SABRmetrics and the MVP, Cy Young, and Rookie of the Year
Baseball is as mathematical as it is spectacular. These multiplicities of understanding the game underpin some of the toughest questions about measuring player performance. What about awards, such as the MVP?
Traditional statistics like batting average and on-base percentage don’t account for situational performance, and slugging percentages are fluffed with singles. So, in discussing player performance, we must account for situational success–even in awards races such as the MVP, Cy Young, and Rookie of the Year.
Clutch hitting will be left to the philosophers. For offensive categories, calculating a player’s Runs Productivity Average, or RPA, accounts for the direct—hits or runs scored—and indirect—runners set in motion—run potential that each player contributes. What is being measured is the amount of scoring situations a player creates versus the league average.
For pitchers, weighing their statistics puts a perspective on how they performed during a single season overall, rather than in comparison to their past. This eliminates the negative effects of bad defense on a pitcher’s ERA.
The problem faced in calculating a position player’s performance is the different jobs each position holds on the field. Calculating defensive runs weighs each player’s stats with the league average at their position, accounting for various differences.
Using these methods, we can answer a few of the tough questions surrounding the postseasona wards race.
We will not examine the races that have a clear front-runner: National League MVP (Kris Bryant) and National League Rookie of the Year (Corey Seager), or the managerial and comeback player awards.
AL Rookie of the Year
Rookie of the Year is an interesting race to measure because both position players and pitchers qualify for consideration. Therefore, we can’t qualify their value with traditional statistics because doing so wouldn’t account for situational play of each player. This means we must use weighted values.
Offensively, Gary Sanchez of the Yankees and Tyler Naquin are the front runners for the award. Both have come up big for their teams in spots, but Naquin has player significantly more than Sanchez.
If we use Linear Weights, we find that Sanchez has an RPA of 9.3 to Naquin’s 11.8. Comparatively, Naquin has a WAR of 2.1 and Sanchez has a 3.4. This is mostly due to the shorter service time for Sanchez because WAR is a direct statistics which doesn’t account for chances created.
More from Statliners
Choosing between the two then comes down to how the voter values the amount of games each one has played.
Michael Fulmer and Edwin Diaz present a similar problem in comparison considering that Diaz is a reliever and Fulmer is a starter. In order to compare the two, we need to weigh their stats against the average for their position.
Diaz pitched eight runs better than the average reliever in 2016 whereas Fulmer was 21 runs better than the average starter. Even with Fulmer’s recent slide, recording a 4.32 ERA in September, he still gives the Tigers a better chance to win than Diaz gives the Mariners.
If I were voting for this award, I would give it to the player who recorded the greatest disparity from his opponent. In this case, Fulmer would receive my vote.
The Cy Young Race
The Cy Young award is supposed to be awarded to the best pitcher in each league. Many critics cite the culpability of this award to be portrayed as a popularity award. Using a weighted method to determine performance only accounts for performance against league averages, removing all chances for popularity to take precedent.
The issue in the American League is whether or not Zach Britton, a relief pitcher, qualifies for the award. Weighing his stats against average relievers gives a more accurate description of his skill and ability than innings pitched and ERA ever could.
Britton pitched 23.3 runs better than an average reliever, while the front-running starter, Rick Porcello, was 26 runs above average. Britton is third highest in saves this season with 45, and has converted 100% of his opportunities. Porcello has the best record in the American League at 22-4.
If we were going to factor in WAR, Porcello has a 4.7 and Britton has a 2.2. As dominant as Britton has been at the end of games, Porcello means more to the success of the Red Sox.
Jon Lester and Max Scherzer create a similar problem in the National League. Lester is pitching 38 runs above average to Scherzer’s 33. We can see that this difference is accounted for in their Cy Young Points—183.4, 181 respectively—though they make the race seem much closer than it is. What’s interesting about the Points ranking is that it has Kyle Hendricks ranked sixth behind Johnny Cueto, though Hendricks ranks higher than Lester with 42 runs above average.
If I had a vote, I would cast it for Porcello and Hendricks.
Unlike the National League race, which seems to be a runaway for Kris Bryant, the American League MVP race is a tight competition between two players: Jose Altuve and Mookie Betts (although a great case can be made for Mike Trout). Both have been irreplaceable for their clubs this season, but how do we find out who is more valuable?
To answer this, we need to account for each one’s presence offensively and defensively.
Altuve created 61 more scoring opportunities versus an average player, while Betts created 69 chances–both numbers are good, but Betts’ is nicer. You can make the argument that without Altuve in the lineup, the Houston Astros would be the same whereas the Boston Red Sox without Betts is drastically different. But, the separating factor in the MVP race for the players is their defense.
Betts has saved 32 defensive runs this year, versus Altuve’s zero. Altuve also trails in dWAR rankings 0.4 to 2.9 respectively.
In order to declare a player to be the most valuable in the league, he must contribute with the bat and glove. In this case, Betts is the American League MVP.