The Better Than Best

October 30, 2020
                            The Better Than Best

            This research started when, working on an earlier study, I happened to notice that Jackie Robinson had a higher WAR (Baseball Reference) than the NL MVP for four straight years (1950-1951-1952-1953).   I was a little surprised at that, and I think most people would be; I believe a lot of people think of Jackie as a pioneer, but actually underrate him as a player.   Anyway, that got me to wondering who else had a higher WAR than the MVP and how many times, so. . .here we are. 

            Some interesting stuff turned up along the way. . ..for example, we all tend to look at the 1964 NL MVP Award, given to Ken Boyer, with some disdain, and this weakens Boyer’s case as a Hall of Fame candidate.  But Boyer gains back that credibility, perhaps, if you realize that Boyer had a higher WAR than the NL MVP in three other seasons—1956 (Don Newcombe), 1960 (Dick Groat) and 1961 (Frank Robinson.)   

            Three, four times having a higher WAR than the MVP isn’t historically super-rare; in history there are 32 players who had a higher WAR than the elected MVP four times, and 74 players who have done it three times.  Of course, most of those guys are already in the Hall of Fame.  Of the 74 players who have had a higher WAR than the MVP at least three times:

48 are already in the Hall of Fame,

2 are not yet eligible,

6 are steroid guys with Hall of Fame numbers, and

18 are players who just haven’t been selected. 

 

Many of those 18 are the "Saberfavorites" that we all like to push for more recognition.  Ken Boyer and Bucky Walters are actually only players in baseball history who won an MVP Award and had more WAR than the MVP three other times, but are not in the Hall of Fame.  

We usually start something like this with "The player who had the most seasons with higher WAR than the MVP was. . ." but that would be boring.   The player who had the most season with higher WAR than the MVP was Willie Mays, but you probably could have guessed that anyway.   Willie had higher WAR than the MVP a remarkable 11 times:

1955 (Roy Campanella)

1956 (Don Newcombe)

1957 (Henry Aaron)

1958 (Ernie Banks)

1960 (Dick Groat)

1961 (Frank Robinson)

1962 (Maury Wills)

1964 (Ken Boyer)

1966 (Roberto Clemente)  

1971 (Joe Torre)

In 1971 the 40-year-old Mays played only 136 games, hitting .271 with 18 homers, 61 RBI, which hardly sounds like MVP numbers, but

(a)  Mays drew 112 walks,

(b)  Leading to a league-leading .425 On Base Percentage,

(c)   And was 23-for-26 as a base stealer,

(d)  While the elected MVP, Joe Torre, is regarded by Baseball Reference as a defensive hookworm disease, leading to a relatively low 5.9 WAR—relatively low for an MVP. 

The top nine are:

 

First

Last

Count

Willie

Mays

11

Hank

Aaron

8

Alex

Rodriguez

7

Stan

Musial

7

Bobby

Grich

6

Roger

Clemens

6

Barry

Bonds

6

Ted

Williams

6

Warren

Spahn

6

 

            Saberfavorite Bobby Grich is kind of our lynchpin for the bit; he’s the only one of those guys who never WON an MVP Award.    It’s Bobby Grich and eight of the greatest players in baseball history.  There are nine more guys who did it five times, but they’re all in the Hall of Fame except the steroid guy:

Mark

McGwire

5

Mickey

Mantle

5

Bert

Blyleven

5

Wade

Boggs

5

Ken

Griffey J

5

Mel

Ott

5

Robin

Roberts

5

Frank

Robinson

5

Mike

Schmidt

5

 

            Blyleven, of course, was a Saberfavorite like Grich until he finally got elected.  At four times we have 14 players:  nine Hall of Famers, two steroid guys, two guys who have loud and active Hall of Fame cheering sections, and Sal Bando:

Sal

Bando

4

Kenny

Lofton

4

Andruw

Jones

4

Rafael

Palmeiro

4

Manny

Ramirez

4

Al

Kaline

4

Jim

Palmer

4

Jim

Thome

4

Pedro

Martinez

4

Jackie

Robinson

4

Duke

Snider

4

Eddie

Mathews

4

Phil

Niekro

4

Greg

Maddox

4

 

