Best 25-Man Roster in MLB History

Sports Illustrated writer Cliff Corcoran has authored an article with this provocative title. This is the team he chose.

 

 

 

 

 

 

I’ll leave you to consider Corcoran’s arguments. Let the debate begin.

50 thoughts on “Best 25-Man Roster in MLB History

  1. David P

    In terms of position players, I don’t have any big quibbles, other than the fact that there’s a clear bias towards older player.

    But starters??? Yeah, I know it’s hard to choose but he used ERA+ as his metric, Which would be fine if there weren’t lots and lots of better tools available.

    Obviously the biggest problem is the inclusion of Kershaw.

    Kershaw has 1901.1 IP and 57.2 WAR and 42.1 WAA.

    Let’s just compare that to Tom Seaver.

    From 1969-1975, Seaver threw 1918.2 IP with 52.9 WAR and 36.8 WAA.

    So a slight edge to Kershaw.

    Except that 52.9 WAR from 69-75 isn’t even half of Seaver’s career WAR.

    Kershaw may belong someday but he was a LOT of work still to do.

    Reply
    1. Doug Post author

      The other odd thing about using the ERA+ yardstick is there are a bunch of pitchers with better ERA+ results than Johnson and Maddux. Of pitchers with 2500+ IP, Pedro ranks 1st, Johnson 11th and Maddux 14th. But, of the pitchers who were passed over with better ERA+ than Maddux, only Clemens and Whitey Ford pitched since World War II, and only Pete Alexander in the live ball era.

      Reply
  2. Dr. Doom

    First of all, who HASN’T done this, from time to time?

    Second, I think Eddie Collins ahead of Joe Morgan is an absolute travesty. I would’ve also strongly, strongly considered Jackie Robinson for that spot, simply because he would be the greatest utilityman in the history of baseball; the only problem is that he was way, WAY too good to be merely a utility guy.

    I like the inclusion of Kershaw. I think these lists are always more fun when we sort of try to project forward. I also think that A.) no one would’ve had a problem with Sandy Koufax being chosen, and B.) Kershaw has already had a career about as good as Koufax’s, so it tracks.

    I don’t get the reliever thing. If it were me,

    David P notes a bias toward older players… only the majority of the pitching staff are Greg Maddux or younger.

    The exclusion of Roger Clemens makes no sense to me, and I would agree with David P that Seaver belongs. But then, I don’t get why you’d have two one-inning relievers on a ten-man staff, and why you’d have two firemen. For my money, give me ten starters. Let Randy Johnson (or Kershaw, or Clemens, or whoever) “close.” They’ll be just as effective as Mariano or Kimbrel.

    My other quibble would be the inclusion of Hank Aaron. I love Hank Aaron and have nothing bad to say about him. However, it’s clearly a pick that values career over peak, and I’d have to say that I would prefer Mickey Mantle.

    Also, I just find it weird when we limit ourselves to “Major League” performance, and then act as if Ty Cobb is a valid pick by Oscar Charleston isn’t. I think it’s pretty clear they were players, and one got to play in the Majors and the other didn’t. I know the stats are easier to cherry-pick because of their ubiquity, but as I’ve gotten older, I feel like so much nuance gets left out of the discussion when we limit ourselves to only white guys pre-1950 when there were equally-deserving black players. I would probably have included Josh Gibson over Carter, and put Charleston on the bench alongside Cobb.

    Still, it’s a fun exercise. If any of you click over there, DON’T READ THE COMMENTS. Yowza. One of the things I love about THIS site is that people are respectful, and generally understand that, if you’re going to pick 25 guys, there are about 50 correct choices. It’s fun to discuss, but name-calling seems a little childish, and it’s basically nothing BUT that. They’re mostly Tweet-length and have all the nuance and depth of a sledgehammer. I think I need a shower after reading them. Ugh.

    Reply
    1. David P

      1) I think you can pretty much flip a coin with Collins vs Morgan. Both should be strongly considered, neither has a huge advantage over the other.

      2) I would have a problem with Koufax and given how long it took for him to get inducted into the Circle of Greats, I’m guesing I’m not the only one.

      3) Fair point re the inclusion of relievers. Particuarly since one of my pet peeves is when people say that closer is a position, It’s not, it’s a role and as you noted the top starters could fill it quite well.

