A Thought About Statistics

The last few years have seen an explosion in the use of advanced and sabermetric statistics to analyze the national pastime. This has lead to  a great divide among commentators and fans as to the value of statistics in judging players, teams, and performances. It seems that this issue has lead to the forming of two separate schools of thought, with no middle ground. There are those who see statistics as being the final arbiter of any dispute about the relative value of a player and team, noting that our eyes and hearts form subjective judgments and therefore often lie. On the other side, there are those who disdain any stat that they did not grow up with, stating that sabermetricians must leave their mother’s basements, put aside the box score, and watch some games. The lack of a middle ground of understanding has lead to this feeling among more casual fans that those who use statistics are spreadsheet obsessed nerds, and those who ignore the numbers are crusty old curmudgeons who refuse to accept the wave of the future. 

I think this dichotomy and disconnect calls for the establishment of some form of conciliatory use of statistics that everyone can appreciate, which is something that we hope to implement on this site. There needs to be an understanding that statistics are valuable in determining player value, but they are not the only evidence that should be considered when analyzing the sport. Statistics continue to be developed precisely because they are not yet perfect, and do not yet quantify everything that goes into a particular game or season.

On the other hand, using only our eyes and our memories leads to similar problems. Firstly, nobody can observe every context within which a game was played, such that our accounts of many players depends solely on statistics. How can I judge Derek Jeter’s defense solely based on my observations when the players I am comparing him to are available to me only through the stat sheet? Additionally, we form attachments to certain players, and extraordinary plays or games will cloud our judgements as to a player’s relative value. Scott Brosius hit a few big home runs, and suddenly a largely average player has become almost a legend. Statistics help level the playing field by eliminating the biases caused by the vagaries of the human memory. 

The solution to these issues is to depend both on statistics and our own judgement. I can use advanced statistics to argue that Andy Pettitte pitched better than his ERA last season and should bounce back, but I can use my eyes to note that he did not look right in the second half of the season and to be a bit worried about him. These two conclusions are not contradictory, as there are unquantifiable areas that must be filled by observation. As Don Ehrke of Dugout Central points out:

My point is that “value” often remains just beyond our grasp. Many fans are on a pilgrimage searching for baseball’s “Holy Grail” – the one statistical tool that measures all things and neatly quantifies them in one simple number. This quest has produced not only Runs Created and Base Runs but Win Shares, VORP, WARP, OPS+ and countless other “alphabet soup” formulas. Yet we never arrive at the destination…….This doesn’t imply that we should cease developing more accurate statistical tools – to create no new formulas is to support the fallacious position that there is no truth and all things are indefinable shades of gray. Using absolute statements to contend that nothing is absolute is irrational. Instead we should enjoy the happy position that baseball both lends itself to statistical analysis yet remains elusively above all attempts to grasp it entirely.

Hopefully, as the sabermetric field ages, we can find some compromise on how best to use statistics and visual observation together. Until then, it is important to remember that neither element is completely useful by itself in all circumstances, and we need to use both to make many of the value judgements that we want to make about baseball.

Where do you stand on this issue? Pro-stats, anti-stats, or somewhere in the middle?

0 thoughts on “A Thought About Statistics

  1. I was never big into stats, at least not prior to blogging, simply because I never knew about them. I wasn’t a big baseball internet head until a few years ago and, at that time, I never really understood the advanced stats so I wouldn’t read texts that utilized them. It wasn’t because I didn’t want to, however. I just could not do it (they can be complicated). Of course, since then, I’ve gotten better with them and have researched them, in-depth. Plus, sites like FanGraphs have been SO helpful in furthering my understanding of advanced stats.

    Anyway, I’m a big judgment guy who supplements judgments with stats. I tend to think that what we perceive visually is often correct. For instance, I remember thinking that Jose Molina caught Phil Hughes and Ian Kennedy better than Jorge Posada did in 2007. I saw it, but I wanted to verify that perception, so I used stats to confirm it (it was true). Visual observation needs to work in tandem with stats. They’re meant to help each, I think and should share an intimate relationship.