 

            At three seasons with higher WAR than the elected MVP we have a whopping 42 players, so I’ll break them down into groups.   26 of them are Hall of Famers:

Lefty

Grove

 

Frank

Thomas

 

Ralph

Kiner

Bob

Feller

 

Roberto

Alomar

 

Bob

Gibson

Early

Wynn

 

Edgar

Martinez

 

Ron

Santo

Brooks

Robinson

 

Carl

Yastrzemski

 

Roberto

Clemente

Rod

Carew

 

Derek

Jeter

 

Tom

Seaver

George

Brett

 

Roy

Halladay

 

Joe

Morgan

Paul

Molitor

 

Bill

Terry

 

Ozzie

Smith

Rickey

Henderson

 

Arky

Vaughan

 

Randy

Johnson

Dave

Winfield

 

Johnny

Mize

     

 

            One is Mike Trout, still active, so that’s 27.   One is a steroid suspect, Albert Belle.   The other 14 are these guys:

Bucky

Walters

 

Bret

Saberhagen

Minnie

Minoso

 

Teddy

Higuera

Camilo

Pascual

 

David

Cone

Luis

Tiant

 

Bernie

Williams

Roy

White

 

Nomar

Garciaparra

Buddy

Bell

 

Ken

Boyer

Dave

Stieb

 

Will

Clark

 

            Lot of Saberfavorites in there.  I’m a "Minnie Minoso for the Hall of Fame" guy.   Love Buddy Bell, Stieb, Roy White, Camilo, Will Clark, David Cone, Bernie.   There are 78 players who had more WAR than the MVP twice, so I’m not going to list them all.   The two most surprising names on that list, for opposite reasons, are Lonnie Smith and Albert Pujols.   Lonnie Smith had higher WAR than the NL MVP in 1982 and 1989.    1982, he hit .307, scored 120 runs and had 68 stolen bases, so that’s not too surprising.   1989 is one of those fluke things, like Mays in 1971, that doesn’t make intuitive sense; he played only 134 games, but led the NL in on base percentage (.415) and hit 21 homers.  He was 11th in the MVP voting, but the MVP, Kevin Mitchell, had a great year with the bat but, like Torre, very negative dWAR. 

            Pujols you might guess would be one of those guys who did this seven times, but actually he beat the MVP only twice, in 2006 and 2010.  Part of the trick there is that by Albert’s era, the MVP Award was less likely to go to some yahoo who caught the attention of the voters.   Some players piled up a lot of chips in this poker game because they played through eras in which the voters made a lot of icky choices.   Andre Dawson won the NL MVP Award in 1987, but there were 15 guys in the league who had higher WAR, so they all pick up a chip—and then the selections in 1988 and 1989 weren’t great, either, Kirk Gibson and Kevin Mitchell.  A lot of people took them to the basket, too.    There have been several eras like that in baseball history—the American League in 1981 (Rollie Fingers), 1984 (Willie Hernandez), 1987 (George Bell), 1992 (Dennis Eckersley).  Don Mattingly (1985) and Robin Yount (1989) are not controversial selections, but neither of them was in the top five in the league in WAR, either.   A player in that era has a lot of chances to have a higher WAR than the MVP—whereas in the NL in the Pujols/Bonds era, it was somewhat more of a challenge. 

            One COULD do an assessment of the MVP voting in an era, I guess, by counting how many players in that league in that decade had higher WAR than the elected MVP.   I’m not going to do that.   (American League, 1940s, 19 players.  American League, 1960s, 27 players.  American League, 1980s, 64 players.  OK, I’m not going to do any more of it.) 

           

            And then, just because it was relatively easy for me to do, I also decided to count how many times each player had more Win Shares than the elected MVP—and that turned up a surprise right off the bat.   Willie Mays, who dominates the list in WAR (11 to 8 over Aaron) actually does not lead the list when we switch to Win Shares.   He falls behind Aaron, 9 to 7.  So why does this happen?

            Two reasons.   One is, Win Shares pays attention to actual wins, while WAR (I think—somebody can correct me if I am wrong)—WAR extrapolates expected wins from runs and runs allowed, but pays no attention to actual wins.  Suppose that two teams both (or each) score 750 runs and allow 700, but one team finishes 95-67, while the other finishes 81-81.   WAR treats the players on these two teams the same, on the theory that deviations from expected wins are just luck and should thus be ignored, while Win Shares bases its analysis on ACTUAL wins, on the theory that you have to explain what actually happened, or what the hell do we need you for?   We know who had pretty numbers without running any advanced calculations; what we are looking for is who had the most actual value. 

            In 1961, for example, Willie Mays and Frank Robinson had essentially identical numbers, so much so that if you swapped their batting lines from 1961 in their career records, nothing at all would look out of place:

Last

G

AB

R

H

2B

3B

HR

RBI

BB

SO

GDP

SB

AVG

OBP

SLG

Robinson

153

545

117

176

32

7

37

124

71

64

15

22

.323

.404

.611

Mays

154

572

129

176

32

3

40

123

81

77

14

18

.308

.393

.584

 

            However, Robinson’s team scored only 710 runs, allowed 653, making them only +57, so we would expect them to finish with a won-lost record of 83-71.  They outperformed that by ten games, finishing 93-61, and thus won the National League pennant.  Willie’s team, the Giants, outscored their opponents 773-655 (+118), so we would expect them to finish 90-64—but they didn’t.   They finished 85-69.  There’s a 15-game swing in the won-lost records, based on scoring runs that won ballgames, rather than just scoring runs. 

            WAR just ignores that, and thus implicitly assumes that the Giants are a better team than the Reds.   My opinion is that this is Dumb, but then, nobody asked me.  But when you apply it to individuals, Mays winds up with more WAR (8.7 to 7.7), while Robinson winds up with more Win Shares, 41 to 38.  I feel strongly that my conclusion is right, Win Shares is right, and theirs is wrong.    And I still expect to win the argument eventually, in history, simply because I am right and they are wrong. 

            The other three splits between Win Shares and WAR (relevant to Willie Mays) are 1956, 1957 and 1971.   1971 we already talked about, Joe Torre hit .363 and drove in 137 runs, while Mays was 40 years old, but WAR still likes Mays.  OK, you can have Mays; I’ll take Torre.  

In 1956 Don Newcombe went 27-7 with a 3.06 ERA, but is assigned an inexplicably low WAR by Baseball Reference.  Bob Friend, pitching in  a much better pitcher’s park, went 17-17 with a 3.46 ERA.   Newcombe had 93 more strikeouts than walks (139 to 46) and a ratio better than three to one, Friend had 81 more (166 to 85) and a ratio less than two to one.  Newcombe as a hitter had a .654 OPS, scored 13 runs and drove in 16; Friend had a .390 OPS, scored 7 runs and drove in 8.

Nonetheless, WAR considers Friend (5.2 WAR) to be essentially even with Newcombe (5.3 WAR).   No doubt Friend has his advantages, hiding somewhere in his numbers. Why Baseball Reference WAR sometimes does stupid shit like that I really don’t know or want to know, but anyway, Win Shares considers Mays and Newcombe to be even, at 27 Win Shares each, both of them AMONG the best players in the league, but neither one at the top.  WAR considers Mays to be 40% better, 7.6 to 5.3.     

            The other difference between Win Shares and WAR, relevant to Mays, is that while both place value on defense, WAR is more trusting of defensive numbers than is Win Shares.   WAR gives Mays an 11-8 lead over Aaron in "better than the MVP" seasons, while Win Shares makes it 9-7, Aaron.   This happens essentially because, while we both agree that Mays was a more valuable outfielder than Aaron, WAR puts Mays ahead by a much wider margin than Win Shares does.  In 1957 Aaron, the elected MVP, has 8.0 WAR, while Mays has 8.3.   Win Shares has Aaron at 35 Win Shares, leading the league, while Mays is at 34, alone in second place.   And I honestly don’t know who is right in that one; I don’t think anyone does, but that makes Aaron +1 and Mays -1 in the Win Shares-to-WAR matchup, and that (combined with the other three that Mays loses in other years) puts Aaron ahead of Mays in the Win Shares count. 

               There are 772 players in baseball history (1931 to 2019) who have had more WAR than the MVP, whereas there are only 528 players who have had more Win Shares than the MVP.   This happens because (1) Win Shares tracks more closely with MVP voting than WAR does, and (2) there are more ties in Win Shares than there are in WAR, thus reducing slightly the number of cases in which anybody is better than anybody. 

            In Win Shares, the top 12 players in this category are:

 

First

Last

Count

Hank

Aaron

9

 

 

 

Willie

Mays

7

Mel

Ott

7

 

 

 

Barry

Bonds

6

Mickey

Mantle

6

Stan

Musial

6

Mike

Schmidt

6

 

 

 

Will

Clark

5

Eddie

Mathews

5

Manny

Ramirez

5

Alex

Rodriguez

5

Ted

Williams

5

 

            In Win Shares, we have 12 players at "four" and 24 players at "3", but I’ll spare you the lists.   Bobby Grich drops all the way to 2 in this count, but then, Joe DiMaggio is at 2 either way—two by WAR, two by Win Shares. 

            You could build this out into a larger study by:

1)     Including fWAR,

2)     Including counts of who LED the league in each category (WAR, Win Shares and fWAR),

3)     Creating a point system for winning the MVP Award, leading the league in one category, or having more than the MVP in some category, and

4)     Doing some studies like the one I suggested about the number of discrepancies between WAR and MVP voting, by league and decade.

 

I’m not doing that stuff; this is enough for me.  I was just interested in how unusual it was for Jackie Robinson to beat the MVP four straight times, and we have an understanding of that now, so I’m done, I’m out.  Thanks for reading.  

 
 

COMMENTS (10 Comments, most recent shown first)

KaiserD2
Defensive Regression Analysis, which many feel is the best historical fielding statistic, confirms that Gus Bell was a dreadful outfielder. in five years, 1954-58, he surrendered -22, -23, -40, -27 and -17 runs, relative to an average centerfielder. the 1956 Reds, by the way, were particularly noteworthy in that despite excellent fielding performances by Roy McMillan, Wally Post, and Frank Robinson, Bell, Ray Jablonski, and Johnny Temple were so dreadful that the teams' fielding as a whole cost them -49 runs or five games. Their pitching, which took a lot of heat at the time, was actually quite good, but the fielding concealed that.
7:52 PM Nov 12th
 
bhalbleib
"I seem to recall Bill making a comment in one of the Historical Abstracts, when discussing Musial, that the voters could have simply given Stan the MVP every year, but got tired of that so decided to give it to someone else every once in awhile."

I think Stan's inability to win after 1948 had more to do with the Cardinals fall from the top of the standings than Stan-weariness by the voters. He finished 2nd to Jackie in 1949 ,whose team won the pennant (the Cards finished 2nd too). He finished 2nd to the Phillies relief pitcher in 1950, whose team won the pennant. (the Cards finished 5th) He finished 2nd to Roy Campanella in 1951, whose team famously did NOT win the pennant, but came about as close as you can to winning one without actually doing so (the Cardinals were a distant third that year). The remainder of the 50s, there frankly were better candidates than Stan every year and, of course, the middling place in the standings that the Cardinals occupied every year didn't help him. He did finish 2nd in 1957 when he won the batting title (and the Cardinals also finished 2nd, boosting his candidacy), but the winner, Aaron, was appreciably better as was Mays (who finished 4th in the balloting) and probably Frank Robinson and Eddie Mathews as well.
1:47 PM Nov 2nd
 
Jaytaft
WAR really hates Gus Bell. In the HBA list of best centerfielders ever, you ranked the 4-time All Star at number 58. In your 2019 "100 Bestest" list, he's ranked number 53, right behind Hall of Famer Lloyd Waner. JAWS ranks him number 167. Hall of Stats puts him at number 193, right behind Craig Gentry, a bench player who had just 1242 career at-bats and 7 career home runs.

"In addition to his strong bat, Bell was one of the most gifted defensive outfielders of his era, regularly placing among the National League's leaders in putouts, assists and fewest errors committed."
--Reds Hall of Fame and Museum
11:02 AM Nov 2nd
 
Brock Hanke
Well, comparing WAR to Win Shares: I'm involved in a project that is trying to determine the best players in baseball in each year. We have done many years now, so it's possible to look at trends. I approach these rankings by using the ordinals - who is first by WAR, who is first by WS, who is second, etc. And I found something. Over time (meaning the time of the seasons, not the time in the development of WAR), WAR has been ranking pitchers higher than WS since at least 1940, and the gap has been steadily growing, not narrowing. Modern WAR pitcher ordinals are, when examining recent seasons, FAR stronger than the WS ordinals for the same season. I am not sure what to make of this, but it does seem like a good place to look, if you're trying to figure out what the differences are between WAR and Win Shares.
4:29 PM Oct 30th
 
MarisFan61
-- Indeed "WAR" ignores actual wins; it is exactly as stated here.
I agree that it is (in your words) DUMB, and for a broader reason than the thing itself: It seems to me that besides being dumb in its own significant right, it reflects a dumb generic view of the game in general, in ways that can impact each and every decision made within the system.
Maybe to put it in plainer English: The thought process behind the Win Shares system, from what I could gather, seems every step of the way to be guided by an underlying consciousness of what wins games. The "WAR" system, again from what I can gather, seems more merely formulaic and removed in all its inner workings.

-- About Ken Boyer: I am not surprised. As I've said here many times, he is the main player of that era on whom I've been surprised that he isn't more recognized and that he hasn't gotten more seriously considered for the Hall of Fame, and one of the top 2 in history for the latter (the other being Keith Hernandez).
2:19 PM Oct 30th
 
kingferris
I see value in both stats...the times I check up for Win Shares is when a player's value is affected by their team deviating from the pythag even though said player's own clutch/high leverage stats run counter. I remember Trout having a low (for him) win shares total at some point during a season a couple of years ago because the Angels were seriously underperforming their run differential (who could have guessed?) But Trout's numbers in high leverage situations were even better than his overall stats; he was just getting burned by his teammates. Wally Berger's another example, because he had very similar stat lines in 1934-1935 but hugely different win shares (last time I checked) because the Braves completely cratered in the latter season. But Berger's clutch numbers were terrific in 1935, better than his overall line with RISP, RISP and two outs and in late & close scenarios. He was just stuck dragging a bad team around his neck.

For Baseball Reference, the biggest problem is that they come up with a team defensive value to gauge pitchers' luck. Sometimes, pitchers on poor defensive teams will still outperform their fielding independent ERA; the line drives are hit at people, they get the 1-out double play with the bases loaded, etc. B-Ref essentially double counts, giving them credit for pitching in front of a bad defense even if said bad defense didn't actually hurt their stat line. Aaron Nola's 2018 WAR (10.2) is a good example of this dynamic. The reverse also happens; pitcher's getting "penalized" for pitching in front of strong defenses even if their run prevention doesn't actually reflect any good luck.
1:43 PM Oct 30th
 
phorton01
Regarding the reluctance of voters to give one player too many MVP awards ....

I seem to recall Bill making a comment in one of the Historical Abstracts, when discussing Musial, that the voters could have simply given Stan the MVP every year, but got tired of that so decided to give it to someone else every once in awhile.
12:15 PM Oct 30th
 
bjames
Very interesting article. To a Non-stat guy like me, it would appear that in earlier eras, there was a reluctance by the voters to give one player "too many" MVP awards.


I don't know. Trout doesn't win it every year, or every year when he could win it. That's probably a little different, but I'm not sure it is measurably different.
11:33 AM Oct 30th
 
337
How far we've come from "I don't do WAR."
10:00 AM Oct 30th
 
archieleach
Very interesting article. To a Non-stat guy like me, it would appear that in earlier eras, there was a reluctance by the voters to give one player "too many" MVP awards. Would this be a factor in some players (Mays for example) racking up a lot of chips in this type of counting?
9:57 AM Oct 30th
 
 
©2020 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy