Updates on the 50 Superstars Project

September 23, 2017
 2017-46

Update on the 50 Superstars Project

              I appreciate everybody’s interest/enthusiasm for the project.   Thank you all for your comments and participation.   Responding to a few comments posted on my earlier article:

1)  Responding to Evanecurb (first post).. .(a) I DID match Carl Yastrzemski against Frank Robinson.   (b) It’s not a hard and fast rule that there have to be a certain number of players from a certain decade.  You can’t make rules for the Almighty.   If there are 5 superstars in one decade and 1 in the next, you don’t make phony selections to create a balance.   (c) Many times you can’t REALLY say what decade a player belongs to.  (d)  The "decade" issue should not be given priority over similar problems.    I match players from the same time period when I can, but I also match catcher against catcher, pitcher against pitcher, Chicago player against Chicago player, San Diego outfielder against San Diego outfielder, and short-career player against short-career player.   It wouldn’t be right to ignore all of those and just make matchups on the timeline.

2)  Responding to Steven Goldleaf—It may well be true that the players of 1990 are a stronger group than those of 1900, but the purpose of creating the 50-superstar list was so that I could use to study another issue, which is the effect of having a superstar in his prime on the direction of the franchise.  To have a strong time-line bias in the group would interfere with that study.  RipCity’s comment on this was essentially right.

3)  Responding to Marisfan—40 names were easy because I was picking 50.   If I was only picking 40, then picking 40 would be extremely difficult, because there would be no tolerance for errors.  I left Albert Pujols, Christy Mathewson and Albert Pujols out of the original 40—an oversight, obviously, but it doesn’t matter because we can cover it in the final 10.   If I didn’t have the final 10, I couldn’t pick 40.

4)  Again Marisfan—It isn’t that studies have been unable to show that there is a meaningful effect for batting order.   It is, rather, that study after study has definitively shown that there is NOT a meaningful effect.   There is a big difference between "I don’t know how to measure that" and a good study that shows no effect.

5)   I don’t know and don’t really care at the moment whether it’s Alomar or Chipper.   Alomar had huge impact on pennant races.   I made up a method for that and studied it one time.   Alomar, like Koufax and Carl Hubbell, was a pivotal player in pennant battles.   Asking people what they think implies that you have to pay SOME attention to what they think. 

 

OK, let’s review the second-round results now and set up the third round.

 

1)  Christy Mathewson defeated George Sisler 91 to 9.  Sisler is out.

2)  Hank Greenberg defeated Joe Jackson 52 to 48, which was the closest vote of the second round.   We will move both players into the next round.

3)  Jackie Robinson defeated Ernie Banks 70 to 30.   Banks is out.

4)  Warren Spahn defeated Don Drysdale 83 to 17.   Drysdale is eliminated.

5)  Roberto Clemente defeated Al Kaline 87 to 13.   Again, I don’t necessarily agree with that.  I kind of think that Kaline may have been the greater player, but honestly, Clemente is more what I need for this study, anyway.    Clemente started slowly, had a clear peak period and then departed suddenly.   Kaline had great years at ages 20 and 21, may have been "actually" better years later, but not obviously better in any case.  Clemente is better for the study.  It’s sort of what I mean by a true superstar, as opposed to simply the quality of the player.   A superstar has a series of high-impact seasons. 

6)  Frank Robinson defeated Harmon Killebrew, 84 to 16.   Killebrew is eliminated.

7)  Dizzy Dean defeated Mickey Cochrane, 58 to 42.   Don’t necessarily agree, but it is what it is.   I already saved Cochrane from elimination once; I can’t save him twice.

8)  Pete Rose defeated Carl Yastrzemski, 56 to 44.   Well. . .OK.   I saved Yaz once; I can’t pull him out of the fire again. 

9)  Steve Carlton defeated Whitey Ford, 79 to 21.  Ford is out.

10)  Tony Gwynn defeated Dave Winfield by the surprising margin of 76 to 24.   OK.   Winfield is gone.

11)  Nolan Ryan defeated Rollie Fingers, 92 to 8.   Fingers is eliminated.

12)  Rickey Henderson defeated Andre Dawson, 92 to 8.   Dawson is gone.

13)  Cal Ripken defeated Jim Thome, 84 to 16.   Thome is eliminated.

14)  Ichiro defeated Mariano, 60 to 40.   I think the competitive performance by Mariano justifies my decision to keep him in the study after he was demolished in the voting in the first round (by Pedro Martinez), but I can’t save Mariano again.   He gone.

15)  Pedro Martinez defeated Buster Posey, 92 to 8.    No Posey.

16)  Albert Pujols is leading Chipper Jones, 83 to 17.   I think I’ll grant Chipper a reprieve, though.  Pujols should have been in the original 40, of course; losing to him is kind of like losing to Babe Ruth.   I’m going to keep Chipper around.

 

              OK, here’s what I’m going to do.   First of all, I’m going to move three players (Christy Mathewson, Rickey Henderson and Albert Pujols) into the "automatic" group, increasing the number of automatic players from 40 to 43.    That will leave us with 15 players in competition—13 winners from the previous round, plus Joe Jackson and Chipper Jones, who I am going to give another shot to because of special problems.  

              As it happens, just by chance I guess, the last fifteen break down as five pitchers, five infielders and five outfielders, if we count Pete Rose as an infielder, and he did play about 60% of his career games in the infield.   So let’s do this.   Let’s do five-candidate rounds for the next three spots, and that will give up 46 players chosen and 12 candidates for that last four spots.   Then we’ll sort that out and get the last four.   OK? 

              Yoseph. . . I’ll look forward to your 50 when you get it done.

 

 

 
 

COMMENTS (16 Comments, most recent shown first)

bhalbleib
Steven: I am further confused by Bill's point about Eddie Collins being the 5th superstar (after Cobb, Speaker, Johnson and Alexander) so he's out, but he included Joe Jackson (who now has been voted through 2 rounds), who clearly is the 6th superstar of that era.
4:00 PM Sep 26th
 
MidnighttheCat
jaybracken,

c'mon, Mariano was just lullin' him into a false sense of security !

:-)
6:06 AM Sep 26th
 
jaybracken
MidnighttheCat already conceded the Ichiro vs Mo point, but selecting the 2001 ALCS as the head-to-head measure between the two leaves out the following:
Ichiro's regular season numbers against Mo: .385/.429/.615

Talking insanely small sample sizes in both cases, but you certainly can't act like this was some kind of one-sided domination, I don't think.
9:47 AM Sep 25th
 
Guy123
As BryanBM observes, not all seasons above 4 WAA (6 WAR) are created equal. So I think a good objective starting point for this effort is "WAR7," the wins a player generated in his 7 best seasons. This rewards players who have the peak impact Bill is looking for, while still requiring some longevity. And it gives appropriate weight to really extraordinary seasons, which isn't true of simply tallying how many superstar seasons a player delivered (as I suggested before). And before anyone objects, no, there is nothing magical about 7 years, it's just what B-Ref conveniently reports. If you want to include Koufax you can use WAR5, and if you want to include Campy you can use WAR3 (and if you're partial to Mark Fidrych, use WAR1).

Comparing Bill's core 43 superstars to the top 43 players in WAR7 (including the top 3 catchers), we find 11 of Bill's automatic superstars don't make the cut, and would be replaced by 11 players with higher WAR7:
Out: Ortiz, Jeter, Campy, Reggie, Yogi, DiMaggio, Brett, Ott, Feller, Koufax, Kershaw.
In: Ripken, Boggs, Matthews, Clemente, Carter, Piazza, Pedro, Ferrell, R Roberts, Niekro, Carlton.

Just eyeballing the list, you can see how relying on intuition rather than an objective measure badly biases Bill's forthcoming study. The guys granted automatic admission generally played on much more successful teams. Look at the radical difference in how often these players reached the postseason:
BJ 11: 86 out of 186 seasons, 46%
WAR7 11: 40 out of 209 seasons, 19%.

The point is not that the 11 WAR7 selections are more "superstary" than Bill's choices -- maybe they are, maybe they aren't. Everyone can have their own opinion on that. But that's precisely the problem -- it's a matter of opinion. That subjectivity contaminates Bill's study, because the selection of superstars -- our independent variable -- is based in part upon the dependent variable (team success). And of course, the final 7 fan selections may also be biased by the success of players' team.
9:13 AM Sep 25th
 
MidnighttheCat
No quarrel on Pedro v. Mariano, even as a Yankees fan, since it is a starter v. closer thing.

But watch the at-bats Ichiro had against Mariano in the 2001 ALCS. He was totally overpowered. No chance at all. Just sayin'. Great player, and everyday position player v. closer, again, I get it. But one-on-one? No contest.
7:00 AM Sep 25th
 
Brock Hanke
hotstatrat - Thanks for letting me know that Kaline still has a good fame base in Michigan. I'm old enough that I was trying to analyze statistics vs. reputations while Kaline was still playing (I was a kid, so I didn't make any great discoveries or anything), and one thing I do remember was that I always thought that Kaline was underrated by the press, and much less famous than he should be. I don't know whether it was because he never quite lived up to his very first years or what, but, at least in St. Louis, he was not nearly as famous as he should have been. I've kept track of this, and I still think that Kaline has that problem. However, if there is one place where he would not have the problem, it's certainly Michigan.
4:25 AM Sep 25th
 
BryanBM
1961-1966 Koufax - at most 5 points
1981-1985 Stieb - most likely 5 points with the other point in 1990

Scoring system might need some adjustments, Dave's 6 oranges aren't the same as Koufax's 4 or 5 apples.
12:16 AM Sep 25th
 
Steven Goldleaf
Sorry--I just re-read the first article Bill posted, where he posted explicitly about leaving Eddie Collins off the list. Whatever it was, it wasn't a dropped card. Still don't get how Lajoie and Hornsby leapfrogged over him to the extent that he doesn't even get to be in a run-off against another candidate, but it was definitely a deliberate choice to leave him off the short list of super-stars.
8:50 PM Sep 24th
 
KaiserD2
I am not going to be doing blow-by-blow critiques of this project of Bill's but I thought some basic data from my forthcoming book (now delayed until early 2018, but with 2017 data) would make for an interesting perspective.

As I explained, using my own statistical methods--adjustments of baseball-reference.com data based on DRA fielding statistics and some changes for pitchers--I have identified every season of 4 WAA or more in baseball history. That definition of a superstar season -- 4 WAA --is the answer to Bill's question in the Keltner test, "If this guy were the best player on your team, coudl you win the pennant?" So, to find the 50 best players history, I would simply find the 50 players who had reached 4 WAA the most times. I wouldn't pay any attention to MVP scores, contemporary opinion etc.

Ruth and Bonds have 17 superstar seasons a piece (including two as a pitcher for the Babe.) Speaker, Aaron and Mays had 16 each, Cobb had 14, Musial and Ted Williams 13, Clemens, Gehrig, Ott, and Clemente 12. Henderson, Hornsby, Frank Robinson, Schmidt, and Pujols had 11. Wagner, Lajoie and Bagwell had 10. Walter Johnson and Christy Matthewson had 9 each. Reggie Jackson, Alexander, Eddie Collins (!), Johnny Mize, Mantle, Eddie Matthews, Yaz, Frank Thomas, ARod, Greg Maddux and Bob Gibson had 8 each. So far we have a total of 35.

At 7 each we have Boggs, Keith Hernandez, Bill Terry, Greenberg, DiMaggio, Billy Williams, Eddie Murray, Edgar Martionez, Jim Wynn, and our own Miguel Cabrera. (Keith Hernandez and Jim Wynn are my picks for the players who have been treated most unfairly in Hall of Fame voting--you will have noticed that nearly everyone so far is in.) That was 10 more, so we are at 45.

At this point, things get sticky. There are 18 people with 6 such seasons, and we have room for only 5. They are Blylevel, Pedro Martinez, Ed Walsh, Joe Jackson, Sisler, Paul Waner, Richie Ashburn (very underrated) Kaline, McCovey, Dick Allen (1 less than his contemporary Jim Wynn, take note), Molitor, Davie Stieb, Barry larkin, Griffy Jr. Todd Helton, Joey Votto, Bobby Bonds, and Al Simmons. I personally think that Bill has always underrated Al Simmons, but clearly there is a perfectly good case fo rnot counting him as one of the top 50 players of all time. I will make no attempt to make distinctions among those 18.

Looking at the list of players that Bill submitted for a vote, a number of them--Ford, Gwynn, Winfield, Fingers, Dawson, Ripken, Ichiro and Rivra and chipper Jones (who cost his team a lot of runs in the field) are not close to the top 50 on my chart.

You may have noticed there are no catchers on my list. Why is explained at some length in the book. It is very hard for catchers to demonstrate top-quality value for very long. None of them has been a great hitter for as much as five years, and for most of the history of baseball there is very little that one catcher can do to create more outs than another.

I'll stop there.




5:44 PM Sep 24th
 
Steven Goldleaf
Totally don't get leaving Collins out, especially (Bill hates it when you quote Bill at him) when he wrote in the first Historical Abstract (Hornsby comment, I believe) that Collins (with Joe Morgan) is the clear choice for All-time #1 2B-man. If something's changed about Eddie Collins' stature over the year, I don't remember reading about it. I think Bill just dropped a card there when making his list of automatic selections--at the very least, he needed to include Collins in his initial one-on-one eliminations.
2:28 PM Sep 24th
 
hotstatrat
Kaline is not low media profile in Michigan. He is still regarded with saintly awe by us baby boomer fans - even though we may have enjoyed more the exploits of Horton, Freehan, Lolich, McLain, McAuliffe, Stanley, Northrup, and Hiller, while Kaline and Cash were the respected elder statesmen of those teams.

Perhaps, we need have a systematic way of measuring newspaper headlines, magazine covers, etc., as a way of weighing superstardom. I think that is the sort of superstar Bill is looking for, at least, it justifies including Reggie Jackson over his superiors Frank Robinson & Al Kaline, his equal Frank Thomas, and, I suppose without any knowledge of his acclaim during his playing years, the very hard to dismiss Eddie Collins.

I agree with Bill about the dangers of accuracy illusion, but if sabermaticians have gone through a stringent process of getting the most accurate measurement possible, you better have some reasonable points about some things that were overlooked.

Why would Bill override WAR and presumably Win Shares record in to order to rate Reggie and Ortiz over Collins and F. Robby? Championships?
Eddie Collins won four of them and was in two others.
With more teams in play, Frank Robinson won two of them and was in four others. Robinson played six years competing agains 7 other teams in the league, all 7 years of 10 team leagues, and his final six years as a regular in the 6 team division single best-of-five play-off format.
Beginning his career in the final year of 10 team leagues, Reggie did win five championships - winning the World Series MVP twice - however, missing one of those World Series championships entirely due to an injury. He was on the losing side in another World Series and made the playoffs four other times.
As we all know, David Ortiz has three championships, but no other World Series appearances, but does have four other post-seasons including one an early one with Minnesota.
So, Reggie and Ortiz do have slightly more impressive post season bragging rights, Eddie Collins and Frank Robinson have had their share. David Ortiz may be the god in Boston that Kaline has been in Detroit and Frank Thomas was in Chicagon (assuming he was), but my national media exposure meter doesn't register Ortiz any higher for the other two or Frank Robinson.

My media meter is admitedly not very accurate, but I don't know who does have an accurate one. I know Reggie practically breaks the instrument in the 1970s and early '80s. Nonetheless, I can't see how we can include Ortiz and leave out Eddie Collins and Frank Robinson. Is there a gigantic Dominican love affair for Ortiz that makes up for this discrepency that Miguel Cabrera lacks from Venezuela?​
12:33 PM Sep 24th
 
Brock Hanke
So far, this has been very instructive and also very funny at times. Sister vs. Vance was "Which great half a career do you prefer?" I assumed that, as soon as the winner ran into someone with a whole career, he'd get creamed. Which is what happened. What I don't understand is how Dizzy Dean keeps winning. I've lived my life in St. Louis. I LOVE Dizzy Dean. He had half a career, just like Sisler and Vance. Mickey Cochrane's career was shortened by injury, but not THAT short. Heck, I'd have voted for Hubbell if I had a Twitter account. Greenberg (or Jackson) is going to run into the same problem next round, I would assume; it's really the only match of this round that puts up two short careers against each other. I agree with you about Kaline and Clemente, but Clemente was unusually famous, while Kaline had a very low media profile. You could sorta see it coming. I've thought about this process, and there's only one player I can think of so far who I think should have been in but who is not: Hoyt Wilhelm. Not the greatest closer of all time, but probably still the greatest reliever.
5:31 AM Sep 24th
 
Robinsong
I found it interesting that Bill did not view Reggie Jackson as even a controversial choice for the top 40 and did not include Eddie Collins as even a candidate for the top 50. By WAR, Collins is ranked 13th by career total and even higher by top 7 seasons. His teams won several pennants. He is the only player in the top 33 by career WAR not still in the running. I think his exclusion is an error. By contrast , Reggie! ranks 8th among right fielders (behind Kaline, as well as Clemente and F Robinson) in both career and top 7. I know the 70s was a lean decade, and I am a big Reggie fan, but I think he should have been in the runoff
5:41 PM Sep 23rd
 
bjames
GFletch

I have a sort of belief about sabermetrics that the more accurate we try to make our measurements, the less I trust those measurements after a certain point. Not sure if you would agree with that or not.


Well. . .I wouldn't be a fanatic about it, but yes, that is how I think. People put a lot of energy into pursuing an illusion of accuracy. There is a generic corollary or the principle that you can't derive three-decimal outputs from two-decimal inputs. Clemente was a great player. Kaline is underrated. This is about as much as I am certain of. ​
4:28 PM Sep 23rd
 
Manushfan
No Al Simmons? Hmmmmm.
4:18 PM Sep 23rd
 
Gfletch
Bill, you once wrote something like, 'We have a responsibility to be accurate only to the point where accuracy can reasonably be expected to affect our conclusions.' Hope I have that right; anyway, I like that, and I like how I think you are taking that attitude with this project (it's not about whether Kaline is a better choice than Clemente, or not, because it doesn't mean much to the project).

I have a sort of belief about sabermetrics that the more accurate we try to make our measurements, the less I trust those measurements after a certain point. Not sure if you would agree with that or not.

Looking forward to the rest of the twitter votes, and the rest of your project when you write it up.
3:59 PM Sep 23rd
 
 
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy