I recently posited a question on Twitter about baseball statistics. The question was in regards to what statistics people use to evaluate players these days. It’s a simple question on the surface, but in reality, it’s a question that leads to complex and often loaded responses. There are a lot of statistics saturating the baseball stratosphere, just as there are a lot of baseball fans, pundits, analysists, etc. with varying opinions about said stats. Baseball statistics and their applications are complex, and can’t really be explored properly through the medium of Twitter.
The easiest of the categories are the traditionalists, for lack of a better term. These folks rely on the quick and easy statistics that have populated baseball for years. Wins, Losses, RBI, Batting Average, ERA, Saves, etc. These are bread and butter stats. One could walk up to any baseball fan alive and strike up a conversation about Batting Average. Most people refer to these as baseball card statistics, which is as accurate of a label as I’ve ever heard ascribed to them. I was a student of baseball card statistics for the longest time. Many years ago as a young lad I started collecting baseball cards and before I gave up the hobby in my early 20s I had amassed quite a collection. For me, it was just as important to gain new information about a player from the back of his card as it was to own the card period.
Sabermetricians are everywhere these days but that doesn’t mean they are all operating from the same statistics. They represent the hardest category to pin down simply because there are so many different pathways of sabermetric thought. There are, however, three main branches of Sabermetricians and it’s based on three specific websites.
First, there are the devotees of Baseball Reference. Fans of Baseball Reference will more likely tout rWAR (their version of WAR), OPS+, and WHIP in their evaluations. More than anything else Baseball Reference is home to the Play Index. I dare say that there isn’t a fan alive who is interested in digging below the surface of statistics who doesn’t have a subscription to the Play Index and make generous usage of that titan of information.
Baseball Prospectus is the home of DRA, DRA-, bWAR (or WARP, their version of WAR), DRC+, BRR, FRAA, etc. They are a statistical powerhouse. Whereas Baseball Reference is concerned with a lot of raw data, Baseball Prospectus, as well as the upcoming third site, are mostly concerned with extrapolating on raw data. Extrapolate they do, and the statistics they provide give a better indicator of who the best players are now, were in the past, and will be moving forward.
The last of the three main Sabermetric sites is FanGraphs. In this land, you’ll find people talking about Spd, fWAR (you guessed it, their version of WAR), UZR, and WRC+ to name but a few of the statistics that form the backbone of their analysis. FanGraphs is very similar to Baseball Prospectus in that they are more concerned with refining raw data and being more accurate in analyzing the past, present, and future. I won’t lie, I tend to visit FanGraphs the least of the three. I still visit and I still use some of their statistics, but I do prefer Baseball Reference and Baseball Prospectus when I’m doing research.
That leads us to the category that I think the majority fall into, the cherry pickers. These people pick and choose the statistics they place the most value in from any source. They’ll freely use any of the aforementioned stats, plus things like wOBA, OAA, Pitchf/x data, and others from sites such as Baseball Savant, Brooks Baseball, etc. I am proudly a member of this grouping, the ability to freely use stats from multiple sites form the basis of all the statistical analysis and research I embark upon. I want to get to the heart of why Batter A had a down year, and cross-pollinating statistics from various sites is necessary to achieve the most accurate portrait of that batters down year.
The problem with the approach that most of us take is that a lot of the time we are offering learned responses from very different starting points. I can argue the merits of a pitcher with Julie, but what are we really arguing about if I’m referencing DRA- and she’s referencing cFIP? This is where a discussion is key. We can’t just throw statistics at one another and hope to come to any sort of agreement or conclusion. Well, I guess we could come to a conclusion but said conclusion will almost always be that one person is right and the other is wrong because we don’t agree with their metric choice.
How we talk about the statistics we use matters just as much as the statistics we use. I use the same base set of statistics when discussing a player. Batting Average, OBP, Slugging, BRR, FRAA, DRC+, DRA, DRA-, WHIP, and bWAR are my go-to base statistics. That’s not to say those are the only statistics I use, and chances are those statistics will be different from your statistics. Next time you disagree with why I think Player A is great, talk to me about it. We can share statistics, see where we’re both coming from, and dig deeper into why we think the way we do. It will be much more productive than just throwing statistics at one another with no end or common reference point in site.
Lead photo courtesy of Dalton Johnson – City News Service