      4) Never actually said that Seaver belonged. Just that, in my opinion, he rates far above Kershaw.

      5) Fair point re: Mantle over Aaron. Oddly the author said he favored peak, though it’s clear that he didn’t use any systematic criteria in making his selections. And what happens when Trout is someday eligible

      Reply
  3. no statistician but

    I’m sure Corcoran wrote the piece with the aims of stirring interest and creating controversy, so I won’t fault him for his choices, even though I don’t agree with his taste in pitching much at all. I’m biassed against Hornsby, but otherwise his starting lineup is very arguable. I’d take Yogi B and J Bench platooning at catcher. The “bench” is a turkey shoot category. Fill in Foxx, Morgan, Musial, Brett, or any of a dozen others and it wouldn’t matter.

    The real problem I see is that the roster concept is flawed. It’s a fantasy team that is based on statistics, not realities. And really it’s just his picks of the best players ever at each position, shoved into a 25-man format.

    We’ve done far better than that here at HHS in the COG, and I say this as one who is highly critical of the COG.

    Bill James picked the best lineups by decade in one of his books, and that, while also flawed, is a sounder and a more interesting approach, something we might try to argue out or refine upon here. But maybe I’m assuming more historical interest among other contributors than actually exists.

    Reply
  4. e pluribus munu

    Like others, I have trouble with the bullpen category here, but Corcoran builds that into the model he’s working with – he wants a roster structured like a contemporary one, and I don’t think it’s productive to argue with a premise of his project. If I were to do that, I’d argue that an ideal team has no DH.

    I’d make some different picks (and I like nsb’s idea of HHS folks making those picks by decade – although perhaps by “era” might be a more interesting and flexible way to approach it), but one thing Corcoran does very well is to anticipate objections and to allow that it would be easy to argue for Mantle, and other outstanding players.

    The one point of disagreement I’d have is with Kershaw, but for reasons different from those cited by others. I’m a great admirer of Kershaw and I root for him. But in this post-season-heavy era, Kershaw has a pretty long and utterly awful post-season record that has contributed to the Dodgers’ enduring Series drought, the longest in their history. Until Kershaw balances this with some real post-season success, I think his fabulous record is significantly tarnished. On the other side, I would never put on this list my actual all-time favorite outstanding player, Sandy Koufax (contra Doom and following David P), but if I were arguing for him, I’d say: look at his performance in the post-season and in the stretch of close pennant races – this is the guy you want on the mound when it counts most! None of the starters on Corcoran’s list comes close. I’d still pick Seaver or (ugh!) Clemens before getting anywhere near Koufax for this list, but I think the comparison illustrates why Kershaw is problematic. I think it will take a whole Kershaw-like regular-season career to overcome a current-level Kershaw post-season record when we’re talking about one of the five all-time best.

    Reply
    1. Doug Post author

      … look at … performance in the post-season and in the stretch of close pennant races. … None of the starters on Corcoran’s list comes close.

      I can certainly recall some memorable and gutsy post-season games from the Big Unit, and from Pedro (even in defeat and passed his prime).

      Reply
      1. alz9794

        The performance by Randy Johnson in the 1995 tiebreaker or play-in or whatever you want to call it game against the Angels was pretty spectacular. CG, 3 H, 1 BB, 12 K, 1 R (in the 9th when they were up 8-0 at the time) is not too shabby. There really isn’t much difference between that game then and a wild card game now, other than one is a regular season game and the other is a postseason game. The consequences are the same.

        And if we’re talking about regular season-postseason combinations, I’ll take Bob Gibson. 9 starts, 8 CG, 81 IP, 7-2 with a 1.89 ERA and 92 Ks with a 0.889 WHIP. And, according to McCarver, Gibson refused to come out of Game 7 in 1964 despite being pretty much gassed and getting rocked (gave up HRs to Clete Boyer and Phil Linz in the 9th), which increased his ERA. He pitched 100 2/3 innings in basically a month and a half (September 2nd through October 15th) in 1964 to push the Cardinals from 7.5 games back of the Phillies (somehow only got 2 points in MVP voting – less than Ruben Amaro) to World Series champs.

        I know Koufax has the prettier postseason ERA (0.95), but he gave up more UER in less IP and, despite the ERA, was “only” 4-3 in the postseason.

        Reply
        1. e pluribus munu

          I thought of adding a comment on Gibson, alz9794. There are plenty of really gutsy post-season performers on a par with Koufax, but they aren’t in the “five best” conversation, and Gibson seems to me the best of them (and not far off a “five best” type list). I don’t think Koufax should be on that list either, but since his name was brought up in that connection in earlier comments, I thought he’d be a good contrast to Kershaw.

          As for Koufax’s unearned runs, most come from the fifth inning of his final game, when consecutive batters reached on a two-base error and a three-base error – easy fly balls right to Willie Davis, who simply dropped. them There’s not a lot a pitcher can do about that. That was one Koufax loss: the others were in games of 0-1 and 1-2.

          Going beyond Koufax’s 0.95 post-season ERA, the comparative WHIP numbers are:

          The Big Train (50 IP): 1.420
          Mad Dog (198 IP!): 1.242
          The Big Unit (121 IP): 1.140 [and, yes, Unit had a terrific 2001 post-season; he was otherwise 2-8]
          The Great Pedro (96 IP): 1.080
          Koufax (57 IP): 0.825 [but where’s his nickname? No one ever actually said “The Left Hand of God” in a game]

          (Gibby’s was 0.889; Kershaw’s is a surprisingly reasonable 1.157.)

          Koufax played in a short-playoff-season era and all his stats are Series stats, like Walter, so his sample of IP is smaller than contemporary players’. That’s why it makes some sense to consider his performance in down-to-the-wire stretch drives as well, which are as close as you can get to DS and CS contexts. In the close ’65 and ’66 races (the only two close races where Koufax was a part of the rotation and uninjured), from mid-September (taking 9/9/65 as the earliest date) on, Koufax’s games over the two years had these scores: 1-0, 1-2, 1-0, 2-0, 5-0, 3-1 // 4-0, 5-1, 11-1, 1-2, 2-1, 6-3 (plus one three-straight-outs Save). Over the 107 IP covered by those games, Koufax’s WHIP was 0.729.

          Reply
          1. Michael Sullivan

            Looking at straight WHIP ignores the fact that Gibson and Koufax were pitching in a historically low run era something of a little deadball era), while all of the others compared here pitched somewhere between a big part and nearly all of their career in some of the highest scoring years. Kershaw somewhat less so, but post steroid ball is still a lot higher scoring than the late 60s. Koufax also pitched in a relatively extreme pitcher’s park. Koufax’s numbers are so extreme that it’s still phenomenal even in context, but I’m not sure his postseason results are really all that much better than Pedro’s, for instance.

    2. David P

      What’s strange is that Kershaw’s postseason numbers aren’t that different from his regular season ones:

      -1.1 more hits per nine innings
      -0.4 more home runs
      -0.3 more walks
      -0.8 more K’s.

      And yet somehow his ERA is 2.21 runs higher. Which strikes me as mostly a sequencing/luck issue. (I checked his doubles and triples and they’re about the same per nine innings).

      Reply
  5. ThickieDon

    I like Kershaw, and all the other pitching selections are top notch, but I would swap Clemens in for Kershaw.

    Reply
  6. no statistician but

    Another problem with this approach generally, if I can diverge from baseball for a moment, is the difficulty of overcoming biases. There’s always at least one assumption lurking in the background of lists like this. Most of us probably remember the craze 17 years ago for naming the 10 or 25 or 100 best whatevers of the Twentieth Century. In my view they all failed miserably in the areas where I have some interest and knowledge. Anyone, for instance, who really thinks The Great Gatsby is the best American novel of that stretch—or of all time!—either has not read the competition, is blinded by something (hype, a love of Scott Fitzgerald, the romance of it’s being the first book he or she ever really got into, etc.), or is incapable of understanding and appreciating fictional prose if it involves density and complexity and depth.

    Just recently I came across a webpage listing the 100 greatest classical symphonies of all time. Five of the top ten were by . . . Sibelius and many others lower down were by Hovhanhess. The desire, conscious or unconscious, to attribute extraordinary virtues to our favorites can be overwhelming, and the corresponding guilt of doing so often makes us give in to spreading the glory around indiscriminately to those we feel are preferred by other people, just to show our impartiality.

