November 16, 2016

ELDEN | Examining Some of Baseball’s Most Obscure Statistics

Print More

In 2003, Michael Lewis published a book that would cause sweeping changes throughout baseball — Moneyball: The Art of Winning an Unfair Game. This book, a bible among the most die-hard baseball fans, opened the eyes of many front office executives to the world of statistical analysis beyond the basics that had been widely accepted in player evaluation.

The low-budget Oakland Athletics’ front office attempted to find the most cost-effective ways of putting together a roster after losing many big-ticket players to free agency. Many teams, including big-market franchises such as the Red Sox and Dodgers, adopted this strategy. The original “Moneyball” approach was to focus on On-Base Percentage (Hits+Walks+Hit by Pitches/Plate Appearances) and Slugging Percentage (Total Bases/Plate Appearances). These statistics were argued to be more important for run-creation, and, therefore, should have been valued more highly than the traditional statistics such as batting average, home runs, runs batted in, or stolen bases.

Recently, these statistics have been explored even more in-depth. Websites like Fangraphs and Minor League Ball perpetuate this movement towards advanced statistics. Bloggers, fans, and front office executives are all shifting away from these basic statistics. There are five major statistics that I will describe in depth: WAR, BABIP, xFIP, OPS+, and ERA+.

WAR, or Wins Above Replacement, is a simple statistic to use but difficult statistic to derive. The formulas for its calculations vary across websites and evaluators. It is essentially all relevant, unbiased statistics compiled into one simple, easy to read number. Below is an example of a formula for the offensive component of WAR.

Fangraphs Formula for hitters: WAR = (Batting Runs + Base Running Runs + Fielding Runs + Positional Adjustment + League Adjustment + Replacement Runs) / (Runs Per Win).

A player’s WAR is the number of wins that they provide their team over a replacement level player (someone who could be found from the ‘scrap heap’ and thrown onto a roster). A rule of thumb is every WAR is worth approximately five million dollars on the open market. For example, Dexter Fowler’s 4.2 wins above replacement last season made him worth ~21 million dollars to the Cubs. Big-ticket free agent Jason Heyward was only worth ~7.5 million dollars with a 1.5 WAR. Fowler was worth 12 million more than his deal last season and Heyward was worth 14 million less. A team with a replacement level player instead of Fowler would have theoretically won four less games.

This method helps teams evaluate more holistically when attempting to determine a player’s worth in trade or free agency. The statistic is usually subjectively calculated depending on who is calculating because of difficult factors to consider such as defense.

BABIP is an interesting and useful statistic that tells us whether a position player’s high batting average is a product of luck or skill in a small-to-mid sized sample. BABIP is simply batting average on balls in play — so any at bat that results in a strike out is not counted. If a player, such as the Nationals’ Daniel Murphy, has a breakout season with a .347 batting average, we can see if this breakout is the result of a new and improved talent level or simply luck.

If a player has a significantly higher BABIP than they have had throughout their career, it is likely the result of luck. Murphy came out of last season with a .348 BABIP to give him a career BABIP of .319 over thousands of plate appearances. This would suggest Murphy is expected to regress back to more reasonable levels next season.

xFIP is essentially a normalized version of FIP and by extension, ERA. Both xFIP and FIP can be read like ERA’s, where an ERA under 3.5 is considered good. One of the most pervasive arguments against ERA is that it does not take the fielding abilities of a pitcher’s team into account. By controlling outside variables, this gives the truest indication of a pitcher’s talent.

xFIP counts 10.5 percent of fly balls produced as home runs, therefore controlling the variability of talent in hitters and propensity to give up home runs in different ballparks.

xFIP = (13*(Pitcher Fly Balls*(League Home Runs/League Fly Balls))+(3*(BB+HBP))-(2*K))/IP + constant

FIP Constant = lgERA – (((13*lgHR)+(3*(lgBB+lgHBP))-(2*lgK))/lgIP)

These formulas do not contain any factors that could be influenced by a team’s defense or similar confounding variables.

OPS+ and ERA+ are two commonly used statistics when evaluating hitters and pitchers. OPS is simply On-Base Percentage plus Slugging Percentage. OPS+ is just OPS normalized to a scale in which 100 is average, and any number above or below 100 is the percentage that their performance was better or worse than league average. ERA+ is the same statistic except with ERA instead of OPS.

These statistics are often complicated to calculate, but extremely easy to read and use practically. If you know how to read these statistics, you are already more knowledgeable than most fans. Most of these statistics are “improvements” or consolidations of old statistics – to evaluate players easier. Baseball is a numbers game- and it all comes down to good, ol’ player evaluation.