Remember me

Explaining the 50 Superstars Project

September 21, 2017
                                           The 50 True Superstars


              OK, let me try to explain why I was trying to identify 50 true superstars.  About 3,000 people responded to some of the polls, so I wanted to say first that I appreciate your help, and I appreciate your responses.   I have a legitimate research purpose here, I think.    This is a spinoff of a spinoff of a spinoff of a study that didn’t work, but that’s kind of the normal pattern of research; you start out studying A, you wind up studying X. 

              The question that started this line of research was, "Is it true that the comings and goings of managers tend to align with the turning points of a franchise?"  I believe this to be true; I believed it to be true initially and I believe it to be true now, but I thought maybe I could demonstrate that it was true.   It appears that I can’t.  

              I constructed franchise strength histories designed in such a way that we can identify objectively the turning points in each franchise’s history.    Then I was entering the points where new managers were hired, with the goal of determining whether it was true that the points tended to align.

              I got about 25% of the way through that process and decided that it wasn’t going to work.   I don’t know; maybe I’ll finish it and maybe it will work, but it doesn’t seem likely.    New managers get hired and fired at all different moments of the season; sometimes it’s hard to match a manager with his first season.  Sometimes effective managers are hired with quite a bit of talent on hand and things start to click immediately; sometimes it takes a year or two.   Managers work on pretty short cycles; often they get fired after one or two years.   Because the cycle is short you don’t have a lot of leeway in the matching process.   Also, bad organizations change managers a lot more often than good organizations do, and this screws up the data.   You can’t really see the effect of a strong manager on a good organization because so many bad organizations are just churning pointlessly through managers, expecting managers to fix problems that they really can’t do anything about.   I still believe the underlying proposition is true (that the hiring and firing of managers are often the turning points for organizations), but I don’t think I’m going to be able to prove that it is true.  That’s alright; I can’t prove that the brakes in my car will work, but I still use them when I need to slow down.  

              Anyway, while that didn’t work, the data describing the strength of each franchise over time was really interesting, and I am planning to write a series of articles about that, about the cycles and turning points in the history of each franchise.    But while I was doing that, I noticed—DUHH—that organizations often turn upward when they come up with a superstar or when that superstar enters his prime, and often head downward when their superstar starts to fade.    The Yankees, for example, have a long, long "up" cycle that runs through 1964, ends in 1965.   What happened in 1965?

              Mickey Mantle got old.   In ’64 Mantle didn’t win the MVP Award but could have, as was normally true.   In ’65 he wasn’t the same player.   The Yankees had stayed strong from 1920 (and a little before) through 1964 by replacing Babe Ruth/Lou Gehrig with Joe DiMaggio, replacing Joe DiMaggio with Mickey Mantle, surrounded by Yogi Berra and Whitey Ford.    When Mantle got old they had several candidates to be the next one, but for one reason or another they all went sideways, so there wasn’t anybody there.  

              That isn’t ALL that happened to the Yankees at that time.   The organization was sold to a corporation, there was chaos in the front office, and there was a lot of other stuff happening as well, so it would be wrong for me to simplify their history to "they lost their superstar."   That’s just one theory.   As one of the participants in the poll correctly pointed out, there ARE teams that have very strong periods with no real superstar; he cited the Dodgers of the 1970s.  

              But I noticed that a lot.   The Red Sox have an "up" cycle from the time Yawkey purchased the team until 1950, a down cycle from 1951 to 1966, an up cycle beginning in 1967.   What happened in 1950?   Ted Williams broke his elbow in the 1950 All Star game.   Although he was still a fantastic hitter after that, he was in and out of the lineup the rest of his career, and drove in 100 runs only one more time (1951) whereas he had been driving in 120 every year before then.    What happened in 1967?   Yastrzemski emerged as a superstar.   Same thing with the Cardinals in the same era; they drifted into a downward spiral after Musial was no longer a superstar; they went back upward when Bob Gibson emerged as a superstar.

              Well, OK, let’s study THAT, then.   I think I can make some sense out of THAT data—if I have a list of superstars.   I mean TRUE superstars; I mean Mickey Mantle, Ted Williams, Stan Musial type of superstars, not Albert Belle, Gary Carter, Al Simmons and Ron Santo type superstars.   I mean the real deal.  

              Once I have a list of superstars I will need to write a formula to mark when each player entered his prime and when he left his prime, but I can do that by the numbers.    So now I have ANOTHER research project:  Define a list of the 50 players in history who should be described as true superstars. 

              Well, let’s see.   I’m not doing the 19th century; I don’t think 19th century baseball is actually major league baseball, but even if I did it is not amenable to the study because of the chaos in the schedule.   New franchises were created and died almost every year in the 19th century; teams lurched from extraordinary good to absolutely terrible in two years.   The organization rules of the 20th century don’t apply to 19th century baseball, so it is hard to study one with the assumptions of the other.

              Since 1900 we have 12 decades of baseball; we are working toward the end of our 12th decade.  For purposes of this study I need a reasonably even number of superstars in each decade.    This isn’t going to work if I have twelve superstars from the 1950s and two from the 1960s.   So, 12 decades, 50 superstars; that’s about four superstars per decade.  

              Some decades it is obvious.   The 1940s, you’ve got Ted Williams, Stan Musial, Joe DiMaggio, Bob Feller; we’re done.    Hal Newhouser; sorry.   You were great, buddy, truly great, but you’re not right for this role.   Ralph Kiner had some huge seasons, but.  . not right.  

              So I tried to draw up a list of 50 players like that.   About 40 of them are obvious.    The last ten are hell on earth to decide.   On the list of 40 players I am going to give you in just a moment there are about 36 that no one reasonably would argue with, and about four that some of you would argue with.   I’ll boldface the ones you might want to argue about and discuss those in a moment, but first the list:

1900s—Cy Young, Honus Wagner, Nap Lajoie

1910s—Walter Johnson, Pete Alexander, Ty Cobb, Tris Speaker

1920s—Babe Ruth, Rogers Hornsby, Lou Gehrig

1930s—Lefty Grove, Jimmie Foxx, Mel Ott

1940s—Joe DiMaggio, Bob Feller, Ted Williams, Stan Musial

1950s—Yogi Berra, Roy Campanella, Willie Mays, Mickey Mantle

1960s—Henry Aaron, Bob Gibson, Sandy Koufax

1970s—Joe Morgan, Tom Seaver, Reggie Jackson, Johnny Bench

1980s—Mike Schmidt, George Brett

1990s—Greg Maddux, Randy Johnson, Roger Clemens, Ken Griffey Jr.

2000s—Barry Bonds, Derek Jeter, Alex Rodriguez, David Ortiz

2010s—Clayton Kershaw, Mike Trout


              Those 40 players include 12 pitchers, 3 catchers, 2 first basemen, 3 second basemen, 2 third basemen, 3 shortstops, 3 left fielders, 7 center fielders, 4 right fielders, 1 designated hitter.    Yeah, I know that some people will argue about Yogi Berra, Roy Campanella, Sandy Koufax and David Ortiz, but I know what I am looking for when I say "true superstar", and those guys fit.   What I am looking for is players who have such an impact (on the won-lost total) that it could reasonably be said that their team is very different with them than without them, not because of the character of their impact but because of the scale of their impact.   Roy Campanella won three MVP Awards; that’s a real high-level impact on the pennant race.    David Ortiz. . . .look, I was there; I know what happened.   I am more convinced by my own experience than by your statistical analysis—or mine.  

              Well, but who else fits that description?   We’re looking for 50, and these 40 are not near the margins in my opinion.   But when you get close to the limits, it gets brutal.    It becomes nearly impossible to say who deserves those last ten spots and who doesn’t.   Sure, I could choose 80 "true superstars" rather than 50, but

              a) That means that I am studying players who have less impact, and

              b) Choosing 80 players isn’t easier than choosing 50; it is actually even harder.   

              The difference between the 80th player and the 85th player isn’t greater than the difference between the 50th player and the 55th; it is less.   Anyone who has any experience in mathematical analysis knows instinctively that that is true.   There’s an illusion that you can avoid the problem by including more players, but you can’t; you just have to deal with the problem at a different point, when it will be even harder.  

              I was trying to identify those last ten players who should be described as true superstars, and I just couldn’t get there.    I started with 20 candidates, and the next thing you know I had 48.   Then I realized that I had forgotten Rickey Henderson; I was thinking about whether Tim Raines should be listed, and I realized "Oh, my God, I forgot Rickey Henderson."   

              So I decided to ask for your help, and 3,000 or more of you did help, and I appreciate your help.   I got a couple of dozen bitchy tweets about how mean it was to do this and pointless and the people who were voting were ignorant and public polls are stupid, but what do you expect; it’s twitter.   I’ll review the results of the polls in a moment, but first I want to point out directly how extraordinarily high the bar is to be listed in the fifty.    It’s the top 20 or 25% of the Hall of Fame.    Eddie Collins.  . .I love Eddie Collins.   Eddie Collins is a true unappreciated superstar.   He had 3300 hits, 700 and some stolen bases, numerous other exceptional skills, and he ranks with Joe Morgan, Jackie Robinson and Ozzie Smith as one of the greatest percentage players ever. 

              Eddie Collins not only didn’t make the 40 players on the list; he didn’t even make the list of 54 players who are candidates for the last ten spots.   I’ve got four superstars from that generation; I’m not going to five.   Among players who are not even candidates for the last ten spots are Jeff Bagwell, Craig Biggio, Orlando Cepeda, Tom Glavine, Juan Gonzalez, Vladimir Guerrero, Todd Helton, Juan Marichal, Roger Maris, Eddie Mathews, Willie McCovey, Paul Molitor, Hal Newhouser, Phil Niekro, Dave Parker, Mike Piazza, Kirby Puckett, Tim Raines, Jim Rice, Ivan Rodriguez, Ozzie Smith, Duke Snider, Willie Stargell, Bruce Sutter, Bill Terry, Arky Vaughan, and Hack Wilson.  

              When I did the twitter polls yesterday, the most common feedback I got was that "your list is totally bogus if you don’t include BOTH of these guys."   Frank Robinson and Carl Yastrzemski; they’ve both got to be there or your list is not right.  Jackie Robinson and Roberto Clemente; you have to have them both.   Ernie Banks and Frank Thomas; you’ve got to include them both.  

              But if you think that way, it’s because you haven’t really tried to do this.   When you get down to what my father would have called nut-cuttin’ time, you realize that it isn’t obvious that Roberto Clemente is a Top 50 superstar.   It isn’t obvious that Miguel Cabrera has to be there.  

              I ran 27 polls yesterday.   I think that by the end of that exercise, most of you had a better sense of how hard it is.   It is a very, very, very high standard to be included in the top 50.  

              OK, we’re a long way from the end of the process.    I am trying to go from 54 players down to 10, and this is what we have done so far.

1)  Cal Ripken defeated Brooks Robinson, 66 to 34, with 2,574 votes so far.   OK, Brooks is out.

2)  Tony Gwynn defeated Rod Carew, 69 to 31, with 2.437 votes so far.   OK, I don’t really agree, but the standard is very, very high, and Carew does not have to be in the Top 50.

3)  George Sisler defeated Dazzy Vance 64 to 17 with 19% admitting they weren’t qualified to vote.   Again, I don’t really agree—I would choose Dazzy, I think—but we’ll drop Dazzy from the process and move on.

4)  Warren Spahn defeated Robin Roberts, 92 to 8.   OK.

5)  Jackie Robinson is leading Roberto Clemente 51 to 49, with 3,147 votes cast so far.   We can’t eliminate Clemente based on that kind of vote, so I’ll include both of them in the next round of comparisons.  

6)  Harmon Killebrew defeated Eddie Murray 66 to 34, with 2,085 votes cast so far.   OK, we’ll move on without Eddie.

7)  Al Kaline defeated Wade Boggs 54 to 46 with 2,149 votes cast so far.   That could be considered a tie, but I think I am going to accept the results of the vote and move on without Wade Boggs.   Sorry.

8)  Frank Robinson defeated Carl Yastrzemski 70 to 30 with 2,313 votes so far.   Somewhat shocked by the result; I thought I had enough Red Sox fans among my followers that the Red Sox would over-perform.

              I am going to leave Carl Yastrzemski in the pool for the next round.   Matching him against Frank Robinson is a touch draw, and I don’t want to prematurely eliminate Yaz.   It kind of came down to Yaz or Boggs in my mind, and I’m keeping Yastrzemski.

9)  Pete Rose defeated Robin Yount 73 to 27.   OK, Yount is out.

10)  Dave Winfield defeated Mark McGwire 56 to 44.   OK, McGwire is out.

11)  Hank Greenberg defeated Charlie Gehringer 84 to 16.   Again, I don’t know that I absolutely agree with that, but there are only 10 spots there for 54 candidates, and I can’t advance a player who loses 84 to 16 to a teammate. 

12)  Chipper Jones defeated Roberto Alomar, 59 to 41.   OK; I asked you what you think, that’s what you think, I’ll go with it.

13)  Albert Pujols defeated Billy Williams, 90 to 10.   Probably should have had Pujols in the first 40, but I accept that Billy Williams is out.

14)  Despite the many criticisms that the polls would have a recency bias, Ernie Banks defeated Frank Thomas 65 to 35 with 2,255 votes so far.   Banks and Thomas ARE, in fact, about even.   They won two MVP Awards each; they have about the same number of WAR.    I don’t really agree, but I accept the vote, and Thomas is out.

15)  Buster Posey is leading Mickey Cochrane 56 to 44 with 8 hours left in the voting.  

              Some people were very irritated by this, and I heard a lot about the recency bias and how twitter polls are stupid.   The fact is that in these matchups, the older player won 15 of the 27 votes.   OK, it is not usually a 1930s player against a 21st century player, and the fact that a 1960s player defeats a 1970s player does not mean that a player from now and a player from 85 years ago are competing on an equal footing.   I will advance Cochrane to the next round of the voting, as I did with Yastrzemski and Clemente, because the vote was close and there could be a recency bias.  

              But I want to say this:  that this is NOT a mismatch.   I think that some of the older voters just don’t appreciate how good Buster Posey is.   It is not a mismatch in terms of games played.  Cochrane played 1,482 games, Posey has played 1,031 so far, and it may be presumed that he will pass Cochrane.    Cochrane had an .897 OPS against an average for the position in his era of .718.    Posey has an .849 OPS against a position average of .714.   Posey has led his team to three World Championships, which is more impressive with 30 teams and several rounds of playoffs than it would be with 16 teams and one post-season series.  Posey is a very good defensive catcher.   It is NOT unreasonable to suggest that he is a greater player than Cochrane, and he deserves to be considered for one of those last ten spots.  


16)  Ichiro Suzuki defeated Miguel Cabrera, 57 to 43, with 2,627 votes so far.  I don’t agree; I would have voted for Cabrera, but I accept the vote and will move on.

17)  Joe Jackson defeated Chuck Klein, 86 to 14.   OK.

18)  Dizzy Dean is leading Carl Hubbell 58 to 42 with 9 hours left in the voting.   I think this vote is wrong, and that Hubbell—who won two Most Valuable Player Awards, one of very few pitchers to do so—was clearly closer to the superstar criteria in my mind than Dean.   Nonetheless, I accept the vote, and will drop Hubbell as a candidate for one of the last ten spots.

19)  Andre Dawson defeated Dale Murphy, 71 to 29.   OK, Murphy is out.

20)  Rickey Henderson, who should probably have been included in the original 40, defeated Ryne Sandberg 91 to 9, 2,374 votes so far.  Sandberg is out.

21)  Whitey Ford defeated Gaylord Perry 68 to 32.   Gaylord is eliminated as a candidate.

22)  Steve Carlton defeated Jim Palmer 73 to 27.   I accept the results, and will eliminate Jim Palmer as a candidate. 

23) Pedro Martinez is leading Mariano Rivera 76 to 24 with 2,411 votes cast.   Despite this lopsided result, I think I am going to have to include Mariano in the next round; I’ll explain in a moment.

24)  Don Drysdale is leading Ferguson Jenkins 54 to 46 with 10 hours left in the vote.  I disagree with this.   I think Jenkins was a greater player than Drysdale and better fits my definition of a superstar; however, I don’t know that either Jenkins or Drysdale is an especially strong candidate for one of the last ten spots, so I will accept the result and move on without Jenkins.  

25)  Rollie Fingers is leading Goose Gossage 53 to 47 with 10 hours left to vote.   A 53-47 margin in a twitter poll is not terribly informative, and probably should be considered a tie.   However, not regarding either man as an especially strong candidate, I will accept the result of the vote, and move on without Gossage, unless he rallies in the voting.

              But if Fingers is in and Mariano is out, is that right?   Don’t most of us now regard Mariano as the greatest closer of all time?   I think that most people do, and reasonably,   Pedro Martinez is a tough first-round match for Mariano, and I don’t want to exclude Rivera because he had a tough draw. I have to be careful about my followers favoring Red Sox players (Martinez) over Yankees.   I’m going to substitute Mariano for Goose, and go forward with both Fingers and Mariano. 

26)  Jim Thome is leading Johnny Mize 59 to 41 with 9 hours left in the vote.   I do some concerns about a recency bias here, but I don’t know that Mize is an especially good candidate at this level anyway, so I will accept the result and exclude Mize.

27)  Nolan Ryan is leading Christy Mathewson 55 to 45 with 2,272 votes cast.   This is (a) a close vote, (b) a tough matchup, and (c) a comparison of players from different time zones.   I’m not comfortable excluding Christy Mathewson, so I will include both Ryan and Mathewson in the next round.


              That gives us 32 players to be included in the next round of the process—27 who won their vote, plus five players who were granted an extension because of extenuating circumstances.   I’ll give people a little time to process this article, and then I’ll start the next round of voting. 


              Speaking tonight at the Plaza branch of the Kansas City Public Library, if you’re in KC. 




COMMENTS (38 Comments, most recent shown first)

(sorry for the typo/confusion -- that sentence is supposed to be:
You think it's not possible that many if not most people would think the players that you're talking about are "more deserving" on this than Carter or Santo?
6:03 PM Sep 23rd
Guy: Funny that in the midst of arguing so strongly against prominent use of subjectivity, you're so subjective in judging what the article is doing and where Bill was coming from.

Like, in saying: "Bill did write that, and we can all see immediately how Mantle/Williams/Musial differ from the latter 4. But we now know that Bill didn't really mean it, because his full list of "unarguable superstars" also includes players who are not obviously more deserving that Carter or Santo (Reggie) or are pretty clearly less deserving (Jeter, Ortiz)."

Let's forget the "obviously," because it's not key for this (yes, I'm using subjectivity) :-) and look at the rest: You think it's not possible that many if not most people would think the players included players that you're talking about are "more deserving" on this than Carter or Santo?
Suppose most people do think so?

You're talking about your subjective judgment of Bill's statements and decisions. Suppose your subjective judgment is outlying?
6:02 PM Sep 23rd
"I mean TRUE superstars; I mean Mickey Mantle, Ted Williams, Stan Musial type of superstars, not Albert Belle, Gary Carter, Al Simmons and Ron Santo type superstars. I mean the real deal." He could have picked any player as one of his four clearest examples. He chose Carter. That has to count for something, no?
Bill did write that, and we can all see immediately how Mantle/Williams/Musial differ from the latter 4. But we now know that Bill didn't really mean it, because his full list of "unarguable superstars" also includes players who are not obviously more deserving that Carter or Santo (Reggie) or are pretty clearly less deserving (Jeter, Ortiz).

As for the idea that the statement itself has to count for something, I agree: it is further evidence that Bill really needs to find an objective way to make the superstar determination, if he wants the study to have value. (Again, my concern is the study to come. If people just want to have a fun, subjective debate about "who's a superstar?", then knock yourselves out. I have no problem with that at all.)

And one correction on my original post: I think Pedro clearly should be "In" as one of the core 40 superstars.

11:58 AM Sep 23rd
(Golly we need an edit function; that looks awful.
I don't mean any "human aspects" about Carter, but, using one's own 'human aspects' of judgment.)
10:39 AM Sep 23rd
That's exactly the kind of thing he's criticizing!
(I agree with you, and (of course) I agree with Bill's using human aspects along with numbers and metrics.)
10:37 AM Sep 23rd
Steven Goldleaf
Guy123--You do realize that Carter was one of the examples Bill gave (in this article!) of someone who doesn't qualify as a true superstar: "I mean TRUE superstars; I mean Mickey Mantle, Ted Williams, Stan Musial type of superstars, not Albert Belle, Gary Carter, Al Simmons and Ron Santo type superstars. I mean the real deal." He could have picked any player as one of his four clearest examples. He chose Carter. That has to count for something, no?
5:57 AM Sep 23rd
Guy: If you'd like, take an informal poll and see how many people think it doesn't fail to meet the sniff test to say that Gary Carter is "clearly" in a higher category than that group of 3 on a thing like what we're talking about, and if such a result isn't a poor comment on the method.

Regarding what you said about intuitive things: Of course they don't determine an issue, necessarily. But thereare instances where a method's result seems so clearly at odds with good intuition (not to mention with other metric findings) that the result is a poor comment on the method. I do say there's little doubt that this is one such instance.

You seem to be muddling some aspects of what I'm saying. I didn't say that a decent method can't produce non-intuitive results, and I didn't say the method stinks; I don't know if it does or not. Again, all I was saying was that this result is a poor comment on the method, and that I think when you've got something like that right near the top of some exegesis, it may tend to discourage interest in reading further.

To be clear: I didn't say either that I think those other 3 are greater or better than Carter. In fact I do think 2 of the 3 are better/greater; Ivan Rodriguez, I don't know quite how to evaluate him, for various reasons. All I'm saying is that it's a poor result when something shows Carter to be higher than those others (when if anything he's lower), and more so when it shows him "clearly" higher, and it surprises me if your own sniff test doesn't tell you the same.
7:35 PM Sep 22nd
I said putting Carter in a "clearly" higher category/group than those other 3, as you did, doesn't meet the sniff test, which it doesn't. It's a poor comment on the approach.
Have you every done this kind of analysis? One non-intuitive result is not a "poor comment" on the approach. Any objective method, consistently applied, will yield some results that don't "feel" right. Often this just means one's intuition was wrong, but sometimes the method has indeed missed something. In this case, it comes down to whether or not Carter was a substantially better fielder, and I don't have a strong opinion on that. But rejecting an approach because you disagree with one output (and w/o evidence) is senseless.

Any objective method will yield a few results that feel wrong. Indeed, it will likely yield a few results that *are* wrong. But it's still what you have to do. If instead you define a player as a superstar because "I know a superstar when I see one," then you have no idea what you are measuring, and even if you find a correlation you have no idea which way the causal arrow points. And if the researcher making these subjective judgments already has theories he is invested in -- Bill admits in this post that he already knows that Gibson and Yaz turned around their franchises! -- then the study has been blown up before it's even finished. GIGO.
4:24 PM Sep 22nd
All I was saying was, what I was saying. :-)
Here's the thing from your post, resulting from that approach based on WAA, that I was commenting on:

"The top 2 catchers are clearly Bench and Carter (both 8 seasons). They are followed by a 3-way tie at 6 seasons among Pudge, Piazza, and Berra...."

I said putting Carter in a "clearly" higher category/group than those other 3, as you did, doesn't meet the sniff test, which it doesn't. It's a poor comment on the approach. Carter was a very good player, maybe a great player. I didn't mean he wasn't. All I was saying was, well, what I was saying.-
2:28 PM Sep 22nd
MarisFan61: what, exactly, are you sniffing?

As I said, there are a lot of ways of looking at this, and they will give you somewhat different answers. If you want to draw the line below Bench, and say it's a 4-way scrum for 2nd place (Yogi, Pudge, Piazza, and Carter), I'm not going to say you are necessarily wrong. Among other problems, measuring catcher defense is very challenging. My point is simply that it's not obvious that Yogi should be in and Carter should be out. More importantly, it only *seems* obvious to you because Yogi played in 14 WS while Carter played in 1. And that's why Bill needs to find a better way of doing this......
1:41 PM Sep 22nd
Let me distill the questions to 1 main one:
(OK, it's not necessarily everyone's main question but I'm saying it's the main question) :-)

Bill, it might seem like your usual thought in such a situation, where 40 players rise to the clear top and then you've got an auditorium-full of close candidates for the next 10 spots, would be to forget about the original intent to get 50 and just say OK it's 40, because those 40 are a real group and the others aren't really in that group and you don't really need any particular number like 50.

How come you didn't just say OK to 40? Was it mainly just because you knew it would be fun to get the last 10 in this way, which admittedly it is?
1:12 PM Sep 22nd
Guy: I'd have to say that your method pretty immediately fails the sniff test and might keep some people from being interested to look further; not your fault -- it's the fault of the metric -- except to the extent that you chose to use "WAR"-related metrics.

At least to me, when such a thing puts Carter in a top category (together with Bench) and leaves Pudge, Piazza, and Berra behind, that's failing the sniff test, and it undoes any confidence in the method.
12:56 PM Sep 22nd
Since the point here is to measure the relationship between having a superstar player and team success, the first thing we need to do is find a way to define “superstar” that isn’t colored by the success of the players’ teams. Unfortunately, by relying primarily on BJ’s intuition, and secondarily on fans’ voting preferences (without guidance on voting criteria), it seems the current approach will define superstar status at least in part based on team success. So let’s compare Bill’s 40 core choices to a more objective measure of whether a player has a large impact on his team’s won-lost record, to see how that might change the list.

We can define superstar performance as seasons with at least 4 wins above average or WAA, using B-Ref figures, which represents roughly the best 10 position players and top 6 pitchers. For catchers, let's llower the threshold to 3 WAA, since the WAR system arguably undervalues catchers a bit. Players who meet these standards frequently are “superstars.” There are many other valid ways to approach this, and WAA alone certainly can’t answer the question. But we need to start with an objective standard, and this is as good as any.

CATCHERS: The top 2 catchers are clearly Bench and Carter (both 8 seasons). They are followed by a 3-way tie at 6 seasons among Pudge, Piazza, and Berra, and looking a little deeper Berra clearly was the weakest of these three – at a minimum, he should have to fight his way into the final 50. Campy trails badly at 3 seasons, behind Dickey, Mauer, Cochrane, and several others.
Out: Campy. In the cage: Berra.
In: Carter.

There are 21 players with 9 or more superstar seasons. Arguably all of them belong in the top 40; at a minimum, you need a strong argument to exclude them. BJ makes a reasonable case for excluding Collins, but 4 others I think belong: Rickey! (10), Clemente (9), Mathews (9), and Pujols (9). Four more players had 7-8 superstar seasons: Foxx, Boggs, Frank Robinson, and Mize. Only Foxx made Bill’s cut, but the other 3 have a strong case to make.

Five of Bill’s choices had 5-6 superstar seasons: Brett (6), DiMaggio (6), Griffey (6), Trout (6), and Morgan (5). Digging a bit deeper, Morgan is a step above the rest (5 amazing seasons, and 5 more just below the 4 WAA threshold), and we assume Trout will add to his total, but the other 3 should have to fight their way in. Finally, 3 of BJ’s non-catcher choices don’t really belong in the discussion: Reggie (4), Jeter (3—at best), and Ortiz (1).

Out: Reggie, Jeter, Ortiz. In the cage: Brett, DiMaggio, Griffey.
In: Rickey!, Clemente, Mathews, Pujols.

There are 11 pitchers with 7+ superstar seasons. Interestingly, BJ omits 3 on his list: Mathewson (10), Pedro (8), and Schilling (7). Schilling is clearly a notch below the others (though arguably stronger than several of BJ’s selections). Another 4 pitchers have 6 superstar seasons – Blyleven, Halladay, Niekro, and Kershaw – but only Kershaw (who presumably has more superstar seasons to come) makes the list. Three of BJ’s selections had only 4 superstar seasons: Gibson, Feller, and Koufax. Gibson had by far the best career of the 3, but he should probably have to fight his way past guys like Halladay and Niekro to reach the inner circle. The other 2 don't make the cut here.
Out: Feller, Koufax. In the cage: Gibson.
In: none.

All told, 11 of the players on Bill's list (not 4) are at least debatable selections -- and that's before we get to the really hard choices! It’s not immediately clear how adopting a more objective “superstar” definition would impact the study Bill is doing. Some of the player’s mistakenly included clearly are associated with periods of team success (Jeter, Ortiz, Jackson, Campy, Koufax). On the other hand, some of those omitted have also been on winning teams (Pujols). My guess is that using an objective measure would undercut Bill’s hypothesis at least a bit, but that’s just a guess.

Of course, there are also the final 10 players to consider, who are now being selected via Twitter, and where the potential for bias is probably even larger. It seems likely voters are at least somewhat influenced by the glow of postseason appearances. And Bill himself may be unconsciously putting his thumb on the scale, such as keeping Rivera and Yaz alive despite their well-deserved electoral defeats. The bottom line here is that this study really needs some kind of non-subjective basis for determining who is and is not a superstar.

12:22 PM Sep 22nd
You're right, I screwed that up. I must have been hypnotized by those 6's. Thanks for the catch. Chipper's MVP record is still more impressive, but the gap is narrower.
11:18 AM Sep 22nd
@wovenstarp "Alomar never once cracked the top 5 in any MVP vote"

This is flat out wrong!

MVP (rank, share)

1991 AL (6, 33%)
1992 AL (6, 30%)
1993 AL (6, 26%)
1996 AL (20, 1%)
1997 AL (22, 1%)
1999 AL (3, 57%)
2001 AL (4, 42%)

9:21 AM Sep 22nd
One aspect of the study I don't understand is whether single-team superstars have any utility in the study. Ripken, Brett, Schmidt.... one thing we can agree on is that you can't draw "just arrived/just left" conclusions from their careers in any way, except arguably twice, and certainly never in the players' peak periods. In that sense I would "vote for" Alomar over Chipper, Rickey over Brett, etc. But I'm clearly missing something about the shape of the study.
8:26 AM Sep 22nd
Well, the higher highs that Alomar reached there are pretty meaningful. I don't think there's much doubt that when Bill writes, in effect, "you guys really took Chipper over Alomar, huh?" that we can take that as an expression of skepticism. I don't know what Bill thinks, after all I was asking for clarification.​
8:22 AM Sep 22nd
I love this sort of thing, and your processes are always fun and interesting. It seems to me, though, that your original 36 plus the small number of additions - to 40, 42, whatever you decide - is the list of droids you are looking for.

Beyond that you are inviting the rank and file to an American Idol competition, where 10 get to join the real Idols. I'm fine with that because I love the drama and excitement - the arguments - but I'm mildly surprised that you are, because you've brought the drama factor to what began as more of an objective study.
7:44 AM Sep 22nd
Managers and turning points: intuitively there does seem to be a correlation, but which is the chicken, which the egg?

Consider the Cubs: just as their young talent is becoming major-league ready, Joe Maddon falls into their lap. Bingo: instant improvement, followed pretty quickly by a championship. They certainly don't win the World Series without Bryant, Rizzo, Russell and Co all growing up at the same time; do they win without Maddon? How can we know? Does Maddon even take the Cub job if their future isn't so obviously promising? There is never a guarantee of victory, but you've got to like your chances a lot better with Theo's Cubs than with the Friedman-less Rays.
7:41 AM Sep 22nd
Also, as to the notion that "it's something about the process that farmers go through when they're working with nuts": yeah, sorta!
1:12 AM Sep 22nd
Theories on the etymology of "nut-cutting time" seem to vary, but if your family is involved with cattle, it will not occur to you to question the phrase in the first place.
1:09 AM Sep 22nd
Without any dog in the fight, I went and got the Win Share data for those guys, using the breakdown categories that Bill used in the New Historical Abstract.
Data are from this site.
(totals done in head; don't take them to the bank, feel free to double check)

Career Win Shares
R. Alomar 375
C. Jones 416

Top 3 years (not necessarily consecutive)
R. Alomar 37, 35, 34
C. Jones 32, 31, 29

Best 5-year run
R. Alomar 132 (btw, never really put together 5 straight real good yrs)
C. Jones 148

By Win Shares, it looks like Alomar had better over-the-top years, but they were scattered, he wasn't consistently year-after-year excellent in the way that Chipper was (Chipper had 8 straight years of at least 23 Win Shares, 9 straight of at least 20; Roberto never had more than 3 in a row of at least 20), and his career total was lower.

If it were me doing this kind of assessment, and doing it by the Win Shares, I might pick Chipper, wouldn't even consider Alomar (never having more than a 3-yr run of at least 20 Win Shares? that doesn't cut it for something like this).

But, did Bill really say he considers Alomar greater??
I don't see it. All I necessarily see is that he wouldn't have expected the vote that occurred. Maybe he meant the margin was greater than he expected; maybe he didn't think either was greater than the other....
12:11 AM Sep 22nd
I'd love to hear your reasoning as to why Alomar is a greater player than Chipper, Bill. Alomar never once cracked the top 5 in any MVP vote, something Chipper did twice, winning outright once. Chipper's WAR numbers on are much more impressive. I don't have Win Shares info or I'd have used that. Chipper topped 1.000 OPS five times, Alomar never did it. Sure Alomar had much greater speed and played a key defensive player much better, but in saying that we tend to forget that Chipper stole 20+ bases twice in his career -- the guy wasn't Pedro Sandoval out there. I don't understand the argument that states that Chipper stands in Alomar's shadow -- although I haven't heard it yet.
11:20 PM Sep 21st
Hi Bill, I enjoy all your articles.

I like to think about teams that were good and went bad or teams that were bad and got good and teams that stayed good for a long time and teams that were bad for a long time. A team I remember is the Milwaukee Brewers from 1978 to 1982. They were good every year. Then in 1983 they were still winning but players like Cooper and Gorman Thomas aged. They stunk in 1984, Ben Oglivie had an off year, Mike Caldwell was hurt and pitched poorly.

The Phillies stunk in 1961 but had a winning record in 1962 because of expansion and were a winning team from 1962 to 1967. Then got bad and started to be good around 1974 through 1984. Managers do make a difference, George Stallings with the 1914 Braves who were also contenders in 1915 and 1916. Gil Hodges and the 1969 Mets. Dick Williams turning the Red Sox into winners in 1967. Terry Francona has been a winning manager with the Red Sox and Indians. Dave Johnson turned the Mets into winners in 1984 and later had success with the Reds, Orioles, Dodgers and Nationals. Dusty Baker has managed a lot of winners.

The Yankees had Babe Ruth and Lou Gehrig together from 1925 to 1934 and won pennants in 1926, 1927, 1928 and 1932. That is it. You would think with two super power hitters in the lineup, they should win the World Series all the time. It is hard to win 95 games every year.

Connie Mack had many ups and downs. He won the pennant in 1914, then gutted the team. They stunk in 1915 and for years after then in 1922 were in 7th place, 1923 6th place, 1924 5th place, 1925 2nd place, 1926 3rd place, 1927 and 1928 2nd place and 1929, 1930 and 1931 1st place. Then gradually down the hill again.

Ichiro joined the Mariners in 2001 and they won 116 games. The team had other winning years with him on the team.

I enjoy Strat-O-Matic Baseball. I play the computer game. I did an experiment with the 1979 Blue Jays. They were bad in their first few years. I added one player at a time, 1911 Ty Cobb, 1920 Babe Ruth, 1966 Sandy Koufax and 1904 Rube Waddell. They were .500 with Cobb and Ruth and adding Koufax and Waddell they became a great winning team. Those four players most likely added 120 Win Shares to the 53-109 Blue Jays.

Take Care,
Tom Nahigian
7:04 PM Sep 21st
Buster Posey seems like a no-brainer to me, especially since only two other players from the 2010s who haven't had nearly the same impact in terms of championships have been selected so far.
5:19 PM Sep 21st
(also please pardon my phumphing redundacy)
5:00 PM Sep 21st
What's the etymology of "nut cutting time"?

I've envisioned it sort of like it's like it's as though the guys who get eliminated are (pardon the expression) getting their nuts cut off.

I can also imagine it's something about the process that farmers go through when they're working with nuts, maybe some final step.

Any assistance on this would be appreciated. :-)​
5:00 PM Sep 21st
MarisFan -- The only other time I've heard the phrase "nut-cutting time" is in Robert Caro's great Lyndon Johnson books; LBJ used the phrase, I think maybe to describe the hard part of the process of giving out committee assignments, or persuading senators on a key vote, or the backroom DNC VP dealings in 1960... something like that...

Re: quality of play issues between early 20th and early 21st century baseball -- I had the same thought, but I wonder whether it matters for the purposes of Bill's exercise. The purpose is to identify the biggest superstars relative to the league, whether the quality of that league is high or low. That way, one can track how having this great player, relative to league, correlates with having a great team, relative to league. There may be some second-order effects of lower quality of play (like higher standard deviation in player skill) that would need to be controlled for, but the quality of play itself wouldn't be an issue. I think.
4:35 PM Sep 21st
Besides the Franks: Thomas and Robinson, the other glaring ommision to my mind is Steve Carlton.

But, sorry, I've put the energy in this as you, but I don't see how those guys are out. Christy Mathewson was clearly a superstar as well, but Pete Alexander was his equal. Not necesserarily top 50 in outstanding-ness, though. Just a superstar in people's eyes - or so I thought.
4:34 PM Sep 21st
I agree with wovenstrap - certainly I would have voted differently in several of these if I had been asked "who was the better baseball player" without reference to being a "superstar."
4:14 PM Sep 21st
I stopped reading when Frank Thomas was eliminated. If you give Yaz a break, Big Hurt deserves one too against a shortstop (half his career) who hit 500+ homers. If David Ortiz makes the cut, how can you leave off Frank Thomas? Makes no sense to me, but this is Bill's article and he can have all the Red Sox he wants.

With 30 teams instead of 16, you should have 5 or 6 guys from the 1990s and beyond, no? I understand as coaching and training improves at the same time as the population of available talent grows, it is harder to be outstanding. I guess this is about how outstanding you are not how skilled you are, but nonetheless, for the sake of justice, I'd prefer to see more recent guys.

4:13 PM Sep 21st
3 stray comments:

-- Jim Bouton used the phrase "nut cutting time" in Ball Four, about a roster-cut-down date. I thought it was a great phrase and I've used it a few times since; I would have been using it lots more except that I never came across it or heard it anywhere else and so I wasn't sure anyone would know it, or even if it was any actual phrase or if Bouton just tossed it out.
Thanks for using it, and thanks to your dad. Now I know I can use it. :-)

-- If 40 names came out pretty easily and you could see it would be so hard to find the 10 others, why didn't you just go with the 40, and be content with 40? Was it largely because you knew it would be fun to get those other 10 in the way that you're doing? (Which it is.) After all, you lead off the article by talking about how studies typically evolve as they go along....

-- About this other thing near the top:
"I still believe the underlying proposition is true (that the hiring and firing of managers are often the turning points for organizations), but I don’t think I’m going to be able to prove that it is true. That’s alright; I can’t prove that the brakes in my car will work, but I still use them when I need to slow down."
Well. :-)
I don't have any complaint about that. I love it. It just makes me wonder why you're not inclined to apply that kind of thinking to all kinds of other things (particularly, some traditional beliefs, like that batting order matters a lot) that studies haven't (yet) been capable of demonstrating to be true; there are a bunch of things I and others have noted that 'seem' to be true and which you and most if not all of the field have dismissed because no studies have demonstrated them. Maybe it's because this thing you're talking about here doesn't yet have as much quantity of studies that have failed to demonstrate it. Anyway....I'm glad to see you applying that kind of reasoning. I'm just more inclined to apply it more freely, which I know can sometimes seem like stubborn rejection of research.
4:08 PM Sep 21st
Steven Goldleaf
My one overarching quibble, however, is that your (correct) dismissal of 19th century baseball from the discussion leads me to a corollary conclusion, that early 20th century is also inferior to early 21st century play--I'd have no problem with a system that increased the number of great players by one (or even two) per decade, because to my mind that's the least by which great players have occurred. So one great star from the 1900s, two great ones from the 1910s, three from the 1920s, etc. (or some such system that increases the raw number of great players over time) would seem justifiable to me.
4:05 PM Sep 21st
Steven Goldleaf
Where are you going to be if I'm not in KC tonight?

I think this is a great technique generally, arriving at truths by direct comparisons between players--I wish we did this sort of thing more, taking two comparables and comparing their body of work against each other's. It reveals more than a discussion or a ranking of twelve or twenty roughly comparable players, though it does require much more work.
3:59 PM Sep 21st
Your involving readers back from 30 years ago in ranking players was terrific and showed that it can work. I love it that you are taking to twitter for this.
3:57 PM Sep 21st
I am unsure why Eddie Collins is excluded when Joe Jackson is not. At least Eddie played in the 20s and could be included in that decade. Jackson really only fits as a 1910s player and as you note, that decade already has 4 better players.
3:26 PM Sep 21st
I thought it was a fun exercise. Many found it to be agony. I think the word "superstar" is slightly distorting -- in the Drysdale vote you can see the effect of that because Drysdale's main character trait is being famous. If the question had been for "impact players" or something, I think you might have seen different results for Yount, Eddie Murray and so forth. Your deference to the voices of a benighted Twitter public is refreshing.
3:16 PM Sep 21st
I understand this is a tough assignment, but you only have three players from the 1900s, and Mathewson isn't one of them. That's the one player missing from the original 40 that sticks out the most. One suggestion: If you're trying to fill out certain decades, why not have players from the same decade matched against each other in the poll? You did this with Carlton and Palmer, Gehringer and Greenburg, which makes sense. Why not pit Frank Robinson against Yastrzemski, and Ernie Banks against Al Kaline?

3:02 PM Sep 21st
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy