Do Gold Glove finalists expose flaws in defensive sabermetrics?

1

November 6, 2017 by jessepell

(Courtesy of MLB.com)

Written by Jesse Pelletier

Every year, it frustrates me to see the finalists for the Gold Glove Award. There are three finalists at each of nine positions for both the American and National League for a total of 54 finalists, so it would be pretty miraculous to have absolutely no issues with the list. 2017 has proven to be no exception.

The first position I checked was center field for the American League. Last year’s AL center field finalists were all from the AL East – Jackie Bradley Jr., Kevin Pillar, and Kevin Kiermaier. This year, the list was sure to be different thanks to Byron Buxton of the Minnesota Twins, who has emerged as one of the best defensive outfielders in the game. Here are the finalists for AL center field:

-Byron Buxton

-Kevin Pillar

-Lorenzo Cain

I hate this list for two reasons. The first is the glaring omission of Jackie Bradley, Jr. In my eyes, Jackie Bradley Jr. has to be the greatest defensive center fielder the Red Sox have ever had. Jerry Remy has even made this claim on air several times. Obviously I haven’t seen too many center fielders suit up for the Red Sox, but how could anyone do it better than JBJ?

(Courtesy of Billie Weiss/Getty Images)

Also, props to CBS Sports for the troll job:

The second reason I hate the list is the omission of Kevin Kiermaier. Kiermaier did not play enough games this year to qualify, but if he did he likely would have been a finalist since it’s notoriously difficult to lose the Gold Glove once you’ve won it. Not only did Kiermaier win Gold Gloves the last two years, but he also won the Platinum Glove (presented to the best overall fielder in each league) in 2015. I don’t think Kiermaier would win the award this year even if he qualified, but it sucks that he doesn’t have a chance to defend his title.

Only six AL center fielders qualified for the award. Mike Trout and Adam Jones were the other qualified players besides Jackie Bradley, Jr. Did JBJ somehow have a poor year according to advanced defensive stats? Did Lorenzo Cain fly under the radar for a whole year while we lauded over the likes of Superman Pillar and the young phenom Buxton? Do defensive statistics even agree with Cain being a finalist?

Defensive metrics aren’t great. The most obvious defensive metric is errors, but outfielders tend to not accumulate many errors. Think of it this way: errors among qualified shortstops range from seven (Jose Iglesias) to 28 (Tim Anderson). Anderson had four times as many errors as Iglesias. If one center fielder has one error and another has four, does that really mean as much? Well, apparently errors don’t mean much at all. This is evident for two reasons.

1.) Jose Iglesias is not a Gold Glove Award finalist, despite having the fewest errors among qualified AL shortstops. Nobody else had fewer than nine.

2.) Lorenzo Cain had seven errors. That’s more than any qualified AL center fielder and tied for the most among all center fielders in the MLB with Keon Broxton. No other qualified AL center fielder had more than five, and Kevin Pillar and Mike Trout were tied for the fewest with one each.

OK, so clearly errors aren’t a factor. So, what is? Below is a table of some of the most important defensive metrics for outfielders excluding errors. The first table presents the stats for each qualified AL center fielder, while the second table ranks each player in that stat and calculates an average rank for each player.

In my mind, the most important statistic is defensive wins above replacement (dWAR). WAR is a very controversial statistic because there is not a universally-accepted formula. It is generally accepted, however, because of how well it tends to work. I imagine dWAR is even harder to compute than offensive WAR, but I’m treating it as the most important metric in these tables.

Range factor attempts to capture how much ground a fielder can cover. Again, probably difficult to calculate but proven to be meaningful.

In the AL Center Fielder Rankings table, the column called AVERAGE computes the average rank in the other five categories while the column WT AVERAGE computes a weighted average. To compute the weighted average, I doubled the players’ dWAR rankings and computed the average with those numbers instead.

The results are… interesting. While Byron Buxton seems to be the favorite among experts, the averages column in the table above has him tied with Pillar. It’s like both of these players are so good, numbers can’t even choose who should win! But it gets more interesting. When I weight dWAR, Buxton wins. But if I weight dWAR and fielding percentage, Pillar wins. If I weight range factor alone, Buxton wins easily. If I weight fielding percentage alone, Pillar wins easily. If I remove errors from consideration, Buxton wins easily. It’s a toss-up, folks. Numbers don’t lie.

Kevin Pillar of the Toronto Blue Jays is never taking his eye off the ball – or his first Gold Glove. (Courtesy of Steve Russell/Toronto Star)

Another interesting thing happens when I remove errors: Lorenzo Cain and JBJ tie for third in the unweighted average. In the weighted average, JBJ would finish third and Cain would finish fifth. Errors and fielding percentage are very redundant, so we should choose one to keep and one to eliminate. Fielding percentage is much more telling since it takes into account the number of putout chances a fielder has had, so we’ll keep fielding percentage and ditch errors.

Once we remove errors from consideration, we get the following results and rankings:

UNWEIGHTED RANKINGS:

1.) Byron Buxton

2.) Kevin Pillar

3.) Lorenzo Cain

3.) Jackie Bradley, Jr.

5.) Mike Trout

6.) Adam Jones

WEIGHTED RANKINGS:

1.) Byron Buxton

2.) Kevin Pillar

3.) Jackie Bradley, Jr.

4.) Lorenzo Cain

5.) Mike Trout

6.) Adam Jones

But let’s consider assists for a second. Outfield assists are a mixture of a fielder’s awareness and arm strength, and accuracy, but also of poor baserunning. How many times do we see a catch at the wall and the fielder easily throws out a guy at first because he just put his head down and ran? Assists don’t always isolate the fielder’s skill, so we could remove assists as well. Since Pillar has the most by a decent margin but everyone else has basically the same number, let’s redo the averages so Pillar ranks first in assists and everyone else ties for second. Those results and rankings are as follows:

UNWEIGHTED RANKINGS:

1.) Byron Buxton

2.) Kevin Pillar

3.) Mike Trout

4.) Lorenzo Cain

4.) Jackie Bradley, Jr.

6.) Adam Jones

WEIGHTED RANKINGS:

1.) Byron Buxton

2.) Kevin Pillar

3.) Jackie Bradley, Jr.

3.) Mike Trout

5.) Lorenzo Cain

6.) Adam Jones

So why is Lorenzo Cain a finalist? He hasn’t won a Gold Glove before, so it’s not pedigree that’s giving him an edge. Cain ranks second in both range factor and assists, but he is outside the top three in the other three categories. He also ranks last in two categories, although those categories are basically the same (errors and fielding percentage). If it were up to me, swapping Cain for JBJ would be a no-brainer.

So did the folks choosing the Gold Glove finalists snub JBJ, or do defensive sabermetrics not tell the whole story? Is there something wrong or biased with how we measure defensive skill? I don’t think I can really say whether or not defensive sabermetrics work properly; not many people are familiar enough with how the stats work to be qualified to answer that. For now, I’m just going to show some numbers and try to convince you something is wrong. It might be the selection committee or it might be the stats, but something is wrong.

Tune in soon for another look at this topic as I look at another group of Gold Glove finalists: AL right fielders.

One thought on “Do Gold Glove finalists expose flaws in defensive sabermetrics?

  1. […] AL Gold Glove finalists at center field. If you haven’t already checked it out, you can find it here. In that piece, I discuss why I chose to favor some statistics over others when assessing Gold […]

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Why read us?

We are the voice of nosebleed fans across the country. Composed of journalists, athletes, take-nadoes and fans alike, we tell it how it is. No take is ever too hot and nothing is out of bounds.

Social media

Archives

%d bloggers like this: