MLB MVP Award: BBWAA voting converges with WAR
The MLB MVP choices by the BBWAA voters have aligned more closely with the leaders in Fangraphs WAR over the last 20 years.
Congratulations to Jose Altuve and Giancarlo Stanton for winning the MVPs of their respective leagues. They were both very deserving MLB MVPs who put up great numbers. They also happen to be the shortest and tallest MVPs ever, which is one of the very cool things about baseball (Altuve is tied with Phil Rizzuto and Bobby Shantz for shortest MVP). In baseball, you can have small guys like Altuve and Dustin Pedroia compete equally with big guys like Stanton and Aaron Judge, not to mention the rare Bartolo Colon types.
The BBWAA first awarded MVP trophies to the best player from each league in 1931. There were MVP-like awards prior to that, but they were inconsistently awarded. The 1931 AL MVP was Lefty Grove, and the NL MVP was Frankie Frisch. Grove and Frisch were fine players, but not necessarily deserving MVPs. According to Fangraphs Wins Above Replacement (fWAR), there were three players in the AL more valuable than Grove and 16 players in the NL more valuable than Frisch. The fWAR MVPs would have been Babe Ruth and Bill Terry.
Of course, fWAR isn’t gospel. There are other metrics like Baseball-Reference WAR and Baseball Prospectus WARP. They all measure the value of a player in one number by incorporating all the things that player does on a baseball diamond. They value a player based on wins above a theoretical replacement player.
When it comes to MVP voting, the baseball writers can use whatever method they choose to determine their picks. Many years ago, their primary assessment tool was their eyeballs and traditional statistics. Most of the time the MVP was on a winning team and very often on a playoff team. Many writers still give extra credit to players on winning teams and playoff teams, but it’s not an absolute. Stanton won the NL MVP this year despite playing on a losing team. Mike Trout did the same last season.
When the MVP Award came into existence, it was not unusual for voters to choose players who were not the most valuable according to fWAR. For one thing, there was no fWAR. For many decades, the narrative of the season greatly influenced the MVP voting. It was more based on subjective feelings than objective measurements. This is still the case for some voters, but objective measurements are being used more and more.
This isn’t to say the MVP should be based on one number like WAR and it doesn’t have to go to the league leader in WAR, but it should go to someone who is at least in the ballpark. Based on my research, BBWAA voters are regularly voting for players who are among the league leaders in Fangraphs WAR (it should be noted that fWAR was chosen because Fangraphs’ Combined WAR Leaderboard is easily accessible).
MVPs and Fangraphs WAR
This year is a good example of the progress by voters. American League MVP Jose Altuve ranked 2nd in fWAR to Aaron Judge (8.2 to 7.5). National League MVP Giancarlo Stanton finished in a tie with Anthony Rendon at the top of the fWAR standings (Rendon was incrementally ahead but both are listed at 6.9 fWAR). The voting was very much in line with the fWAR rankings.
This year’s BBWAA picks are typical of the voting trends over the last five years. In the AL, all of the last five MVPs have finished in the top three in fWAR. In the National League, the last ten MVPs have finished in the top three in fWAR and nine of them were the league leader in the metric. The lone exception in the last 10 years is Ryan Braun in 2011, who was third in fWAR behind Matt Kemp and Roy Halladay.
It wasn’t always like this. Looking back to 1931, here’s the breakdown of how often the BBWAA MVP was the league leader in fWAR.
29%–1931 to 1989
35%–1990 to 1999
45%–2000 to 2009
63%–2010 to 2017
For nearly sixty years, the MVP was the league leader in fWAR just under 30 percent of the time. This started to increase slightly in the 1990s and again in the 2000s before really jumping in the last eight years (particularly in the NL).
There’s also been an increase in how often the MVP was among the top three in fWAR. This has been particularly true over the last eight years, when 75 percent of the AL MVPs and 100 percent of the NL MVPs were in the top three in fWAR.
48%–1931 to 1989
45%–1990 to 1999
50%–2000 to 2009
88%–2010 to 2017
This increase in how often the MVP finishes at or near the top of the fWAR rankings has also meant really bad choices are increasingly rare. As mentioned above, the NL MVPs have been very good choices over the last 10 years. Before this, the back-to-back picks of Ryan Howard in 2006 and Jimmy Rollins in 2007 were two of the worst picks in the last 20 years.
Bad Choices by the BBWAA
In 2007, Jimmy Rollins finished eighth in fWAR and was almost two wins worse than David Wright, but still won the NL MVP. The Phillies made the playoffs by a single game over the Mets and Rollins led the league in games, plate appearances, at-bats, runs, and triples. He also hit 30 homers and stole 41 bases. Still, Wright was the more valuable player based on fWAR.
In 2006, Ryan Howard won the MVP despite finishing ninth in fWAR. He led the league with 58 homers and 149 RBI, but was still not as good as Albert Pujols. This was a slightly unusual win because Howard’s Phillies didn’t make the playoffs while Pujols’ Cardinals did. Many times, that would have been enough to put Pujols over the top. In this case, Howard’s big home run and RBI totals likely made the difference for the voters.
Despite those two seasons with questionable picks, the NL has a better recent history than the AL. When Miguel Cabrera won back-to-back MVPs in 2012 and 2013, he was 3.9 fWAR worse than Trout the first year and 3.0 fWAR worse the second. Trout also finished ahead of Josh Donaldson in 2015 but “The Bringer of Rain” took the trophy.
Before this season, Trout led the AL in fWAR five straight years and has just two MVP Awards to show for it. Trout could easily have five MVP trophies in his wood-paneled den where he likes to track the weather. If he hadn’t missed one-fourth of this season, he likely would have led the AL in fWAR for a sixth straight year.
The AL had a couple of really awful picks in the early 2000s. Justin Morneau was 38th in fWAR in 2006 when Grady Sizemore was the league leader and outplayed him by 4.1 wins. Sizemore led the league again in 2008, but Dustin Pedroia took the MVP over him despite finishing fourth in fWAR.
In 2002, it was Miguel Teajada (27th in fWAR) who stole one from Alex Rodriguez, just as Jason Giambi (5th in fWAR) had done two years earlier. The Tejada over A-Rod pick was particularly bad because A-Rod was worth more than double the fWAR as Tejada, beating him 10.0 to 4.5. Tejada played on the Moneyball Oakland A’s, who won 103 games on a low-budget payroll, so he had a strong narrative.
The two worst picks in the last 40 years were also in the American League. For some reason, the AL seems to have had more really bad MVP picks than the NL during this time. In 1996, Juan Gonzalez was 52nd in the AL in fWAR, yet was named MVP. It should have been Ken Griffey, Jr. (or even Alex Rodriguez).
Juan Gone had a great year at the dish but was terrible in the field. He also played in the second-best park for hitters, which helped his numbers. He was worth 3.5 fWAR. Griffey (and A-Rod) were both very good at the plate and in the field and both were worth more than 9 fWAR.
The Juan Gonzalez MVP in 1996 was the second-worst choice over the last 40 years. Number one on the list is Willie Hernandez in 1984. Hernandez was the Tigers closer who won the Cy Young and MVP Award by pitching 140.3 innings with a 1.92 ERA. He was worth 3.2 fWAR. Shortstop Cal Ripken, Jr. hit .304/.374/.510, with 103 runs, 27 homers, and 86 RBI in 162 games (of course). He was worth 9.8 fWAR.
Players “Robbed” of Multiple MVPs
As mentioned above, Grady Sizemore and Alex Rodriguez were both “robbed” of MVP trophies two times, but they are hardly alone. There’s an impressive list of players who have led the league in fWAR but failed to win the MVP on multiple occasions. Including Sizemore and Rodriguez, there are 17 such players, many of them Hall of Famers. Carl Yastrzemski, Ken Griffey, Jr., Lou Gehrig, and Jackie Robinson are on this list.
Four players have had this happen three times. Trout was mentioned above. He’s joined by Mickey Mantle, who won back-to-back MVP trophies in 1956 and 1957 and led the AL in fWAR both years. He also led the AL in 1955 and 1958, but lost the MVP race to Yogi Berra and Jackie Jensen. He could have won four straight. In 1961, for the third time in his career, Mantle led the league in fWAR but didn’t win the MVP. His teammate, Roger Maris, won the award with his record-breaking 61-homer season.
Randy Johnson led his league in fWAR three times and didn’t win the MVP in any of those three seasons. In 1995, Mo Vaughn took the honors. Chipper Jones stole it from him in 1999 and Jeff Kent won it over the Big Unit in 2000. A decade earlier, Rickey Henderson was the top player in 1981, 1985, and 1989, but Rollie Fingers, Don Mattingly, and Robin Yount won the MVPs in those seasons.
Two players, Roger Clemens and Mel Ott, have been “robbed” of the MVP four times. Clemens spread his non-wins out in the 11 years from 1988 to 1998. Jose Canseco and his 40-40 season topped Clemens in 1988. Dennis Eckersley’s 51-save season in 1992 gave him the MVP, despite ranking 60th in fWAR. Ken Griffey, Jr and his 56-homer, 147-RBI season won over the voters in 1997. Juan Gonzalez and his 45-homer, 157-RBI season beat out Clemens in 1998. Clemens is the rare pitcher to win an MVP Award, but he could have won five.