    Statistics, of course, are supposed to be impartial, but we all know that that is hardly the case. The interpretation of statistics is what we do here, and how much agreement do we find? Statistics are something of a map, but they aren’t the territory, aren’t the reality.

    Reply
    1. alz9794

      I don’t know anything about symphonies, so to me it’s unclear if Sibelius having 5 of the top 10 and Hovhanhess having many others lower down is reasonable, unreasonable, or just that someone likes Sibelius.

      It’s certainly possible that an extraordinary talent should dominate a list, and other greats make perhaps one appearance on a top 100 list, while others don’t crack it at all. There have been thousands of player seasons in MLB, even if one limits that to position players. Using WAR for position players, the top 100 seasons (really top 101 as there is a 13 way tie for 89th place with 9.4 WAR) break down as:

      10: Ruth
      8: Hornsby
      7: Mays
      6: Bonds
      5: Ted Williams, Cobb
      4: A-Rod, Gehrig, Mantle
      3: Yaz, Eddie Collins, Trout
      2: Ripken, Banks, Honus, Jackie, Morgan, Lajoie, Henderson, Musial, Speaker
      1: Beltre, Rosen, Pujols, Harper, Biggio, Brett, Sisler, Aaron, Foxx, Griffey, Larry Walker, Boudreau, Schmidt, Betts, Petrocelli, Yount, Carew, Santo, Sosa, Shoeless, Terry Turner

      It’s 42 different players, though half of them only appear one time. We can argue about the measure being used (WAR), we can argue that the measure doesn’t incorporate everything we would like (Hornsby’s probably not the team leader that Pee Wee Reese was, and that isn’t captured; defense isn’t measured the same over time, etc.), we can argue about the fact that I’m only using MLB seasons (no Oscar Charleston, no Josh Gibson, etc.) – but based on this measure Babe Ruth had 10 of the top 101 position player seasons in MLB history (really 6 of the top 13), which seems like something that would get a head nod from most baseball players.

      There are also great players who do not make the list: Mel Ott (8.9), Frank Robinson (8.7), Eddie Mathews (8.3), Roberto Clemente (8.9), and Cap Anson (6.9 – though his single season numbers are muted by the number of games played early in his career) are the 5 position players with the most career WAR who do not have a top 100 season.

      On the other hand there are some non-all-time greats on that list: Rico Petrocelli and Terry Turner had about a quarter of their career WAR in that one season. Generally they were in the 2.5-4.5 WAR range per season, but for one year everything seemed to align for them. Mike Schmidt also makes the list only one time, but if we were to extend out the list to the top 200 seasons he would pop up a few more times, whereas Petrocelli and Turner aren’t anywhere close to making another appearance in the top 500, probably not the top 1000.

      Reply
      1. Michael Sullivan

        If you do know something about Classical symphonies and composers, you’d know that while Sibelius is certainly great and worthwhile, and might deserve a couple entries somewhere on a top 100 list, he’s hardly vastly superior to *many* other composers in general. He’s more of a borderline COGer than a Ruth.

        But of course, classical music is a lot more subjective than baseball, so it’s certainly possible that one particular listener feels he stands out far above the rest, and who are the rest of us to judge their taste? It’s *very* clear, however, that their taste (in this case) is not very representative of classical music listeners, players or scholars as a whole.

        Reply
    2. Voomo Zanzibar

      Sibelius’ WAR was inflated by his home-piano advantage.
      And Hovhanhess virtuosity is muddled by suspicions of PEDs.

      Reply
      1. Mike L

        Sibelius basically ran dry the last 30 years of his life. No notable outlet. I’m not sure we could call him a Steve Blass, but essentially he may have run out of pitches.

        Reply
        1. Dr. Doom

          … but running dry for 30 years is not a surprise. There’s a great quote from David W. Barber, a music history humorist (one of the world’s greatest and oddest jobs) that goes something like, “Mozart set the standard for prodigies by dying young.” And there’s a lot of truth to that. Rather leave them wanting more than run dry.

          Reply
          1. Mike L

            That’s the variant on a Tom Lehrer line “It is a sobering thought, for example, that when Mozart was my age he had been dead for two years.”

