Friday, April 16th, 2021

Introduction to FanGraphs, WE and WPA

April 18, 2010 by  
Filed under New Media, Statistics and Analysis

Given that my Washington Nationals were being pummeled 11-7 by the Milwaukee Brewers on this Sunday afternoon, I thought my time would be better spent talking a little bit about FanGraphs ( and some key statistics being used to analyze games. is a company located in Arlington, Virginia that provides statistics for every player in Major League Baseball history. The website, among other things, plots the course of every game, showing which hits and outs changed outcomes.  For self-described data “geeks” like myself, it is as if we had died and gone to heaven!

For every MLB game, after every at bat, Fangraphs calculates the probability that one team or the other will win the game.  You can see from the graph at the right, that the chances of the Brewers winning the game, win expectation or WE, became almost 100% as they scored 10 runs in the first inning against the Nationals.  The innings are plotted along the horizontal axis and the probability of winning plotted on the vertical axis ranging from 100% probability for Nationals at the top to 100% probability of the Brewers winning at the bottom.  Each games begins at the middle of the vertical axis where both teams have a 50/50 chance of winning.

I know, I know, you are saying “How is the WE calculated and how is it used?”.  WE stands for the percent chance a particular team will win based on the score, inning, outs, runners on base, and the run environment.   The WE and other metrics being talked about in this article are explained in more detail with The Book:  Playing the Percetages in Baseball by Tom Tango, Mitchel G. Lichtman, and Andrew E. Dolphin

Let’s get back to the all-important Nationals-Brewers contest. If you notice the far-right hand tail of the above graph, taken during the Nats’ portion of the fifth inning, you see the curve go up slightly.  This reflects a 3-hit rally the Nationals began against starting pitcher Doug Davis.  After Adam Kennedy singled, the WE for the Nationals rose to 1.5%.  After a single by Alberto Gonzales, the WE rose to 2.3% and after another single by Willy Taveras, which loaded the bases, the WE for the Nationals rose to 3.7%.  Ultimately, the Nationals went on to make the score 5-10 with Adam Kennedy flying out with the bases loaded.  Kennedy’ out reduced the Nats WE from 10.3% back down to 5.2%.

Win expectancy (WE) is a useful tool to keep a running tally of each team’s probability of winning.  But the statistic actually has more value when it is added across time for each player to give an approximation of that player’s value in team wins.  This statistic is called WPA (win probability added): WPA is the difference in win expectancy (WE) between the start of the play and the end of the play. That difference is then credited/debited to the batter and the pitcher. Over the course of the season, each player’s WPA for individual plays is added up to get his season total WPA.

The WPA seems to be a pretty good barometer of player worth.  The WPA leaders for 2009 were Albert Pujols at 8.24 and Prince Fielder at 7.79.

Understanding the run-scoring potential or incrementality of every potential play gives managers an upper hand.  It is why the conventional wisdom of sacrificing a runner to second may be conventional, but may not actually be wise.  More on that to come with subsequent articles.

To some casual baseball fans, these statistics might seem like overkill.  However, the statistics and analytics revolution within baseball is already well underway.  Bill James was the beginning, Moneyball brought it out to the mainstream, and books like The Book and companies like are making this type of analysis mandatory for all MLB teams.  Fortunately, you won’t need a fancy calculator, slide rule or even abacus.  Just an internet browser will do.

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!

Better Tag Cloud