Mel Ott never won an MVP Award but finished in the top ten in the voting six times. In four seasons in the 1930s, he led the NL in fWAR but didn’t get the trophy. Chuck Klein in 1932, Dizzy Dean in 1934, Carl Hubbell in 1936, and Ernie Lombardi in 1938 all won MVPs despite finished behind Ott in fWAR during their MVP-winning seasons.
Ted Williams and fWAR
Ted Williams was one of the greatest hitters of all-time and a two-time MVP winner. He could have won seven. Including his two MVP seasons, he led the AL in fWAR seven times in the eight full seasons he played between 1941 and 1954 (he missed some years due to WWII and the Korean War). In five of those seasons, another player won the MVP.
Joe Gordon got him once. In 1942, Williams led the AL in runs, homers, RBI, walks, and all three rate categories (.356/.499/.648). He was worth 11.6 fWAR. Joe Gordon was worth 8.8 fWAR and wasn’t better than Williams in anything except for steals, yet he got the trophy. Of course, Gordon’s Yankees won 103 games to the 93 won by Williams’ Red Sox.
Joe DiMaggio got him twice. The first time was the famous 1941 season in which Ted Williams hit .406 and was worth 11 fWAR. DiMaggio (9.8 fWAR) had his 56-game hitting streak and won the MVP Award. That was somewhat understandable because DiMaggio had a great season and kept fans spellbound with his hitting streak during the season.
The 1947 season was a much worse choice. Williams hit .343/.499/.634 and led the league in runs, homers, RBI, and walks. He had 692 plate appearances. DiMaggio hit .315/.391/.522 and came to the plate almost 100 fewer times. Williams more than doubled his fWAR total, 10.5 to 4.9, but the voters went for DiMaggio.
Yogi Berra also twice won MVP awards in seasons in which Williams led the league in fWAR. In 1951, Williams led the league in on-base percentage, slugging percentage, and fWAR (7.1) but finished 13th in MVP voting. Yogi Berra finished fifth in fWAR, but took home the trophy. It was a similar story three years later when Williams again led the league in on-base percentage, slugging percentage, and fWAR, but Berra won the MVP despite being worth 2.5 fewer wins than Teddy Ballgame.
Willie Mays and fWAR
Teddy Williams may have missed out on five MVP Awards, but he isn’t the all-time leader in the category. That would be the great Willie Mays. The Say Hey Kid missed the 1953 season because of military service. He came back in 1954 and won the NL MVP Award for the first time. He would win it again in 1965. In the ten years in between, he led the NL in fWAR seven times but didn’t get the MVP in any of those years. He could have won nine MVP trophies during a 12-year period.
In three of the seven seasons in which he missed out on potential MVP trophies, Mays’ advantage was within 1 fWAR. These are not egregious choices. Hank Aaron in 1957, Ernie Banks in 1958, and Sandy Koufax in 1963 were all close enough in value to Mays to be acceptable winners.
Live Feed class=inline-text id=inline-text-5
House That Hank Built
In four other seasons, Mays’ advantage was at least 2.5 fWAR. The first such year was 1955, when Roy Campanella (5.7 fWAR) won the hardware over Mays (9.0 fWAR). Campanella finished ninth that year in fWAR. Mays had almost 150 more plate appearances than Campy and led the league in home runs and slugging percentage.
Dick Groat won it in 1960, despite having 2.5 fewer wins above replacement than Mays. Groat was 10th in the NL in fWAR but his Pittsburgh Pirates team went from 78 wins to 95 wins and Groat was their best position player. There’s the narrative of the season popping up once again.
In 1962, Dodgers shortstop Maury Wills broke the single-season stolen base record when he nabbed 104 bases. He led the league in plate appearances, at-bats, triples, and steals and won a close race over Mays for the NL MVP Award (209 points to 202 points). Despite all those steals, Wills was 11th in the NL in fWAR and 5.2 wins behind Mays.
Finally, in 1964, Mays led the NL with 10.5 fWAR. He had a typical Mays-like season when he hit .296/.383/.607 and blasted 47 home runs, leading the league. The MVP went to the Cardinals’ Ken Boyer, who hit .295/.365/.489. Boyer’s Cardinals won the NL pennant in a close race over the Phillies (one game back), Reds (one game back), and Giants (three games back). If Mays’ teammates had been a little better, he likely would have won the trophy.