      2. no statistician but

        The symphonies of Sibelius basically run in the 120-160 OPS+ range. Beethoven, Brahms, Mahler—the best ones go over 200 and the least make it to 150+. Saying that, by the way, isn’t meant to slight the great symphonies of Schubert, Schumann, Tchaikovsky, Dvorak, Shostakovich, Prokofiev, Bruckner and one-offs of people like Berlioz which also score high, or the best works of Haydn and Mozart, which, because of their era, are smaller scale gems.

        Bach, Handel, and Vivaldi, among others, predate the symphonic form. But I digress.

        Reply
        1. Dr. Doom

          I think all of us who are music nerds can agree that putting 5 Sibelius symphonies in the top 10 of all-time is a little bit like saying that Mel Ott had five of the greatest baseball seasons of all-time. No disrespect to Ott or Sibelius, but… c’mon.

          Reply
        2. Mike L

          My daughter is a Mahler enthusiast. I think he’s a little derivative–and maybe relies on only two pitches too much. The 8th has echoes of Beethoven’s 9th. But Brahms lifts from Ludwig as well, so maybe I’m being hard on him.

          Reply
          1. Mike L

            Rick. He was a righty, and as everyone knows, there are virtually no left-handed conductors.

          2. Dr. Doom

            I doubt that I’ve ever more enjoyed a thread on this site than this lovely classical music diversion, ending in a Rick Mahler reference. I knew there was a reason I loved this community.

          3. Dr. Doom

            It’s a lovely piece, Mike, and it was odd to think that your last name might actually not just be the letter “L.” Agreed with just about all you said. As a funny anecdote, I went to a Lutheran college, sang in I think 9 or 10 different choirs before I graduated, as many as 6 in one semester. Anyway, we had this big Christmas extravaganza every year – 200 vocalist choir, full symphonic orchestra, art on the walls and hanging from the ceiling of the chapel, poetry readings, dancers – the whole nine yards. It’s like a 2.5 hour thing. Anyway, there’s audience participation – mostly Christmas carols, usually only one or two. So, one year, they had broken it down as to who was to sing: verse 1, all; verse 2, musicians; verse 3, singers. The “singers/musicians” dichotomy is still a favorite with my parents.

          4. Mike L

            You mean “Doom” is not your real last name? Great story about Christmas concert. And I can see it. While there are some double major vocal + liberal arts students at conservatories that are part pf colleges/universities, vocalist/musicologists are fairly rare. Hopefully you still sing.

          5. e pluribus munu

            Second Doom’s comment on your essay, Mike. I’d like to know whether there was a particular thought (yours or your daughter’s) that led you to link to the Op. 131 in particular, apart, perhaps, from the connection with Schubert.

          6. Mike L

            Thanks, EPM. I was trying to make a point that some of the greats were willing to explore different structures and to innovate.
            My daughter sent me Op.131 to show that Beethoven does this a lot–he mixes genres, or cross-pollinates styles. He put a typical opera styles in his string quartets (movement 3 of op. 131) and again more famously, in the 4th movement of the 9th symphony. Also, Op. 131 just sounds different–not that “Beethoven-y” and textured in a way that feels almost modern, but not remote, like some modern music can be. Beethoven was willing to experiment.
            To bring it back to Doug’s post, the bigger issue comes down to relevance and appealing to a younger audience. Classical music and baseball suffer from some of the same issues—an aging fan base, very high cost to view live, and an embedded history that makes the past better than the present. Look at the Best-25 Man Roster. Loved reading it–but it’s what baseball fans do–compare Mozart to Beethoven to Brahms.

          7. e pluribus munu

            Thanks, Mike. That’s a clear thought on Op. 131. Beethoven was mixing genres early, like his send-up of Italian opera in the adagio of Op. 31/1, but I really like your notion of the late quartets as “modern, but not remote.” I think Beethoven often has an unusually effective interplay of reanimated “classical” elements in a highly innovative context, something that “post-modern” composers now try to achieve, but generally with less success. (And with Beethoven, the late quartets and sonatas retain their modern/post-modern edge almost 200 years later!)

            In a baseball context, I’m not sure that there may not be some similar elements. It’s true that baseball is uniquely backward looking among sports, but I’m not sure it’s in the sense that the past is always better. In the early ’60s, for example, when the stolen base first became headline news (although it had begun its revival a few years earlier), there was a sort of post-modern sense of reviving a classic art in the context of a new era. Ty Cobb and the long-forgotten Bob Bescher became hot topics, but not nostalgic: the point was that Maury Wills was going to outdo them. (Of course, the year before, the reactionary Ruth-celebrant Ford Frick had done all he could to ensure that Maris and future 162-game schedule players would forever fail to be Ruthian.) I think that contemporary fans, especially the aging ones, continually play with this tension of past and present: hoping to see records broken by new players, while hoping to see the greatness of the idols of their youth confirmed. When I look at the Best-25 list, it seems to me I see a pretty even distribution of eras, stretching from Wagner and Cobb to Kimbrel and Kershaw, and the article looks forward to a future with Mike Trout on the list.

            Overall, I think Bill James put an end to the relentless nostalgia of the Robert Smith-style “Golden Age” view of baseball’s past. I think HHS is a good distillation of how the “aging fanbase” comes to grips with the ambiguities of quality across eras in a way that makes the tension between past and present part of the texture of interest. (However, that does not apply to baseball announcing, at least in my case, where the classic masterpieces of Barber and Harwell completely overshadow the contemporary works of Knowitall, Ego, and Hype.)

          8. Kahuna Tuna

            But Brahms lifts from Ludwig as well, so maybe I’m being hard on him.

            If your [lineup] card includes two masters like Johannes and Ludwig, you’re not going to mind if they copy each other occasionally.

        3. Michael Sullivan

          I’d put a couple of those “smaller scale gems” from the classical period up with anything from BBM. Just because the lack the bombast and scale of the romantics and post-romantics doesn’t mean they aren’t every bit as worthy.

          Also too on your point that looking at symphonic form only leaves out a WORLD of incredible music. If I were to make a top 100 of my favorite/considered-greatest musical works, at most 5-10 of them would be symphonies.

          Reply
          1. Mike L

            I have a couple of personal smaller-scale favorites. Mozart’s Rondo in D (KV 382) and his Clarinet Concerto (622). And Dave Brubeck’s “Take Five”

    3. Paul E

      N S B
      “The desire, conscious or unconscious, to attribute extraordinary virtues to our favorites can be overwhelming, and the corresponding guilt of doing so often makes us give in to spreading the glory around indiscriminately to those we feel are preferred by other people, just to show our impartiality.”

      Men of merit exist in every generation but mankind, in general, prefer the meritorious of their own generation. But, tough to tell if “guilt” is responsible or it’s just bad opinions.

      My quibbles? I agree – give me 10 SP’s. Morgan over Eddie Collins. Ted Williams in LF; Frank Thomas at DH, Mantle has to be somewhere (ahead of Aaron?) Berra ahead of Carter and not real crazy about including the steroid guys (BONDS vs. Musial; CLEMENS vs. Seaver; AROD vs. Mathews/Brett/Ripken) over reasonable alternatives.

      Reply
        1. Steven

          I’d go with a starting rotation of Lefty Grove, Bob Feller, Sandy Koufax and Walter Johnson. 1245 complete games, there. Three relievers: Rollie Fingers, Hoyt Wilhelm and Goose Gossage. All can throw multiple innings. That frees up space for both Musial and Pujols. I’d keep three catchers: Bench, Berra, and the perpetually overlooked Ted Simmons. No Carter. I’d take Roberto Clemente, too.

          Reply
  7. robb

    I have a couple of small quibbles, mainly for including Bonds but not Clemens. He makes the case based on Bonds before 2001 to exclude PED years, but then includes the 48.5 war after that in his argument to get him on his all time roster. For THIS list, I’m not certain “favoring” peak does it. Give me Cobb, Aaron or Musial. If PED doesn’t disqualify then you have to include Clemens, with the third most pitching WAR, and just below Tom Seaver pre-2001. He also had a peak run of 7 seasons ave 8.2 war and a later mini-peak of 9.3 for 3 years. Both could be in HOF based on pre-2001, just don’t think Bonds makes THIS list w/o Clemens. I’m very upset that either used, and think MLB and the HOF have to come to terms with it some way, but I don’t think you can include one but not the other for a list like this.

    Reply
  8. oneblankspace

    I’d put Cal Ripken (Jr., of course) in there ahead of ARodriguez. Maybe ahead of Wagner.

    And maybe Rickey Henderson ahead of Kimbrel. With those starters, you don’t need an extra large bullpen.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *