By In Stuff

The MVP Formula

As you might imagine, I did not plan on writing a 7,000-word post on MVPs. This was supposed to just be a two paragraph aside about Babe Ruth in my last Hall of Fame post. But, you know, sometimes these things get out of hand.

The starting point was this: Babe Ruth is, more or less, incomparable. I mean that literally — it’s more or less pointless to compare any player in baseball history to him. Satchel Paige is like this too, for different reasons. Ruth was a great pitcher AND hitter. And he so thoroughly dominated his era. It seems to me he will ALWAYS be the best player in baseball history, no matter who come around. This makes his unique in sports, I think. In football, I think, it’s possible for a player to come along who the majority will decide is better than anyone before him. It’s obviously true in basketball — the “greatest player in basketball history” fluctuated many times between Russell and Chamberlain to West and Robertson to Kareem to Larry and Magic to Jordan, and there are people now throwing Kobe and LeBron into the conversation. In hockey, it will be tough for anyone to ever be better than Gretzky to moderate hockey fans — the Great One certainly seems the greatest ever to me — but I know intense fans who insist Gretzky isn’t the best ever RIGHT NOW.*

*This has been explained to me at great length — stuff about how Gretzky only played one side of the ice, how the sport has changed and would never allow him to just stand behind the net, and so on — but to be honest I’ve never really understood it. I like hockey, and have been watching it more closely this year, but I’m still a helpless novice when it comes to the sport.

But Ruth … how can you beat a hitter who outhomered every team one season AND ALSO led the league in ERA when he was 21 AND ALSO went 3-0 with an 0.64 ERA in the World Series AND ALSO led the league in OPS every year but one from 1918 to 1931? You can’t. Ever. Williams may have been a better pure hitter — Bonds too, in his bulked up years — but Ruth will always have that pitching trump card.

Which made me wonder: The BBWAA started its MVP award in 1931 — coincidentally, the last year Ruth led the league in OPS. But what if the MVP had started in 1900. How many MVP awards would Babe Ruth have won? I mentioned that he led the league in OPS 12 times. He led the league in WAR 11 times. Perhaps more to the time, he led the league in homers 11 times, including 1918 when he was still a pitcher (he started 20 games, went 13-7 with a 2.22 ERA).

I think he was the best player in the league every year but two from 1919 to 1931. He also was probably the best pitcher in the league in 1916. Thing is, that many of those years he wasn’t just the best player in the league — nobody else was on the lead lap.

Here are the players I think MAY have been contenders to win the MVP instead of Ruth:

1919: Maybe Shoeless Joe Jackson

1920: Nobody

1921: Nobody

1922: Ruth’s first off year.

1923: Nobody

1924: Nobody

1925: Ruth’s second off year.

1926: Nobody

1927: Lou Gehrig

1928: Gehrig

1929: Jimmie Foxx

1930: Gehrig

1931: Gehrig

Could you imagine if they gave the award in 1919 to Shoeless Joe Jackson? That would make this Ryan Braun controversy look like an episode of Bewitched*. Realistically, Ruth had to win at least six awards — 1919-21, 1923, 1924 and 1926. Nobody was even close to him. But he was probably the best player every year from 1927-31 also. I’m sure the BBWAA at the time would have TRIED to give awards to Gehrig or Foxx just to mix things up. But Ruth probably wins eight or nine, maybe 10 MVPs. And a Cy Young in 1916.

*My daughters have taken a sudden liking to the show “Bewitched.” And it reminds me of something that has always bothered me — was Darrin insane? I say yes. He had a wife that looked like Elizabeth Montgomery and could do pretty much unlimited magic. She constantly wanted to use that magic to HELP him. And he basically yelled at her a lot and told her to STOP using that magic and become a suburban housewife? Huh? … The question goes double for Larry Hagman in “I Dream Of Jeannie?”

But this led me to a whole different question — and this ridiculously super sized post. Bill James had a fascinating series of articles on his Web site last week about MVP award bias. He had a fun way to try and root out that bias: He looked at all the MVP choices that clashed with the Win Shares leaders. Then he tried to find out WHY the voters chose that MVP rather than player Bill’s stats tell him was the best in the league that season. He looked at various categories. Did that player have more RBIs than the Win Shares winner? More hits? More stolen bases? Fewer strikeouts? Did he come from a bigger city? Did he play a more demanding position? Was race a factor? Was age a factor? And so on.

Obviously, it’s just a fun study — Bill concedes the point that the study could actually show bias in Win Shares rather than bias in the MVP voting (but he doesn’t think so). I fully expected to find that MVP voting has been driven by:

1. RBIs

2. Team success

3. Various theories about clutch play and leadership.

And Bill found some evidence about those things. But he developed a surprising theory. He determined that perhaps the biggest bias in MVP voting has been AGAINST hitting. Now, by “hitting” he means hitting value — on-base percentage, slugging percentage, runs created, that sort of thing.

And looking at Ruth’s years made me wonder: How often does the best hitter in the league NOT win the MVP award? And when they don’t win the MVP award, well, why not?

Of course, picking the best hitter in the league every year is a trick unto itself. Some might say the batting champion was the best hitter. Some might say the home run leader. Some might say the RBI leader. Some might say the on-base or slugging champ. Here, we would normally delve into some of the more advanced statistics — use wOBA or Base Runs or WAR Batting Runs or Runs Created or OPS+ or something.

But instead, I’m going to use raw OPS — the simple addition of on-base percentage and slugging percentage. No park factors. No league adjustments. Just OPS. Why? Well, I don’t want this thing to become about the statistic or choice. OPS seems to be mainstream enough now that people aren’t arguing about it much.

OK: There have been 163 MVPs since the BBWAA started the award in 1931. The odd number is because of 1979 when Keith Hernandez and Willie Stargell tied for the award. Neither one of them was the OPS leader that year — that was DAVE KINGMAN, believe it or not — so we’ll keep referring to 162 awards.

Of those 162, how often do you think the OPS leader was MVP? I realize that nobody was figuring OPS until recent years, but the point is not figuring the statistic but determining the best hitter in the league. And the best hitters — Ruth, Williams, Musial, Mays, Mantle, McCovey — usually lead the league in it OPS. Every Triple Crown winner has also led the league in OPS. Just about all of the classic great years — Ruth in ’27, Williams in ’41, Musial in ’48, Mays in ’54, McCovey in ’69, Brett in ’80, Thomas and Bagwell in ’94 and so on — are also OPS leaders. So how often has the OPS leader won the MVP? Three-quarters of the time? Half the time?

Answer: 56 out of 162.

That’s just a touch over one-third, if you are scoring at home.

That seems low to me. I’m not saying the OPS leader should be MVP every year — heck, I just pointed out that in 1979 Dave Kingman led the National League in OPS, and Dave Kingman absolutely should not have been MVP. But assuming the OPS leader is often the best hitter in the league, I would have expected more than one-third of the MVPS to be OPS leaders.

OK, so that leaves us with 106 awards where the OPS leader and MVP were different.

Of those, 21 MVPs were pitchers. That could be a different post — the value of hitting vs. the value of pitching.

In another 39 cases, I decided that the difference in OPS between the leader and the MVP was too small to decide conclusively that the OPS leader was a better hitter than the MVP. I put the cutoff at 10% difference because I did not want to leave any doubts. OPS is, like every individual statistic, narrow and flawed. In 2000, Manny Ramirez had a 1.154 OPS. Jason Giambi had a 1.123 OPS. I don’t think you could say with any confidence that MannyBManny was a better hitter that year than Giambi — in fact, when you look at various other variables, I think it’s pretty clear that Giambi was the better hitter. There are a few other cases like that, where the difference was too small, so I threw them out.*

*In most of those cases, I should add, I think the OPS leader WAS a better hitter than the MVP, but I don’t want this to devolve into an argument whether Darryl Strawberry or Kirk Gibson had the better 1988 season.

And now we are left with 44 MVPs whose OPS were at least 10% lower — meaning (in almost every case) that they were CLEARLY not as good a hitter. Still, they won the MVP. Why? Yes, I looked at ALL 44 of them. That’s why this post is so insanely long.

In the end, I think there are three dominant reasons why the MVP was not the best hitter.

1. The voters thought that the MVP provided more defensive value.

— This could mean that the MVP was widely viewed as a better defender or (more often) that the MVP played a much more valuable defensive position.

2. The MVP played on a winning team while the best hitter did not.

3. The voters decided that the MVP provided various intangibles (leadership, clutch hits, great base running, etc.) that the best hitter did not.

There are other smaller reasons — RBIs sometimes play a role, batting average played a role (especially in the early years), and voter fatigue (a player having won the award already) may have played a part too. Also, sometimes, the voters just fall in love with a good story.

Let’s see what we found:

1931: Frankie Frisch over Chuck Klein

Difference in OPS: 22.2%

Reasons: 1, 2, 3.

Comments: Rogers Hornsby technically led the NL in OPS that year but he only played in 100 games … Frisch had what I call the MVP trifecta. He was (correctly) viewed as much more valuable defensively than Klein, his Cardinals won the pennant, and he was probably more lauded for intangibles than any player of his era. Fordham Flash, they called him. These things made up for a huge difference in offense. They voters may have been right in this case though not for the right reasons — Klein’s high OPS was largely an illusion of the hitter-crazy Baker Bowl where he played. Frisch really was better. But Bill Terry was almost certainly the best player in the National League that year.

1934: Mickey Cochrane over Lou Gehrig

Difference in OPS: 28.3%

Reasons: 1, 2, 3.

Comments: Another MVP trifecta. This is one of the most egregious choices in the award’s history — Gehrig won the Triple Crown in 1934, for crying out loud. But the stars were aligned for Cochrane. 1. He was a catcher — and catchers get Papal Dispensation in MVP voting because it’s such a strenuous position; 2. Cochrane’s Tigers won the pennant. 3. Cochrane actually served as the Tigers manager as well as catcher, so that meant intangibles galore. Gehrig never had a chance — he actually finished fifth in the voting.

