Getting Defensive – A Brief Look at Advanced Defensive Stats

colby

A friend of mine recently asked me a question about Colby Rasmus. 

He said that at one point during the season he looked at some defensive stats and found that Rasmus was ranked as the best defensive CF in baseball.  He couldn’t remember what stat he was looking at, however, and asked me if I might know.

That question has led to this post.  I have often been curious about advanced defensive statistics, mainly because they are so controversial and different.  Defense, and how good a player is at defense, is fundamentally subjective.  Almost every aspect of the great game of baseball is cut and dry, black and white, with no grey area.  A home run is a home run.  A triple is a triple.  A strikeout is a strikeout. 

But that is not the way it works when measuring defensive value.  An error in one person’s mind may not be an error in another.  A fantastic diving play by one player may not necessarily be a fantastic diving play by another – he may have no chance to reach it, or may reach it with ease (not needing the dive).

So how can we accurately measure defensive value?  The short answer is that we can’t.  Nobody knows and yet everybody knows, all at the same time.  There are a variety of advanced statistics out there, and each one tells us something different.  All have something important to say; none are right, and yet none are wrong. 

Here is a brief introduction of some of the most popular:

The Basic

Fielding Percentage

The most straightforward defensive statistic out there, and also the one that was relied upon for so many years.  It is calculated as (Putouts + Assists) / (Putouts + Assists + Errors).  It basically tells us how many times a player turns a ball that is hit at him into an out.  According to fielding percentage, the top defensive players at each position in the major leagues are:

FieldPerc

There are so many flaws with this stat, however, that whatever it tells us is virtually worthless, such as:

– An error is defined by the official scorer, who is different in every stadium.  It is an extremely subjective concept

– It doesn’t factor in the type of play made (i.e. hard screaming liner or soft grounder vs. routine gound pall or pop-up).

– It doesn’t factor in range.

To get around these flaws, we need to go more advanced.

The Advanced

Defensive Runs Saved (DRS)

DRS is a very complicated stat meant to capture a players total defensive value, and is calculated by the Fielding Bible.  It rates individual players as above or below average on defense, and is measured in runs saved.

It is meant to separate the great fielders from the good and bad fielders, by measuring how many more (or fewer) successful plays a player will make than the league average.  As per Fangraphs, “if a shortstop makes a play that only 24% of shortstops make (based on computer comparisons and a database of similar plays – location and ball speed), he will get 0.76 points.  If a shortstop blows a play that 82% of shortstops make, he will lose 0.82 points.”

The stat takes into account all aspects of defense, including controlling stolen bases (for catchers and pitchers), handling bunts, how well infielders turn double plays, how well outfielders prevent extra bases from being taken with their throwing arms, how well OF prevent home runs with leaping catches, and an evaluation of the  players range.  As you can see, very complicated.

However, as with most defensive stats, it’s not perfect.  It is compiled by human scorers, and thus likely contains human error and subjectivity.  Moreso, the stat really requires three years worth of data (and a three year average) to really be meaningful.  However, just for kicks, here are the top fielders this season using DRS:

DRS Best

And here are the worst ten fielders:

DRS Worst

For context, the best Blue Jay so far in 2013 is Jose Bautista (T-18th, +8), and the worst is Maicer Izturis (T-1006th, -5).

Ultimate Zone Rating (UZR)

Ultimate Zone Rating works similar to DRS.  It is an advanced metric that uses play-by-play data recorded by Baseball Info Solutions that attempts to estimate a fielder’s defensive contribution above or below the average fielder at his position.  A UZR of zero, means that the fielder is exactly average, based on other players at the same position in the same year.

Fangraphs has an excellent primer on this stat, and gets into far more detail than I will here.  But in a nutshell, it is calculated in a simlar manner to DRS with a few differences, including:

– UZR does not differentiate between single outs and double plays, as DRS does

– DRS adjusts for individual parks with more precise zone buckets, while UZR uses a more blanket adjustment

– UZR uses multi-year samples, while DRS adjusts year-to-year based on league fielding averages

– It seems that UZR doesn’t give “bonus” credit for HR saving catches

Otherwise, the two are fairly similar – both essentially give and subtract points (runs) based on how effectively a fielder turns a batted ball (with a similar trajectory, velocity, and zone on the field) into an out – and should yield similar results.

Let’s see.  Here are the best UZR:

UZR Best

And here are baseball’s worst:

UZR Worst

Again, for context Toronto’s worst is Edwin Encarnacion at -3.0.

Interesting to note that while 7 players make the “best” list for both stats, only two – Lucas Duda and Shin-Soo Choo – appear on both “worst” lists.

Defensive WAR

Defensive WAR, or dWAR, is a baseball reference statistic that measures how many wins above replacement a player contributes to his team purely with defense.  It attempts to combine a zone rating measure (in this case, total zone rating), with defensive runs saved, giving one nice round number.  Since it is measured in wins, it is easily comparable to WAR.

The major difference with dWAR, is that it adjusts the previously calculated defensive metrics for a player’s position.  Not all positions are created equal.  Shortstop, Centre Field, and Catcher are very difficult (SS and CF both require extensive range to be effective, and catcher is just plain hard), whereas first base is fairly easy – for the most part you just catch balls thrown at you.  It’s not much of a surprise that the slowest, and least athletic players most often play first base (think Albert Pujols, Prince Fielder, Adam Lind, Mark Teixeira, etc.).  Similarly, corner outfielders and second baseman are discounted as well.

In terms of defensive WAR, here are the best in baseball in 2013:

dWAR Best

And the worst:

dWAR Worst

In terms of Blue Jays, the best of the bunch is the man who got this entire column started, Mr. Rasmus, with a dWAR of 0.8. (T-26th in the majors).  At the other end of the spectrum is Edwin Encarnacion, T-907th at -0.6.

No matter what stat you like best, a few things are certain: Manny Machado, Andrelton Simmons, and Carlos Gomez are great defensive players.  Lucas Duda and Shin-Soo Choo are not.  And no matter how bad guys like Izturis and Bonifacio looked early this season, take comfort in knowing that there are lots of guys that are way, way worse.

As for Rasmus? While not elite, it is true that he is an above average defender, and – at least according to DRS and dWAR – better than Mike Trout.

Leave a Reply

Your email address will not be published. Required fields are marked *