When David Appelman decided to implement a full-time writing staff back in April, Fangraphs was still finding its way. Complete with win probability metrics, game graphs, and pitch type information, Fangraphs offered all of the traditional stats with interesting extra numbers. Statistics like WPA, Leverage Index, and WPA/LI helped not just to explain the performance levels of players but also the story aspect of these contributions. By adding a writing staff to a very stats-oriented website, the goal became to provide insightful analysis while incorporating these fantastic but not terribly utilized metrics.

Curious about how the toughness of Mariano’s save situations differed from those of Bobby Jenks? Which players posted higher wins above average totals due to clutch hitting? The highest percentage of sliders thrown in the league? Which batters swung at the most pitches out of the zone? The answers to all of these questions were available at Fangraphs, which helped to separate the site from other incredibly useful baseball websites.

Over the next few months, changes and additions overtook the site. We were adding new metrics, altering the leaderboards, providing more information on the player and team pages, and doing whatever possible to improve the site from a user-friendly and information standpoint. As the weeks continued to roll by, even more metrics were added, and it soon became somewhat tough to keep up with all of the additions. Over the next few months, both Dave Cameron and I will break down these statistical additions in the hopes of developing a much better understanding across the blogosphere. For the inaugural post in this semi-regular series we are going to discuss two of the offensive additions, both of which tie into one another.

### wOBA

To grasp the importance and usefulness of wOBA we must first understand why stats like BA, SLG, OBP, and OPS fall short as evaluative barometers. Batting average measures the relationship of hits to at-bats but has two major flaws: every hit counts the same and walks or other times on base are not counted. A single is clearly not worth as much as a double, triple, or home run, but average does not care. And, while a walk is not “as good as a hit,” it does carry value. BA ignores the walk entirely and solely counts recorded hits as times on base while measuring them against at-bats, not plate appearances.

OBP does include walks but again fails to differentiate between the hits. SLG does differentiate between the hits, but does so in an inaccurate fashion, and ignores walks. Essentially, the ideal solution would be a metric that properly weights the different hits, values walks and other game events, and places the end result on a familiar scale.’

Luckily, a metric fitting this bill exists, and it is called wOBA. Weighted On-Base Average, developed by Tom Tango and used predominantly in The Book: Playing the Percentages in Baseball, takes the linear weights of each game event and scales the result similarly to OBP, where .330 is about average, .370+ is great, and below .310 or so stinks.

Properly weighting the events refers to the ability of each to produce runs. Ultimately, scoring runs is the name of the game on offense, and metrics like OPS overvalue hits while undervaluing walks. wOBA weights every possible outcome of a plate appearance very accurately, providing a telling window into the offensive success of a player.

Now, not every season brings with it the same exact run environment, as certain eras scored more or less runs than others. This leads us into a discussion of run environments, which are very important when discussing wOBA and wRAA. The environmental factor refers to how easy or difficult it is to score runs in a given context. This directly reflects on the value of a play. In a low-scoring season or stadium, it is less likely that a single or double will produce a run than in, say 1998, or at Coors Field. The reasoning deals with the idea that whatever is causing this low run environment—be it dominating pitchers, the stadiums, etc—makes it less likely that the hitter following the guy that just hit the single or double will get the base hit required to score a run.

The wOBA added to the site reflects this since the weights for different events are calculated on a year to year basis. A double might be worth .595 runs one season and .525 runs in another, and it would be inaccurate to apply one set of weights for the entire course of baseball history. The wOBA on Fangraphs is adjusted for run environments from a seasonal standpoint but not yet in terms of park factors. However, as we will discuss in the next section, the park adjusted result can be found in another area on the site.

As an example of wOBA in action, take a look at the 2008 campaigns of both Kevin Youkilis and Hanley Ramirez. Youk beat Ramirez in the OPS department by almost 20 points, .958 to .940. Both had great years, but OPS pegs Youkilis as the better hitter.

In terms of their slash lines, Youk hit .312/.390/.569 while Ramirez hit .301/.400/.540. Our wOBA also factors stolen bases into the mix, of which Ramirez had a 35 to three advantage. The baserunning advantage, as well as a 10-point advantage in OBP, gives Ramirez a .405 wOBA to Youkilis’ .402.

Another example would be a comparison of Ramirez and Ryan Ludwick. Ludwick hit .299/.375/.591, for a .966 OPS, almost 30 points superior in OPS to Ramirez. Despite a 50 point advantage in SLG, Ramirez’s 25 point lead in OBP and lead in stolen bases places his wOBA a mere point behind Ludwick’s .406 mark. By evaluating these players solely on their slash lines or OPS, we get drastically different results, with the reasoning that the components of the slash lines do not incorporate all pertinent plate appearance events and that OPS undervalues the OBP aspect.

These were the 2009 major league leaders in wOBA:

1) Pujols .458 2) C. Jones .446 3) M. Ramirez .432 4) Bradley .423 5) Berkman .419 6) Holliday .418 7) Quentin .414 8) A. Rodriguez .413

wOBA is available on the player pages and leaderboards at Fangraphs, and, in a perfect baseball world, would be one of the primary evaluative barometers.

### wRAA

wOBA also has another tremendous usage, as it easily converts to offensive runs above average. Titled wRAA, Weighted Runs Above Average, this conversion offers the offensive component in our player valuations. The formula to convert wOBA to wRAA looks like this:

((wOBA – lg wOBA)/Scale) * PA

In English, that is the league average wOBA subtracted from the individual wOBA. The difference is then divided by a scale to normalize for the aforementioned run environments. This scale is generally around 1.15, which represents the weights that season are 115 percent greater than the “standards” weights calculated. In doing so, we more closely match that season’s OBP. This scale is only necessary to convert the wOBA figure into runs above average.

Once the difference between individual and league wOBA meets the wOBA scale, the quotient is multiplied by the batter’s total number of plate appearances. Using Ramirez once more, who has a .405 wOBA in roughly a .332 wOBA league, with 693 PA, we see that he is worth +44 runs above average. Due to the PA aspect, wRAA rewards playing time while wOBA does not. Ramirez may trail Ludwick by one wOBA point, but leads him by five actual runs. And Ludwick has a four-point lead over Youkilis but falls just one run ahead in wRAA.

wRAA, like wOBA, is not adjusted for park environments on the Advanced section of the player pages and leaderboards, but is on the Win Values sections. Under the pseudonym ‘Batting’, the wRAA figure produced from the wOBA conversion formula incorporates park factors. Hanley Ramirez clocks in at +44.3 runs, overall, but +43.8 when the adjustment for parks are taken into account. Youkilis slips from +38.1 to +36.2. And Ludwick actually increases his output from +39.5 to +40.7. After the park adjustments, Ramirez leads Ludwick by three runs, not five, and Youkilis falls from one run behind Ludwick in runs above average to four.

The 2009 wRAA major league leaders were:

1) Pujols 68.9 2) M. Ramirez 56.3 3) C. Jones 52.2 4) Berkman 50.2 5) Teixeira 46.1 6) Holliday 46.0 7) H. Ramirez 44.3 8) A. Rodriguex 41.8

And Barry Bonds 2001-2004:

2001 .539 119.0 2002 .546 111.1 2003 .506 82.0 2004 .538 108.8

Fangraphs has made plenty of incredible additions lately, but wOBA and wRAA are two of the most important from an evaluative standpoint. These metrics properly value offensive contributions and produce tangible runs above average results. Hopefully the matter of time before these two stats become more widely adopted is very short. Until then, know that you can find them for anybody on the site dating back to 1974.

## Leave a Reply