Lofton’s points.

       Reply ↓
       1. Mike L February 28, 2019 at 3:50 pm

          Lofton was a terrific player who gradually eased down to good, and then useful. But he was playing at a time when huge dinosaurs ruled the Earth, and is thought of as one of those B+/A- guys that help win games, but aren’t first rank stars. In a different Era (or an era where half the folks didn’t have needles in their backsides) he would certainly have done better than a one and done 3.2 % of the HOF vote.

          Reply ↓
          1. Paul E February 28, 2019 at 7:39 pm

             when you consider that it took Tim Raines 10 full years to get into Cooperstown, it’s probably not all that surprising that Lofton, though obviously inferior (by the eye test) to Raines, got so little respect from the BBWAA. Lofton was surrounded by Thome, Belle, Ramirez, Juan Gonzalez, David Justice, Travis Fryman, Alomar(s) and that might tend to obscure his contributions.

             Reply ↓
    3. Michael Sullivan February 28, 2019 at 3:48 pm

       Chicks may dig the long ball, but not as much as HOF voters. To those who understand the stats, his stood out. He’s in pretty much every hall out there except Cooperstown, including the COG which is about half the size.

       Reply ↓
    4. Bob Eno (epm) February 28, 2019 at 6:22 pm

       Lofton’s point seems to me exactly the right one to make: “I’m not saying what Pete Rose did was right, but his numbers that he put up were real numbers. If it’s all about numbers, guys who cheated the game shouldn’t be in.”

       Reply ↓
42. bells February 28, 2019 at 8:15 pm

    Main ballot: Wallace, Lyons, Allen

    Secondary ballot: Smith, Irvin, Randolph

    Reply ↓
43. Dave Humbert February 28, 2019 at 9:25 pm

    Vote:
    Main ballot: Ramirez, K. Brown, Wallace

    Secondary ballot: Eckersley, Coveleski, Petitte

    Disappointed that Wallace and Dahlen could not build more momentum despite their qualifications (especially Wallace’s primary value being post 1900 in the first place), and that those who support Brown but not Ramirez or Ramirez but not Brown are ensuring neither gets in and we continue the same tired arguments about suspected use/what to discount. At least Eckersley appears to be headed to the main ballot, it’s been a long time coming…

    Reply ↓
44. Michael Sullivan February 28, 2019 at 11:17 pm

    Decisions, Decisions. Put Lyons in for sure even though I think he’s probably my 4th pick on this ballot? Or set up a potential 4 way runoff that includes two guys I’d rank higher but also Manny.

    Well, here goes nothing:

    Primary: Allen, Brown, Nettles

    Secondary: Randolph, Reuschel, Irvin

    Reply ↓
    1. Bob Eno (epm) March 1, 2019 at 12:32 am

       Michael’s vote does create a four-way tie for frontrunner among Dick Allen, Kevin Brown, Ted Lyons, and Manny Ramirez. Don Sutton is one vote behind. Ken Boyer is poised to drop back down to the Secondary Ballot. If there is still a ballot uncast, it could determine the outcome. . . . Half an hour to go.

       Reply ↓
45. Doug March 1, 2019 at 12:30 am

    For any last minute voters out there, there are 30 minutes to go to break a three-way tie between Brown, Lyons and Manny.

    Reply ↓
    1. Michael Sullivan March 1, 2019 at 12:31 am

       I thought I was getting Allen in the tie as well?

       Reply ↓
       1. Bob Eno (epm) March 1, 2019 at 12:43 am

          Doug, You’ve got seven votes for Allen on your spreadsheet, which is what the other three have as well. My count is the same.

          Reply ↓
          1. Doug Post authorMarch 1, 2019 at 2:48 am

             You’re right. We do have a four way tie.

             Reply ↓
46. Bob Eno (epm) March 1, 2019 at 1:14 am

    Here are the vote totals I have for the Primary Ballot (they match Doug’s sheet):

    Primary Ballot

    21 ballots submitted

    ———————————————–50% (11)
    7 – Dick Allen, Kevin Brown, Ted Lyons*, Manny Ramirez
    6 – Don Sutton*
    ———————————————–25% (6)
    5 – Bobby Wallace
    4 – Andre Dawson*, Graig Nettles, Ted Simmons*,
    3 – Richie Ashburn*, Bill Dahlen,
    ———————————————–10% (3)
    2 – Ken Boyer*, Luis Tiant
    1 – Hideki Matsui*, Bengie Molina*,

    What I have for the Secondary Ballot differs from Doug’s sheet. The discrepancy is because Doug has me voting for Eck, Helton, and Irvin. I actually voted for Coveleski, Irvin, and Smith.

    Secondary Ballot

    20 ballots submitted

    12 – Dennis Eckersley*
    7 – Stan Coveleski*, Todd Helton, Willie Randolph
    6 – Monte Irvin*, Minnie Minoso*
    5 – Rick Reuschel
    4 – Reggie Smith*
    3 – Andy Pettitte,
    2 – Bobby Abreu
    ————————————10% (2)
    1 – R.A. Dickey*

    Voters: Voomo, Gary B, koma, Bruce G, opal611, Doom, epm, Jeff B, Andy, JEV, Hub Kid, Richard C, Paul E, Doug, Chris C, Mike L Primary only), Josh D, mosc, bells, Dave H, Michael S.

    Reply ↓