Our Blog

Choosing Measures Wisely – A Lesson from the American League MVP Race

Posted August 31, 2011 4:35 PM by Mark Cutler

A recent debate stirred up by a Wall Street Journal sports article got me to thinking about the importance of choosing measures to determine the impact performance is having on your organization.

To summarize the article from Aug. 13, "The Obvious MVP ... in 1991," for you, it argued that the New York Yankees' Curtis Granderson would be the favorite for the American League's Most Valuable Player (MVP) award if it were 1991 because he was tied for first in runs batted in (RBI), second in home runs, led the league in triples, and even had 22 stolen bases. However, over the last 20 years, the major statistics (measures) by which MVP voters gauge a baseball player's value have shifted, largely due to the Sabermetrics movement led by Bill James and Billy Beane.

In the long history of Major League Baseball through the mid-1990s, article author Daniel Barbarisi says, the statistical categories upon which people relied to help them choose an MVP were runs, RBI, and home runs. From 1985 to 1995, the AL MVP award went to the league leader in runs or RBI seven out of 10 times.

Since Sabermetrics and Money Ball have taken hold over the last 20 years, things are much different. Now "OPS is king," says Barbarisi. OPS is on-base percentage plus slugging percentage, the combination of two more traditional baseball statistics. OPS has become so influential, in fact, that since 2008 the MVP in both the AL and the National League has been the leader in OPS five out of six times. Only Albert Pujols, in 2009, led his league in runs or RBI.

The thinking, according to Barbarisi, is that runs and RBI are more dependent on the team around a player because to score runs a batter behind you has to hit well and to get RBIs the batters in front of you have to get on base. OPS, on the other hand, combines the percentage of plate appearances that the batter gets on base with the total bases per at-bat, two statistics that are based solely on the performance of that player.

So, with OPS king, the leading AL MVP candidates are Toronto's Jose Bautista and Boston's Adrian Gonzalez instead of Granderson.

Does this mean that Granderson is any less valuable than Bautista or Gonzalez? I would argue, no, that all it means is that the voters have changed – for the time being, anyway – the measures they use to define value.

In fact, I think that in sports where there is so much talk about players that are "clutch" or "choke" in high pressure situations, RBI may be a better measure of value because it means a player is getting a hit when it has more value to his team – when someone is in scoring position.

Regardless, I think it is a great example of how organizations have to choose their metrics wisely in order to measure what they believe is truly valuable.

Filed Under Measurement