1935: Gabby Hartnett over Arky Vaughan

Difference in OPS: 13.6%

Reasons: 1, 2, 3.

Comments: Give another one to the catcher … Vaughan’s 1935 season is one of the best ever for a shortstop. He hit .385/.491/.607. But catcher trumps shortstop. Winner (Hartnett’s Cubs) trumps also-ran (Vaughan’s Pirates). And Gabby as a nickname trumps Arky. To be fair, Hartnett was a marvelous defensive catcher and he did hit .344 — I don’t want to get away from the point that the MVP winner often had a great season — but Vaughan was a lot better offensively and did not get the award.

1937: Charlie Gehringer over Lou Gehrig

Difference in OPS: 12.4%

Reasons: 1

Comments: Here is our first really baffling one. Gehrig’s team won. Gehrig, at this point in his life, certainly had any advantages in the sportswriters’ minds as far as intangibles go. Gehringer did win the batting title, which carried a lot of weight in those days, and he had never won the MVP award. It’s also possible Gehrig split votes with his young teammate Joe DiMaggio. I think the writers were probably able to convince themselves that Gehringer’s defense at second base made up any difference there might be in hitting.

1938: Ernie Lombardi over Johnny Mize

Difference in OPS: 11.7%

Reasons: 1, 3

Comments: Well, we might need to add a fourth category especially for the early years — batting title. Lombardi gets the catcher advantage. But I’d say the fact he won the batting title — and won it on the last day after refusing to sit, a la Ted Williams in ’41 — was probably the difference maker here.

1940: Frank McCormick over Johnny Mize

Difference in OPS: 18.3%

Reasons: 1, 2

Comments: Poor Johnny Mize. He had another one taken away from him, this time by a slick-fielding first baseman named Frank McCormick, whose Reds won the pennant.

1941: Joe DiMaggio over Ted Williams

Difference in OPS: 15.9%

Reasons: 1, 2, 3

Comments: See DiMaggio over Williams in 1947.

1942: Joe Gordon over Ted Williams

Difference in OPS: 21.5%

Reasons: 1, 2, 3

Comments: One of the most famous slights in MVP voting history … Gordon really did have an excellent season. Williams’ season was clearly better offensively, but the voters, as they often will, went with defense, winning and leadership over performance.

But maybe I should pause here to make a point again: I think there is a difference between giving an award to an UNDESERVING player and giving an award to, perhaps, a LESS DESERVING player.

Take the National League MVP award this year. I thought Matt Kemp was most deserving of the MVP award. I would have voted for him unquestionably. I thought when you considered park factors, he was the best offensive player in all of baseball. And while there was some difference of opinion about his defense, he did rate well in some defensive statistics and he won a Gold Glove. I thought he was the guy, absolutely.

But, Ryan Braun ALSO had an MVP season. It may not have been as good, all things considered, as Kemp’s. But it was certainly MVP quality. The fact that he won the MVP (obviously before we knew of the possibility of a failed drug test) did not thrill me — I think in the end he won it because he had better teammates — but I don’t think it’s one of the all-time bad decisions because Braun was really good in 2011.

And the same goes for Joe Gordon in 1942. He was not as good as Ted Williams. He wasn’t really CLOSE to as good as Ted Williams. But he was really good. His MVP season stands up well against other MVP seasons.

1944: Marty Marion over Stan Musial

Difference in OPS: 30.7%

Reasons: 1, maybe 3

Comments: This one doesn’t stand up well. It’s one of the worst decisions in MVP history, but, hey, there was a war going on. Marion was a dreadful offensive player — even that season — and Musial was, of course, Musial. But Marion was an all-time defensive shortstop, and that carried the day.

1947: Joe DiMaggio over Ted Williams

Difference in OPS: 19.4%

Reasons: 1, 2, 3

Comments: Bill James made a fascinating point about the famous DiMaggio-Williams MVP races. He says that in 1941, there was an argument to be made about which one was better. Williams had the better offensive year, but when you took into account park factors, defensive contribution, the hitting streak and so on, there was at least a good argument to be made for DiMaggio.

But in 1947, there really wasn’t even an argument. Williams’ offensive numbers outclass DiMaggio’s in every way. He hit 30 points higher, slugged 100 points higher, scored 27 more runs, drove in 17 more RBIs, there was no argument, But, Bill says — and I think it’s a great point — once voters decided that DiMaggio’s defense and intangibles eclipsed Williams’ hitting, they lost their sense of perspective. In 1941, defense and whatever else voters believed about DiMaggio things MIGHT have made up the difference (I don’t think they did, but that’s an argument). In 1947, when Ted Williams won the Triple Crown, there was a much, much bigger gap and those things almost certainly DID NOT make up the difference. But the voters chose DiMag anyway.

1947: Bob Elliott over Ralph Kiner

Difference in OPS: 12.1%

Reasons: 1, 3

Comments: They called Bob Elliott, “Mr. Team.” They did not call Ralph Kiner that.

1948: Lou Boudreau over Ted Williams

Difference in OPS: 11.2%

Reasons: 1, 2, 3

Comments: Well, here’s a case where it’s not entirely clear that the OPS leader was the better hitter. Boudreau was second in offensive WAR to Williams by only .3 points. Williams had the advantage of a better hitters park so the numbers are not entirely clear. Boudreau actually had the highest WAR, though Williams has the advantage in Win Shares. In any case, Boudreau’s Indians won, he was considered a defensive whiz (and inventor of the Boudreau Shift specifically to thwart Williams) and he was player/manager. I suspect that’s why he won.

1949: Jackie Robinson over Ralph Kiner

Difference in OPS: 11.8%

Reasons: 1, 2, 3

Comments: Ralph Kiner was a defensive liability who played on lousy teams and was widely viewed as a terrible team player (largely because Branch Rickey hated him and was never shy about saying so). With those three things going against him, he had absolutely no chance to win an MVP award no matter how well he hit — and despite being a truly great hitter for seven years in a row, he received exactly one first place MVP vote.

1950: Phil Rizzuto over Larry Doby

Difference in OPS: 13.1%

Reasons: 1, 2.

Comments: One thing I would love to see explored more — maybe I should do it myself — is the difference in integration between the National and the American League. We’ve heard the National League story again and again (and it is a fascinating story) but the AL story is fascinating in its own right because the two most prominent teams in the league — the Yankees and Red Sox — were so reluctant to use black players.

And so you see one picture when you look at the 1949-59 MVPs in the National League: 1949: Jackie Robinson; 1950: Jim Konstanty; 1951: Roy Campanella; 1952: Hank Sauer; 1953: Roy Campanella; 1954: Willie Mays; 1955: Roy Campanella; 1956: Don Newcombe; 1957: Hank Aaron; 1958: Ernie Banks; 1959: Ernie Banks.

That’s nine African American MVPs in 11 years.

The American League MVPs over that same time: 1949: Ted Williams; 1950: Phil Rizzuto; 1951: Yogi Berra; 1952: Bobby Shantz; 1953: Al Rosen; 1954: Yogi Berra; 1955: Yogi Berra; 1956: Mickey Mantle; 1957: Mickey Mantle; 1958: Jackie Jensen; 1959: Nellie Fox.

Right: That would be zero. The first African American MVP in the American League was Elston Howard in 1963.

That’s not to say that there was racism in the AL voting. No, there was racism in American League MANAGEMENT. The Washington Senators did not have a player of color until 1954. They Yankees waited until 1955. The Tigers — largely because of the racism of Detroit owners Walter Briggs Sr. and Jr. — did not have a black player until 1958. And the Red Sox, rather famously, were the last time to integrate in 1959. That’s half the American League right there. There were a few excellent players of color in the American League in the 1950s — Larry Doby, Minnie Minoso, Al Smith, Elston Howard, Vic Power — but not NEARLY as many as in the National League.

Doby had a pretty good case for MVP in 1950, 51 and ’52 and he finished second in the voting in 1954. Minnie Minoso had a good case in 1954 too, and he almost certainly should have done better in the voting in 1959.

1951: Yogi Berra over Ted Williams

Difference in OPS: 17.4%

Reasons: 1, 2, 3

1954: Yogi Berra over Ted Williams

Difference in OPS: 25.5%

Reasons: 1, 3

Comment: Once you have determined that MVP voting is strongly biased toward (1) key defensive players, (2) on winning teams, (3) who were lauded for their clutch hitting and leadership, it’s easy to see why Yogi Berra won three MVP awards over players who had much better offensive seasons. Yogi, in many ways, was the perfect MVP candidate for his time.

And Ted Williams, in many ways like Ralph Kiner, was the Bizarro MVP candidate of his time. He, of course, played left field and with particular disinterest, he wasn’t often on winning teams, and sportswriters never tired of calling him a loser and a terrible teammate, even if the charge was often baseless.

In other words when Yogi and Williams, clashed, Williams would never win.

1955: Yogi Berra over Mickey Mantle

Difference in OPS: 21.4%

Reasons: 1, 3

Comments: Mantle was much, much better than Yogi in 1955, but the Yogi MVP magic was still burning.

1958: Jackie Jensen over Mickey Mantle

Difference in OPS: 10%

Reasons: Mantle Fatigue?, RBIs (late add)

Comments: Ted Williams actually had the highest OPS in 1958, but he only played in 121 games so I’m going with Mantle … Mantle and Rocky Colavito both had markedly better offensive seasons than Jensen in 1958. Jensen did win the Gold Glove in 1959, so there might have been a sense he was a great outfielder though I don’t see too much evidence in stories of the time that he had a great defensive reputation — plus Mantle played center field. I’m guessing that because Mantle had won the MVP award in 1956 and ’57, the voters just wanted someone different. And … wait … I just noticed something. Jensen did lead the American League in RBIs in 1958. So, we might be entering the RBI zone.

1959: Nellie Fox over Al Kaline

Difference in OPS: 18.2%

Reasons: 1, 2, 3

Comments: Kaline played in 20 fewer games than Fox, so that’s a consideration here too. But mostly, it was the White Sox year and the top three players in the MVP were all White Sox — Fox, Luis Aparicio and Early Wynn. Minnie Minoso had about as good a year as Fox, but finished 12th in the voting.

1960: Dick Groat over Frank Robinson

Difference in OPS: 23.7%

Reasons: 1, 2, 3

Comments: Groat was not viewed as a good-fielding shortstop, but he was a shortstop, and the Pirates won, and Groat was seen as a leader on that team. This year was controversial, but not because of Robinson — this was the year that Roberto Clemente fans (and Clemente himself) thought he should be the MVP since he was (obviously) a great right fielder, and he led the Pirates with 94 RBIs. Anyway, it doesn’t matter: Willie Mays had a far superior season to any of them. Mays probably should have won at least five more MVPs when it comes down to it.

1961: Roger Maris over Norm Cash

Difference in OPS: 13.6%

Reasons: 1, 2, 3 (Home Run Record)

Comments: Maris was viewed as a significantly better fielder than Cash, and the Yankees won, but I don’t think this one applies to our theory of “the best hitter” not getting the award. I suspect nobody in 1961 (and few people now) would have accepted that Cash had better offensive year than Maris who, after all, had just broken Babe Ruth’s home run record. Cash’s 115 point advantage in on-base percentage and 40-point advantage in slugging would have been scoffed at. Mantle probably had a better year than both of them, all things considered.

1962: Maury Wills over Frank Robinson

Difference in OPS: 31.1%

Reasons: 1, 3, Story

Comments: One of the most famous miscalls in MVP voting history, but again it wasn’t Robinson who was jobbed. It was Willie Mays, who was better than Willis just about every way imaginable AND his team won the pennant.

I would argue that the voters fell in love with Wills because he had just finished of one of the most jolting seasons in baseball history — nobody in the NL had stolen more than 50 bases since Max Carey in the early 1920s, and then he stole ONE HUNDRED AND FOUR. It was one of those jaw dropping achievements that inflamed the imagination. It was, in retrospect, a fabulous but almost certainly overrated season … still, people who witnessed it never forgot, and there was a concerted effort (by the great Jim Murray, among many others) to get Wills into the Hall of Fame, even though he had a lifetime 88 OPS+ and was probably not a good defensive shortstop (despite two Gold Gloves). That’s how deeply the image of Wills stealing 104 bases seared in the mind of people who watched it.

As for Frank Robinson, he led the National League in OPS three straight year and got one MVP out of it.

1964: Ken Boyer over Willie Mays

Difference in OPS: 13.7%

Reasons: 1, 2

Comment: It’s probably absurd to suggest that anyone could have won an MVP over Willie Mays with DEFENSE as a bias, but I think that might have been the case in 1964. I mean, the big difference here is that the Cardinals won the pennant with that famous comeback over the sad ol’ ’64 Phillies. But Boyer was also a brilliant defensive third baseman, or he had been in his younger days, and while Mays was probably the best defensive center fielder ever in the Major Leagues, he was getting older and I can’t help but wonder if people weren’t taking that part of his game for granted. I don’t know — maybe not. It probably just comes down to the Cardinals winning.

1964: Brooks Robinson over Mickey Mantle

Difference in OPS: 12.4%

Reasons: 1

Comment: Mantle’s year was better offensively, but the voters determined that Robinson’s defense more than made up the difference. In this case, WAR agrees with the voters. Win Shares gives a slight edge to Mantle.

1965: Zoilo Versalles Over Carl Yastrzemski

Difference in OPS: 16.2%

Reasons: 1, 2, 3

Comment: One of the most famous “defense first” choices; Versalles did offer some offensive punch. He led the league in doubles and triples and and he hit 19 homers. His 115 OPS+ certainly wasn’t bad. he just had a very low on-base percentage. The real argument that year did not involve Yaz; it was whether to go with Versalles defense or his teammate Tony Oliva’s offense. The vote wasn’t close, 19 first place votes for Versalles, one for Oliva.

1966: Roberto Clemente over Dick Allen

Difference in OPS: 12.8%

Reasons: 1, 3, Story

Comment: Clemente was just such a joy to watch play … I think that was his intangible. Top two in WAR that year were both pitchers — Koufax and Marichal. Mays had another fabulous year. Allen led the league in slugging and OPS. But Clemente was just so infectious, such a wonder, he did so many things well, I think that won him the award (see Ichiro, 2001).

1970: Johnny Bench over Willie McCovey

Difference in OPS: 11.7%

Reasons: 1, 2

Comment: McCovey probably did have a better offensive year than Bench in 1970, but most of that difference was tied up in walks (McCovey walked 83 more times than Bench) and as Bill James points out, many of the MVP voters DESPISE walks. They don’t feel neutral about walks. No, they despise walks — walks show a lack of gumption or something. What’s interesting is that in 1970, Tony Perez had a decidedly better offensive year than Bench — higher average, much higher on-base percentage, slightly higher slugging percentage. But Bench, of course, was a miraculous defensive catcher. Miraculous was really the word they used then. Nobody had seen a catcher quite like him — with that arm and that athleticism and all of it.

1973: Pete Rose over Willie Stargell

Difference in OPS: 19.3%

Reasons: 1, 2, 3

Comment: Rose was such a force of nature that people seemed perfectly willing to overlook his sudden, and rather shocking, decline in power. From 1965-71 — a time for pitchers — Rose hit double figure homers every year and slugged a rather robust .460. Starting in 1972, though, he became a different hitter, a slasher who cut down on his strikeouts, hit a lot of ground balls and line drives. He led the league in doubles five times, but only once more hit double figure homers (and only 10 at that). In 1973, he led the league in hitting (.338) and punched out 230 hits. It was, unquestionably, a terrific year. But it was limited too. It seems almost certain that Joe Morgan, who hit 21 more homers, stole 57 more bases, walked 45 more times , had a better year despite hitting only .290. But nobody was going to see that in 1973 — Joe Morgan the announcer wouldn’t have voted for Joe Morgan the player (“Wait, he hit only .290?”).

1974: Steve Garvey over Willie Stargell

Difference in OPS: 14.1%

Reasons: 1, 3

Comment: Rose and Garvey’s brand of offense (.312 average, no walks, lots and lots of hits) was the style of choice in the early 1970s. Garvey did hit 21 homers, but his .342 on-base percentage makes him a pretty bad choice for MVP, especially because there was a great choice in 1974: Mike Schmidt. He led the league in slugging, led the league in homers, walked 95 times and played Brooks Robinson defense at third base. Schmidt was the best player in the the league. Nobody saw it just yet.

As for Stargell, he was a great hitter in the early 1970s … but it isn’t until he developed the reputation as a leader in the late 1970s that he won his MVP award over better hitters.

1976: Thurman Munson over Hal McRae

Difference in OPS: 11.4%

Reasons: 1.

Comment: A catcher vs. a designated hitter? No contest: McRae didn’t get a single first place vote. George Brett was probably the best choice that year, though the most fitting choice might have been Mark Fidrych, who electrified baseball that year.

1979: Don Baylor over Fred Lynn

Difference in OPS: 14.9%

Reasons: 2, 3

Comment: One interesting MVP voting question to me is this: When do you give credit to a player for being a “winner?” In 1979, Baylor’s Angels won the American League West while Lynn’s Red Sox finished a distant third in the East. So, that would seem to give Baylor the voter’s edge — and indeed, I think it did. But Lynn’s Red Sox actually won MORE GAMES than Baylor’s Angels. Baylor, of course, got loads of credit for being a leader, and 1979 was the year for that; this was also the year that Pops Stargell won his MVP.

1985: Willie McGee over Pedro Guerrero

Difference in OPS: 11.2%

Reasons: 1, Batting Average.

Comment: McGee hit .353 to Guerrero’s .320, which I suspect evened up the score for most voters, though Guerrero’s on-base percentage was almost 40 points higher and he hit 23 more home runs. But I think the voters got it right, at least between these two, because Guerrero missed 25 games. There seems almost no doubt at all that Dwight Gooden was the best player in the league anyway, but his team didn’t win.

1987: Andre Dawson over Jack Clark

Difference in OPS: 15.1%

Reasons: 1, 3, RBIs, Story

Comment: I would say this is the most bizarre choice in MVP history … but I also think I understand it. It’s bizarre because Dawson did not even have a GOOD year, much less a great one. By WAR, it was his ninth best season. HIs .328 on-base percentage is the lowest to ever win the award in the National League. It’s bizarre because the Cubs were TERRIBLE, last place, and we all know how much winning plays a role in the MVP voting. It’s bizarre because Dawson was no longer playing a prominent defensive position — he had not been a center fielder fore three years — and even though he was still viewed as a good defensive outfielder, everyone also knew that his legs were shot. It’s just bizarre

So why did it happen? Well, there’s the obvious — he led the league in RBIs (and homers) which tends to send many MVP voters into the “There’s gold in them thar hills” dance. But with Dawson, I think there was something more. There was a compelling story. Dawson was one of the classiest players in the game. He was badly mistreated by owners in the shameful collusion scandal.* He showed up at Cubs spring training with a signed blank contract, which was at first called a ‘Dog and pony show” and then Dawson signed for a substantial pay cut. He promptly went out and hit 49 homers and drove in 137 runs which, even considering the context, are jaw-dropping power numbers. It was a triumphant ending, a Disney ending — longtime Cubs fans STILL get teary-eyed thinking of that season — and the MVP vote was simply the triumphant music. Sure, Ozzie Smith had a better year. Tony Gwynn had a better year. Eric Davis … Tim Raines … Dale Murphy … the aforementioned Jack Clark … numerous others had better years. But none of them had a better story.

*I was reminded by a couple of Brilliant Readers just what Tim Raines lost in his career because of the Labor Wars. In 1981 — Raines rookie season — he stole 71 bases in 88 games. I’ll repeat that: He stole 71 bases in 88 games. Over 150 games, that would equal 121 stolen bases — he was on pace to set the Major League record in stolen bases. He had that chance taken away by the strike. He might have won the MVP award had that season been played out. … In 1987, he was a victim of collusion and missed the first month of the season. Even with that, he hit .330, banged a career high 18 homers, stole 50 bases while being caught only five times and led the league in runs scored. Another month, it might have been impossible to deny him the MVP. … And then, in 1994 and 1995, when he was still an every day player, he lost maybe 50 games because of the labor nonsense.

Raines finished his career 395 hits short of 3,000, and I’m not sure that the lost games cost him quite that many hits, but he would have been quite a bit closer. More, though, he might have won two MVPs. I think without labor woes, Tim Raines would already be in the Hall of Fame.

1995: Mo Vaughn over Edgar Martinez

Difference in OPS: 13%

Reasons: 1, 3, RBIs

Comments: Vaughn wasn’t exactly a defensive whiz, but with Edgar being a full-time DH by 1995, you would have to say that defense played a small role. There were the intangibles — Vaughn had many big hits. And with the MVP voters often being RBI-happy, Vaughn had more RBIs. It’s not a sensible stance — Edgar had a much better offensive year than Mo Vaughn. But this is how MVP voting goes.

The more surprising turn in 1995 was Mo Vaughn winning over Albert Belle because Vaughn doesn’t seem to have ANY of those historical advantages over him. Defense: Draw. Both teams were winners — Belle’s Indians were actually a lot better. They had the same number of RBIs. But Belle hit 17 points higher, slugged more than 100 points higher, hit 50 homers and 50 doubles, and scored 23 more runs. The value really wasn’t close. But nobody liked Albert Belle and everybody liked Mo Vaughn, and I think it came down to that.

1995: Barry Larkin over Barry Bonds

Difference in OPS: 12.2%

Reasons: 1, 2, 3

Comments: Larkin needed all three biases — defensive value, a winning team and the intangibles of leadership and clutch ability — to win the award, though Bonds wasn’t a factor in the voting. He finished 12th.

1996: Juan Gonzalez over Mark McGwire

Difference in OPS: 15.6%

Reasons: 2, RBIs

Comments: A pretty dreadful choice, though it was probably Alex Rodriguez — with one of the greatest shortstop seasons in baseball history — who probably deserved the MVP. McGwire did have almost 100 points of on-base percentage, and almost 90 point of slugging over Juan Gone despite playing in a much tougher hitting park (he led in OPS 196-145). And Gonzalez did have 144 RBIs, which was a lot more than A-Rod or McGwire. But if we talk about RBIs, we once again have to turn to Albert Belle who had MORE RBIs for a better team than Juan Gone and also 40-plus points in on-base percentage.

The two Juan Gone MVPs are among the most baffling in baseball history.

1998: Sammy Sosa over Mark McGwire

Difference in OPS: 16.2%

Reasons: 1, 2, 3, RBIs

Comments: Whatever you may think of Mark McGwire and how he got strong enough to hit all those home runs, nobody could deny that he performed under the most intense glare imaginable. Heck, stadiums were filling up just to watch him take batting practice. And then he went out and hit .299/.470/.752. He absolutely DWARFED Sosa’s offensive numbers. Using runs created, McGwire created FIFTY more runs than Sosa. There’s no way that Sosa made that up in defense or intangibles. But, hey, the Cubs won, and Sosa had more RBIs, and he didn’t just win the MVP, he won it almost unanimously (30 first place votes to 2).

1999: Ivan Rodriguez over Manny Ramirez

Difference in OPS: 17.3%

Reasons: 1, 3.

Comments: Pudge got the special catcher’s dispensation and edged Pedro in one of the weirder votes in baseball history (Pedro had more first place votes but, famously, was left entirely off two ballots) … A word for for MannyBManny, though. The Cleveland Indians from 1994 to 2002 had some of the most amazing offensive lineups in baseball history. They reached two World Series. They led the league in runs scored three times and finished second another three. They had players lead the league in runs (Albert Belle, Robbie Alomar), hits (Kenny Lofton), doubles (Belle), triples (Kenny Lofton), homers (Belle), RBIs (MannyBManny, Albert Belle), OPS (MannyBManny, Thome), stolen bases (Lofton), slugging percentage (Belle, MannyBManny, Thome) and walks (Thome). They did not have a single player win the MVP award.

I don’t like RBIs as a defining statistics, but we know the MVP voters do. In 1999, Manny Ramirez drove in ONE HUNDRED SIXTY FIVE. That was the most in baseball since Jimmie Foxx. Jimmie Bleepin’ Foxx. He finished tied for THIRD in the MVP balloting. Whew. What does an enigma have to do around here to get some respect?

2000: Jeff Kent over Todd Helton

Difference in OPS: 12.2%

Reasons: 1, 2, 3.

Comments: I included No. 3 here because the intangible was Coors Field. Helton’s raw numbers were other-worldly, but so it goes when you hit in a balloon factory. Helton was a great hitter wherever he went, but at Coors Field he hit .391 and slugged .758. Even with that you could make the argument that Helton was the better choice, but with Kent playing a more significant position (though not well), with the Giants winning and with Coors Field muddying up the whole thing, Kent got the nod over his teammate. He also got the nod over his longtime friend Barry Bonds.*

*Looking for the irony font.

2001: Ichiro Suzuki over Jason Giambi

Difference in OPS: 26.3%

Reasons: 1, 2, 3

Comments: Giambi’s A’s won 102 games in 2001, which would seem to blunt the winning bias. But the Mariners famously won 116. … Suzuki definitely had all of the biases going for him — he was a great defender, his Mariners won like crazy, people talked about his intangibles all the time — but I really think he won the MVP for about the same reason that Maury Wills won the 1962 MVP: Nobody had seen anything quite like him. He came from Japan, and he slapped and chopped and blooped and slashed 242 hits. He stole 56 bases. He threw like Clemente. And, like Clemente, he was irresistible. Nobody in 1962 really knew how to measure the value of 104 stolen bases, and so I think people probably overvalued. Nobody in 2001 really knew how to measure the value of Ichiro and so, again, I think people probably overvalued. Oh, don’t’ get me wrong, I love Ichiro, he was wonderful, he is wonderful, he’s one of my favorite all-time players, and that 2001 season was certainly MVP worthy.

But I think Giambi, with a .447 on-base percentage and 200 point higher slugging percentage, was more valuable no matter how well Ichiro threw or ran.

2002: Miguel Tejada over Jim Thome

Difference in OPS: 23.2%

Reasons: 1, 2, 3

Comments: A shortstop vs. a not-so-hot first baseman? Check. A winning team vs. a loser? Check. A “clutch” player vs. a perceived non-clutch? Check. Tejada’s vote went right by the playbook and Thome never was even in contention. He finished seventh.

2006: Justin Morneau vs. Travis Hafner

Difference in OPS: 14.9%

Reasons: 1, 2, 3, RBIs

Comments: To be fair, Hafner only played 129 games, but this Indians bias has to stop. I really thought Joe Mauer was the MVP in 2006. And I thought Derek Jeter had a better year than Morneau. But the MVP voters have always loved Morneau. Best I can tell, Morneau does indeed seem like a lovable guy.

2007: Jimmy Rollins over Chipper Jones

Difference in OPS: 15%

Reasons: 1, 2, 3.

Comments: Another by the book choice — Rollins played the more important defensive position, his Phillies won and Rollins was said to be a team leader. Rollins’ 2007 season was indeed remarkable — 20 doubles (38, actually), 20 triples (exactly), 20 homers (30), 20 stolen bases (41), a Gold Glove at shortstop, a brilliant season. Chipper with 150 more OPS points was probably more valuable, though, and so was Albert Pujols.

2008: Dustin Pedroia over Milton Bradley

Difference in OPS: 13%

Reasons: 1, 2, 3, 4

Comments: Bradley only played 126 games, plus he’s Milton Bradley … this one isn’t a viable comparison.

And there you go — I am certain that nobody got to this point in the article, but if you did, I think we can take away this: It is rare that the best hitter in the league wins the MVP award BECAUSE he’s the best hitter in the league. It might have happened a few times — Larry Walker in 1997, Joe Torre in 1971, Willie McCovey in 1969, Bagwell and Thomas in the strike year of 1994 and so on — but in general to be MVP you need to play a key defensive position, have his team win, be imbued with great and potentially undetectable qualities.

RBIs help too.

2 Responses to The MVP Formula

  1. For Jim Thome to be snubbed in 2002 is as big an injustice as any snub in the game.

  2. Kurt says:

    I think you are dead wrong about the 1995 AL MVP vote. Albert Belle absolutely should have been the winner, not Mo Vaughn or Edgar Martinez. Belle was by far the best player than year and everyone knew it. Belle hit 50 home runs and 50 doubles that year when it was rare to hit 50 home runs. Mo Vaughn only won because Belle was a jerk and the sportswriters hated him. I refuse to believe that Edgar Martinez, as a DH, could possibly compare to Albert Belle in 1995.

    I personally think that if Belle hadn’t developed his hip problem, he would have been widely considered as a Hall of Famer, although I strongly suspect he was probably a juicer.

Leave a Reply

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