Final Report on the 50 True Superstars Project

September 26, 2017

Final Report on the Top 50 Superstars Project


              So, my project to identify 50 players who could be identified as "true superstars" has wound up identifying 52 players.  The whole point was to make it to 50, but I only got to 52.  What happened was, in the semi-final vote I had Ichiro Suzuki matched up against Tony Gwynn, and Gwynn was leading 53-47 with about 1500 votes cast, so I figured that race was over and created a final matchup of Gwynn against Ripken.   But then, after I had started the last vote with Gwynn, Ichiro surged past Gwynn in the late voting, and won that contest 51-49.  Ripken then beat Gwynn (also) 52-48, but 51-49, 52-48; these aren’t good margins.   So I decided the hell with it, and just put all three of them on the list.  The final list is

1900s—Cy Young, Honus Wagner, Nap Lajoie, Christy Mathewson

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, Jackie Robinson

1960s—Henry Aaron, Bob Gibson, Sandy Koufax, Roberto Clemente

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

1980s—Mike Schmidt, George Brett, Cal Ripken, Tony Gwynn, Rickey Henderson

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

2000s—Barry Bonds, Derek Jeter, Alex Rodriguez, David Ortiz, Albert Pujols, Ichico Suzuki,

            Pedro Martinez

2010s—Clayton Kershaw, Mike Trout


              After the voting I realized that I could construct "power charts" comparing all 54 of the players by using the method we use to compare basketball teams or football teams which haven’t played each other.   Florida plays Arizona and loses 71-67, then Oregon plays Arizona and loses 84-48, you figure (based on that one matchup) that Florida is probably stronger than Oregon.   Same thing here.  Cal Ripken beats Brooks Robinson 66-34 and beats Jim Thome 85-15,you figure Robinson is a stronger candidate than Thome.  You make calculations based on ALL of the common opponents, you have power rankings.

              You have to make some assumptions.   You can’t assume there is a straight-line relationship.   In other words, the fact that Albert Pujols beats Billy Williams 90-10 doesn’t mean that Pujols is 9 times better than Williams in the eyes of the voters.  You have to make assumptions, which means that you can be wrong.  But given that, we can make power ratings for all 54 players in the competition for the final ten spots (which turned out to be 12).   The rankings are.






Rickey Henderson



Albert Pujols



Pedro Martinez



Roberto Clemente



Jackie Robinson



Frank Robinson



Nolan Ryan



Christy Mathewson



Pete Rose



Warren Spahn



Cal Ripken



Carl Yastrzemski



Tony Gwynn



Ernie Banks



Steve Carlton



Mariano Rivera



Ichiro Suzuki



Brooks Robinson



Frank Thomas



Robin Yount



Chipper Jones



Miguel Cabrera



Ryne Sandberg



Joe Jackson



Hank Greenberg



Rod Carew



Al Kaline



Andre Dawson



Harmon Killebrew



Roberto Alomar



Wade Boggs



Billy Williams



Dave Winfield



Jim Palmer



Don Drysdale



Mark McGwire



Ferguson Jenkins



George Sisler



Whitey Ford



Eddie Murray



Jim Thome



Dale Murphy



Buster Posey



Goose Gossage



Rollie Fingers



Johnny Mize



Dizzy Dean



Mickey Cochrane



Robin Roberts



Gaylord Perry



Carl Hubbell



Charlie Gehringer



Chuck Klein



Dazzy Vance



              My analysis of the data suggests that Carl Yastrzemski would be an even match against Tony Gwynn if they were pitted head to head, and that Warren Spahn would probably beat Ichiro.   Maybe I’ll test those names and see what happens, but anyway, that’s like any other tournament; there is luck involved in who you have to play, and the teams that make the Final Four aren’t always the four best teams.  

              I don’t really care who makes the list; I mean, my idea was that any of the 54 candidates would be acceptable to me as one of the last ten "true superstars", or I wouldn’t have put them in the vote.   I listed them because I couldn’t decide, so I thought I would let the fans decide, and they did a good job, and we had fun doing it and I wound up with a good list.   I’ve got no objections to it.

              However, I will note that when I analyzed the data, I had to admit that the "recency bias" which was much discussed while the voting was going on did turn out to be a larger factor in the results than I thought it was at the time.   I mostly matched up a player from the 1930s against another player from the 1930s and a player from the 1950s against another player from the fifties or sixties, which tends to hide the recency bias.   But when you run the power rankings, you can see that there is a concentration of 1930s candidates at the bottom of the list, and that some of the more recent stars perhaps did better in the voting than was fully warranted.   I will now use the list to study what it was that I was trying to study, which was the connection between having a superstar in his prime and the direction of the franchise.   Thank you all for playing.  


COMMENTS (34 Comments, most recent shown first)

Nice article as usual. Thank you.

Missing in action: Tom Seaver, Mickey Mantle, Ted Williams and Johnny Bench.

Take Care,
Tom Nahigian
10:09 PM Oct 6th
337. Ah, yes. I see my mistake now
1:42 PM Sep 30th
Herb, that's just a mistake on your part. Schmidt appears on Bill's original list, along with Brett, representing the 80s. Collins, however, was excluded by name from that list, and never AFAiK had a run off against anyone, despite appearing on many lists including Bill's as the greatest 2bman of all time. In the first Historical Abstract Bill listed him and Morgan as essentially tied for the #1 spot. Players Bill considered Collins' inferior, such as Gehringer and Lajoie and Hornsby, either got placed on the original list of Superstars or got a runoff. I don't get it .
11:03 AM Sep 30th
I don't see Mike Schmidt on the list of 54 either.
10:40 AM Sep 30th
Can anyone explain to me how Eddie Collins didn't warrant a run-off head-to-head contest with, say, Charlie Gehringer? I'd argue that he actually warrants an automatic inclusion on Bill's top 40 list, but not to get a run-off at all seems to me absurd.
6:43 AM Sep 29th
I completely agree with Guy 123 and his position on Jeter. I don't know if that makes him happy or no, but I say that he is 100% correct. And that's all I have to say about that.
7:57 PM Sep 28th
definition of a superstar "a high-profile and extremely successful performer or athlete"-sounds like Jeter to me....
7:10 PM Sep 28th
Does this not assume you know his actual value?
Not precisely, no. It assumes that I am aware of the biases in the BIS data used for WAR calculations for much of Jeter's career, and that I'm familiar with the most careful studies of Jeter's fielding (Tango's WOWY methodology). A fair reading of that evidence suggests that Jeter cost the Yankees many more runs in the field -- at least 70, probably more -- than the estimate used for WAR.

Whatever debate there once was about whether Jeter's fielding was historically bad -- some said he was really better than his raw stats implied because the Yankees generated fewer opportunities at SS -- that case should now be considered closed. In his final 3 seasons in the field (2011, 2012, 2014), the Yankees' average AL rank for SS assists was 15th (last). In the four recent non-Jeter seasons (2013, 2015-2017) NYY has ranked 8th in SS assists. Even if we discard his miserable 2014 season, the Yankees averaged 391 SS assists in 2011-12. In the three post-Jeter years they have averaged 438 assists, a difference of 47 outs! It's been an enormous transformation, far greater than what BIS/DRS was able to capture.

Jeter was a great hitter and baserunner, but his fielding was cosmically bad. There's simply no way he is one of the 50 top superstars. The interesting question is whether he should have even made the fan ballot of 54.

6:52 PM Sep 28th
Except his banker
6:29 PM Sep 28th
"fielding metric that significantly *overestimates* his actual value"-guy123 referring to Derek Jeter.

Does this not assume you know his actual value?

I mean I thought some of the sabr crowd thought they had figured it all out-but jeez come on. Nobody knows his actual value.​
5:59 PM Sep 28th
Now I get it.
You're pretending for this moment to be a slave to "WAR," which I'm pretty sure you aren't really.

In any event, even if one is a slave to that, he shouldn't have that be a basis for finding it puzzling that others have views that don't follow it.
4:03 PM Sep 28th
It's interesting to see the differences between fan perceptions and peak WAR. The fans chose not to admit 5 of the top-50 WAR7 Players:
Boggs (56.2 WAR7)
Yaz (55.5)
Roberts (54.8)
Carlton (54.3)
Perry (52.8)
Both Yaz and Carlton came very close to election. Robin Roberts and Gaylord Perry may be a bit undervalued by fans, though neither are what you'd call automatic selections. Boggs (ranked 31 out of 54) is perhaps the player whose WAR most exceeds fans' perceptions.

Fans were not given a chance to evaluate 5 of the top-50 WAR players, and 2 of the 3 top catchers. None of the 5 non-catchers are obvious top-50 superstars, but they do outrank plenty of guys on the 54-man fan ballot:
Ferrell (55.0)
Niekro (54.5)
Matthews (54.4)
Santo (53.8)
Newhouser (52.4)
Carter (48.3)
Piazza (43.1)
Perhaps there is a subtle pattern here, in terms of which superWAR players seem obviously (to Bill) not to be "superstars." I don't see it.

(Jeter, FYI, has a WAR7 of just 42.2 -- and that’s using a fielding metric that significantly *overestimates* his actual value. Gwynn, Ryan, Ichiro, and Rose are all in the same ballpark (41-45), with Reggie and Koufax a tick higher. Ortiz trails at 35.)
12:00 PM Sep 28th
Next most puzzling thing on here:

"Jeter pretty clearly fails that test."

Odd coming from someone who's so into sabermetrics, because it would seem that a big part of the science of the field is knowing when something is clearly shown and when it isn't.

It isn't "clearly" shown by any test.

BTW one might have thought that even if in no other way, you would have realized it ipso facto by Bill's having included him.
10:52 AM Sep 28th
I have trouble understanding anyone finding it puzzling. You should know, very well, what's the basis (or bases) of [Jeter] being seen as a superstar.

And you should know, very well, that Jeter wasn't selected by fan voting. He was included by BJ, who defined "superstars" clearly as players having the greatest impact on their teams' seasonal win-loss records. Jeter pretty clearly fails that test. Moreover, BJ included Jeter as an "obvious" choice that "no one reasonably would argue with." So yes, a very puzzling selection.
7:29 AM Sep 28th
Ahh, the list by decades is the top 40, and the next list is the also-rans.

Them being in different formats is what threw me.

Oh, and if we're talking about impact of guys joinging/leaving teams, you gotta start and end with Lonnie Smith.
5:00 AM Sep 28th
Most puzzling thing on here:

"Jeter is a really puzzling selection."

I can well understand people (highly sabermetric ones anyway) disagreeing with such a pick.
I have trouble understanding anyone finding it puzzling.

You should know, very well, what's the basis (or bases) of his being seen as a superstar.
Even if you think it's misguided, including if you think those bases are fictional. But the beliefs are wide, and strong, and if they're a mystery to anyone, you haven't been paying attention. :-)
9:50 PM Sep 27th
Fans did a pretty good job overall. At the end of the day, I only see 3 fairly clear omissions based on peak WAR: Boggs, Yaz, and Carter. Fenway's peculiarities may impact 2 of those 3, and of course the fans never had a chance to vote for Carter.

Looks like 9 players on the list fell fairly far short of having the needed impact, based on peak WAR, and 4 were picked by the fans (Rose, Gwynn, Ichiro, Ryan). Based on that list, it seems the fans' bias could be as much an over-emphasis on batting average and batting titles as a recency bias. The other 5 were picked by BJ: Jeter, Reggie, Koufax, Campanella, Ortiz. Reggie is at least within spitting distance of the minimum standard, and Bill made clear why Ortiz is on the list. Campanella and Koufax are obviously special "supernova" cases (they didn't contribute long enough to make my list, but YMMV). Jeter is a really puzzling selection.

4:57 PM Sep 27th
Bench, Rose, Seaver and Morgan were on the same team for about 1.5 years. Wasn't at any of their peaks, but still...
3:06 PM Sep 27th
Not that it's important, but list Bonds in the 1990s, which balances out the list, and since Bonds played the full decade, as opposed to the next one, when he was retired due to his inability to get a job.
2:42 PM Sep 27th
Brock Hanke
AH! It took me several minutes after reading shthar's comment, but your list of 54 here is the list of those who did NOT make the 40, but who did make the 54. That's where Hornsby and Gehrig and Foxx and Musial and those guys went. They are already on the list. That means that the top ten from the list make the Final Fifty (or so). The only clear poor results the I see in the Top Ten are Clemente and Ryan. #s 11 and 12 are Cal Ripken, Jr. and Carl Yaz. But I'd put Carlton (#13) above any of the four, so what do I know?
2:14 PM Sep 27th
...and once again, Hubie Brooks gets the shaft.
12:28 PM Sep 27th
Is "Ichico Suzuki" supposed to be Ichiro or Chico Salmon? Honestly, that's a tossup for me.
12:10 PM Sep 27th
Sitting on my desk here at work is a small yellow box labeled "Collect-A-Books Premier Edition 12 Superstars". Inside are 12 little booklets of current and past baseball Superstars, circa 1990:

- Don Mattingly
- Nolan Ryan
- Jerome Walton
- Kevin Mitchel (sic)
- Tony Gwynn
- Dave Stewart
- Roger Clemens
- Darryl Strawberry
- George Brett
- Hank Aaron
- Ted Williams
- Warren Spahn

It says "Buy all 3 boxes to complete your set of 36 superstars." That seems pretty definitive. If you have a list of 50 (or 52) that doesn't include Jerome Walton you got some 'splainin to do.
10:47 AM Sep 27th
I assume you didn't read the other articles or forgot, 77royals, but Bill is using this list to study the impact of superstars joining/leaving teams on their record/history. Paige and Gibson wouldn't really help for the study.
10:10 AM Sep 27th
Not disputing the outcome, but I have to admit I'm surprised that there were only 3 selections from the '20's. I would have thought there would be more.

And it's a shame there isn't a way to get at least Satchel Paige and Josh Gibson in somewhere.
2:53 AM Sep 27th
Rogers Hornsby & Lou Gehrig dissapear? Some bias.
1:38 AM Sep 27th
Big surprises in the order on what we might call the "virtual power ratings" list?
Quite a few, I think.

May I suggest these as big surprises, based on seat-of-pants combination of metrics and traditional reputation:
(not talking about slight stuff, like whether Mathewson seems to belong 1 spot higher so he'd be ahead of Ryan)

-- Kaline being as low as 27; I think usual consensus would have him ~10 spots higher

-- Billy Williams (32) being so high and Winfield (33) so low, not just because they seem in the wrong order but because it doesn't seem they should be close to adjacent

-- Cochrane (48) is surprisingly low

-- Hubbell likewise; would seem to be more comfortable a dozen to a-dozen-and-a-half spots higher

-- Gehringer!! Actually I don't know that this ranking (52) is lower than the current consensus view, to the extent there is one, since his stock seems to have declined in recent decades, but if Carew is #26 and R. Alomar is #30, I have trouble not seeing Gehringer a couple dozen spots higher than how he showed on this.

Of course some of these probably involve the "recency bias" that Bill mentions.
10:58 PM Sep 26th
(OK -- on closer look, I guess indeed Spahn wasn't intended to be included.)
10:11 PM Sep 26th
Names that stick out as not belonging: None :-)
(Great list)

BTW, I wonder about another possible inadvertent omission: It sure looks to me like it's implied that SPAHN belongs.
Was he omitted on purpose?
10:08 PM Sep 26th
Never mind....didn't read the FINE PRINT....votes only for FINAL TEN
10:04 PM Sep 26th
I don't see Yogi or Teddy B either
10:03 PM Sep 26th
Thanks. That 50/52 thing got me again.
9:48 PM Sep 26th
I don't see Mantle
9:47 PM Sep 26th
I think you forgot to list Frank Robinson and Nolan Ryan as part of your 52.
9:26 PM Sep 26th
©2021 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy