Why The MVP Won’t Win MVP

Mike Trout is the best player in baseball. He once again put up a ridiculous slashline in 2020 – .281/.390/.603 – although it was somewhat disappointing considering what he has done over the past half decade. He may be high on the list, but it is safe to say he will not defend his American League Most Valuable Player (MVP) Award this year.

The award is not about recognizing the best player in each league, but rather – as the name suggests – the most valuable. That wrinkle is an interesting nuance that leaves plenty of room for interpretation and biases. How we choose to define value can make all the difference in who takes home the award and legacy that comes with it. So first, let us start by looking at how each season ends.

The Nationals’ wild run to the World Series last year was something straight out of Hollywood. The horrid start, the comeback in the wild card game, triumphing over the one seed Dodgers, and finally the game seven win against the Astros. From a non-business perspective, they achieved the ultimate goal: win the world series.

Perhaps that is where the evaluation should begin as well: Which player contributed the most to their team winning the World Series? Juan Soto would probably have won the award with a good regular season, great postseason, and plenty of signature clutch moments along the way. Using Baseball Reference’s championship win percentage added (cWPA), Soto added about 42.6% to the Nationals’ chances of winning the World Series that year over an average player.

That number may seem high… because it is. The closer to the end of the World Series you get, the more wild the swings in a team’s championship chances are. Soto’s regular season cWPA was only 2.0% – practically irrelevant next to what he did in the playoffs. His one run single in game 7 alone increased his cWPA by 12. Thankfully MVP voting occurs before the playoffs or else it would always go to whichever random player landed the knockout punch in October. Looking at only regular season cWPAs is better, but it still has the same pitfalls. 

In the 2008 season, neither MVP race was a run away. Dustin Pedroia was picked from a mixed bag in the American League while Albert Pujols edged out Ryan Howard in the National League. Know who cWPA said was supreme? John Danks. He had a good season (12-9 record, 3.32 ERA, 195 IP), but by no means MVP worthy. He did not even receive a Cy Young Award vote. So why did he have the highest cWPA? Game 163.

After 162 games, the Twins and White Sox stood tied for first in the AL Central. Their record was worse than the wild card Red Sox, so whoever lost the tiebreaker would not see the postseason. Danks toed the rubber for the White Sox and pitched eight shutout innings en route to a 1-0 victory, securing Chicago’s playoff opportunity. That one outing was incredibly high leverage and Danks delivered.

He massively increased his team’s world series chances within the regular season, but should that one game carry so much weight in his candidacy? Should it reign over the season-long contributions of someone like Pedroia? There is a valid argument to be made that cWPA reflects players stepping up in high leverage situations and is a solid statistic in finding a season’s most valuable players. If you think so, let me present your 2020 Most Valuable Players:

Not too shabby. The soon-to-be unanimous Cy Young winner Shane Bieber and Mike Yastrzemski – the unheralded driving force of a team that was in the playoff race up to the last day. Both are solid candidates, but of course cWPA can be a controversial line of thought. Those who don’t think that divisional games and games toward the end of the season are particularly impactful could scoff at the disproportionate weighting. Why should the 89th win be worth more than the 1st? 

Win Percentage Added (WPA) takes this approach. Clutch moments within a game are still weighed more heavily, but each game is viewed the same. Doing so can be both a strength and a weakness. It prevents the performance in any one game from skewing the statistic, but it also devoids context from how important each win actually is. If a player on a bottom-feeder team adds five wins, is it as valuable as the player that adds five wins on a team that barely makes the playoffs? Most would probably say no.

To reveal the issue with WPA, we have to zoom out and look at some of the other top candidates. For instance, Liam Hendriks finished behind Bieber in AL WPA. Looking solely at WPA, Blake Treinen was the most valuable player in baseball in 2018 and Kenley Jansen was in 2017. The point of the closer role is to maximize leverage and naturally WPA with it. It is certainly a decent way to gauge the most valuable player in each league, but it would be quite divergent with how voters have historically conducted themselves. Nevertheless, it is a valuable statistic to at least consider:

Huh, Bieber and Yastrzemski again. The fact both remain shows they did not rely too heavily on being on playoff fringe teams to inflate their cWPAs; rather they have both high WPAs and cWPAs due to generally being productive and clutch in games.

That idea of “clutch” does not bode well with most baseball stat nerds. Research shows that clutch is almost entirely luck, but even if Yastrzemski’s clutch hitting was lucky, was it not still valuable? Lines get quite blurry here. The widely accepted statistic Wins Above Replacement generally tries to take luck out of the picture and put everyone in context neutral situations.

There are multiple versions of WAR, but I will be using baseball reference’s WAR (rWAR) as it focuses more on the value a player contributed as opposed to the underlying skill of the player. Since becoming mainstream around 2015 it has proved quite powerful in predicting MVPs (probably because writers have started leaning on it). Looking at the leading candidates, we can see why:

WAR tends to produce names better aligned with our perceptions because it looks at season stats like we do. No MVP voter has the time to check if Mookie Betts’ home runs were more clutch than Yastrzemski’s, just that he had more. Betts and Yastrzemski were generally equal in contextless offensive production, but Mookie is a far better base runner and defender giving him the edge in rWAR.

Bieber, once again, stands atop the American League ahead of DJ Lemahieu (2.9 rWAR) and Jose Abreu (2.8 rWAR). Based on the various metrics I have presented, he was the most valuable player in the American League. That being said, he is not the AL’s MVP front runner. What hasn’t been accounted for? Bias.

In high school I wrote a pretty basic statistics paper in math class using binomials to show how voters have bias against players on subpar teams and against pitchers. It was a subset of MVP voting research I was doing at the time, but it shows how the divergent definitions of value can shake up results. Voters have simply proven to prefer players on winning teams, especially playoff teams.

It has also historically been popular to insist pitchers can not achieve the same level of value – despite what WAR and WPA say – as position players because “they don’t play everyday”. In 2018, Christian Yelich won NL MVP with a bWAR of 7.3. Aaron Nola, Max Scherzer, and Jacob deGrom finished 13th, 10th, and 5th with bWARs of 9.7, 9.2, and 10.3. deGrom really got a double punch with being a pitcher on a losing team, but WAR suggests he was a full three wins more valuable than Yelich! With a gap of only 0.3 over the top position players, Bieber does not stand much of a chance.

In my opinion, Shane Bieber and Freddie Freeman should win MVP, but Jose Abreu seems to be the favorite for the American League. Both Abreu and Freeman currently sit as favorites because they were central pieces to lineups on good playoff teams, they ranked near the top on advanced metrics (enough to satisfy the statistically savvy), and they put up impressive traditional statistics (hits, RBIs, batting averages).

Value can be difficult to put a finger on and that is doubly so in a shortened season. The metrics have less time to pan out and agree on a victor leaving voters to cling onto what they trust most. When it comes down to it, we do not trust metrics we do not understand. I present to you the 2020 “Most Valuable” Players:

All numbers taken from baseball-reference.com