Did Consistency Decide the 2017 AL MVP Race?

The 2017 MVP race in the American League was an important one, or at least for me. Growing up, a lot of awards races ended up shaping the way I view baseball now. In 2010, King Felix won the Cy Young award with an amazing  2.27 ERA in 249.2 innings pitched but only a 13-12 record. Pitching wins no longer mattered. In 2012, Mike Trout put up what I think is the best overall season in my memory, dominating in hitting, defense, and baserunning, but still losing to Miguel Cabrera’s triple crown. Conventional stats overshadowed Trout’s season. In 2016, Mookie Betts put up an amazing 9.5 WAR season for a division winner, while Mike Trout put up an even better 10.5 WAR season for the subpar Angels. It turned out you didn’t have to play for a playoff team to win MVP.

The 2017 AL MVP race was pretty fun for me despite not having different baseball cultures clashing. Two guys were the obvious candidates, winner Jose Altuve and runner-up Aaron Judge. Traditional stats loved them; Altuve slashed .346/.410/.547 and stole 32 bases with 112 runs. Judge slashed .284/.422/.627 with 52 homers, 114 RBI, and 128 runs. Not only that but both players were the best player on a playoff team. New-age metrics loved both players too; Altuve had an 7.6 fWAR, .405 wOBA, and a 160 wRC+. Judge had a 8.2 fWAR, .430 wOBA, and 178 wRC+. With WARs so close, they were basically both the Most Valuable Player.

Yet, Altuve won MVP rather easily. This doesn’t necessarily mean that people thought he was way better than Judge but rather most people agreed that Altuve had the slight edge in such a close season. I’m sure traditional reasons factored into Altuve’s win. Altuve played for a division winner with 101 wins, while Judge’s team was a Wild Card team with 91 wins. Voters perhaps liked that Altuve had been due for an MVP while Judge was just a rookie. Altuve had a very high batting average, which we all still appreciate, that led the league and was 62 points higher than Judge’s. Judge, while walking at a great rate and even beating out Altuve in OBP (meaning he literally got out less), still struck out at an alarming volume of 208 Ks compared to Altuve’s 84. I think Altuve probably got a push being a traditionally “pure” hitter. Not only that, but the perception that Altuve was better overall player because he’s small and quick while Judge is huge probably existed too, although Fangraphs likes Judge’s outfield defense more, while not being as big on Altuve’s baserunning as you’d expect.

Honestly, I would have voted for Altuve as well. Of course, not for any of those reasons that I listed above. Rather, I viewed the two as equally valuable but achieving their value in two fashions: Judge being more volatile and Altuve more consistent. From a stats standpoint, a player’s final value is their final value but last year was the year I started looking at how that value was distributed across a season. From the beginning of the second half to the end of August, Judge hit .179/.346/.344, which over 191 plate appearances is a pretty long time to be a bad hitter. This definitely hurt Judge’s reputation, especially as it became headline news everyday that he added a game to his strikeout streak. It’s really the only reason I decided on Altuve. They ended the season with equal value but the human in me would want the guy who had great numbers by being consistently great.

I know in past awards races, volatility has had somewhat of a factor. In 2004, Vlad Guerrero had a huge push after hitting .363/.424/.726 in September. Jake Arrieta won the 2015 Cy Young over Kershaw and Greinke after having a historically great second half ERA of 0.75. The mindset here is that the games get more important as the end of the season approaches, therefore they get a push for playing their best when it matters the most. Judge’s case is a little different because it’s more about being great, then bad, then great again in September but it’s definitely similar in terms of player volatility in awards races. Judge’s huge September (ridiculous 1.352 OPS) however wasn’t enough to win. Obviously I believe each game is equally important because, well that just makes sense, but either way, these players were clearly better at certain times more than others and it had an impact on how we viewed them.

With consistency being a factor that came to my attention, I thought it would be fun to visualize how a player’s season looks. I think it’s a neat tool if you’re deciding between two equally great players.

As expected, Judge’s line has more curve to it while Altuve’s looks closer to linear with less dips and rises. What if we looked at a stat’s average by month? That way we can see how they performed by the numbers as opposed to just seeing their cumulative WAR slowing down or continuously rising. As suspected with Judge’s slumps, he dips to an average player for a two-month span while Altuve consistently remains well above average.

All of this is based off my original assumption that consistency IS better than volatility, just based off of logic. I think it would be fun to actually see how consistency correlates with overall season stats. Do players like Altuve in 2017 typically have better seasons, or do players like Judge in 2017 typically have better seasons? How would I even define a stat like consistency? Looking at the graph above, I could just use Rsquared from a regression line from each season; the closer to 1 Rsquared is, the closer to a linear line it is, and therefore the more consistent they are. That wouldn’t make as much sense though, since I’m not really predicting the x variable (games played) for WAR. I think what would make most sense is if I divided a players seasons by months, or weeks, or 50 at-bat increments, really anything and use standard deviation to see how much they vary from their final season averages. So for now I can just have fun visually tracking a player’s season…but stay tuned.