Today I want to discuss the stat that I love to hate the most. First, however, I want to reiterate that my opinion of specific statistics is in no way intended to undermine their value but to emphasize that an over dependence on a specific statistic can actually undermine the importance of other qualitative variables that are not quantifiable. Anyways, the statistic I want to talk about is the WAR stat, or the Wins Above Replacement stat which is now used ever more frequently in professional baseball. The statistic was formulated by statistician Sean Smith with the intent of measuring the value of a player expressed in wins.The stat is based on an individual players averages compared to league averages and then again to those of a theoretical replacement player. Now the formula for this stat is highly complex and takes into account an array of variables, which is precisely my biggest issue with the stat. even though it takes into account fundamental stats such as On-Base Percentage, slugging percentage and other traditional baseball stats, the formula is also dependent on the use of other arbitrary numbers. For example, WAR calculates speed, fielding proficiency, and in an attempt to connect generational and level of field playing gaps, stadium configurations. Lets start with the calculations made for speed. Stolen bases are calculated which would be fine except for the fact that a host of variables contribute to a players ability to steal a base. First and foremost is how was the ball delivered from the pitcher to home plate. A wild pitch from the pitcher or poor fielding from the catcher often times leads to an uncontested steal. It would not be uncommon for this to inflate a runners stolen base numbers by at least 15% over the course of a season. Another category used is fielding. This is where the use of arbitrary numbers comes into play. In baseball it is widely viewed that some positions are more difficult to play than others. While there may be some truth to that it still remains highly subjective and therefore difficult to quantify. WAR, however, applies a rating system and gives a rating for each individual position that ranges from a -8 to a +8. I do not know how these rating are justified as the end all be all, but it is taken into account regardless. I want to move on to my least favorite part of the stat which is the comparison of the player being analyzed to its potential replacement. Numbers indicate that replacement players put up numbers under the league average. My argument to this is how do you calculate a a hypothetical players ability, and how do you not take into account that one players replacement could potentially be better than another. Again there are a number of variables that go into a players potential. For example, while a replacement player may not be as effecient offensively as his counterpart he may actually be better defensively and in turn helping the team prevent opposition runs and thus more wins. I just do not know how you justify a hypothetical players worth. To sum it all up my view is that the WAR stat depends on too many arbitrary factors for me to use it as my sole means of evaluation of player talent. Again, I do want to stress that I believe it has value just do not become over dependent on it. Next week I will discuss the sports statistics that I do like and how I feel their best used.
