What Are “Sabermetrics” Part Seven

Last week, in Part Six this series, we looked at a statistic called Pythagorean Win Expectation which uses the differential between runs allowed versus runs scored to generate an “expected win total” for a specific team, over a fixed amount of time (usually a full season). Over the next couple of articles, I am going to attempt to shed some light on a pair of similar statistics, called Wins Above Replacement (WAR) and Value Over Replacement Player (VORP).

The cool thing about the listed formulas is that they can be used for both pitchers and hitters, similar to BABIP which was discussed earlier in this series. The not so cool thing is that they can be complex and the base level established for comparison in each formula (replacement level player) is less then scientific.

So, for this week’s discussion, what is WAR? Well, it is more then just a pretty good album from the rock band U2 (maybe not as funny as I thought). Getting serious, according to Wikipedia, “WAR calculates the total number of wins that a specific player adds to his team over the course of a season, by comparing the player’s performance with that of a fictitious replacement level player”. Said “replacement player” can be viewed as the equivalent of a AAA player who might only play as a replacement for an injured player and whose skills are “far below league average”.

Another way of defining a replacement level player is to identify them as a player you can get “for next to nothing, off of the scrap heap”. You can see that the listed definitions are based largely on opinion, which leaves a lot of room for interpretation (and argument). However, the concept is easy to understand (expressing a player’s worth, compared to a basic replacement player where there will be a drop off in performance).

How do you figure this mystical statistic out? Easier said then done, as you will see shortly. WAR is the sum of the win values that a specific player’s offense, defense, pitching, playing time, ball park, year and league context, expressed in the form of runs added or runs prevented (nice and simple). Going against the grain of my other contributions in this series, it is much easier to research a player’s WAR figure on any reputable statistics website (you will thank me later, especially if you try and figure out how to calculate WAR for a pitcher).

My goal here is to provide a basic understanding of how it is generated and then focus on what the expression actually means. For example, Jose Reyes contributed a WAR of 6.2 over the 126 games that he played in 2011. Simply put, if Jose Reyes was replaced at the start of the year by a replacement level player (meaning a WAR of zero), the Mets would have won six fewer games during the 2011 season. Not a big deal when you finish twenty-five games out of first place, but a huge contribution regardless.

In theory, anyone on a major league roster should have a positive WAR figure (over the theoretical baseline of zero). A player posting a WAR below zero is causing more harm then good and could be replaced by our loosely defined “replacement player” to the team’s overall benefit.

Most regular position players will accumulate a positive WAR. A “good” player will usually finish in the 3 to 5 WAR range over the course of a full season. Very good players can provide a WAR figure of over 5 and an excellent season is anything over a 9 (Jacoby Ellsbury, for example, posted an impressive 9.4 WAR figure which was the highest among batters in 2011).

Another way to look at WAR is to use the expression to see if a specific player “justified” their salary for a given period. You arrive at this figure by multiplying that player’s WAR by the average value of a win for a particular season (obtained separately using total dollars spent and dividing by the total number of wins for all of MLB). By doing so, you would have a total value, expressed in dollars, that a specific player contributed for the season in question. You can compare the fixed salary to the dollar amount that the player was “worth” and spawn all kinds of arguments and discussion on blogs and sports radio!

So Mets fans, as we slog deeper into the relatively quiet offseason, you most definitely will hear different talking heads refer to a player’s WAR, at some point. Remember, that is simply how far above replacement level that player is thought to be. The higher the WAR, the more valuable the player (but like any statistic, they are much better at defining past contributions and then they are at predicting future performance).

If Jose does leave for the Marlins, or another team, the Mets need to find “six wins above replacement level” either at SS (unlikely) or a combination of improvement at other positions (much more likely), to replace his offense.

Next week, we will address VORP and show that it is a valuable, yet similar statistic to WAR.

Enjoy your Sunday afternoon and get ready for the Winter Meetings, where hopefully something exciting will finally happen this offseason.