Insights from a Decade of Offense and Defense Data

Modern baseball statistics are often made easy to interpret. Many statistics available come pre-normalized and with many facets of a player’s production baked in. This post focuses on two such statistics, FanGraphs’ Off and Def metrics. Off, a truncation of Offense, rolls up just about all aspects of a player’s offensive profile, including their base running, all the while controlling for factors like the park they play in. Def, representative of a player’s defensive value, makes a similar effort to broadly capture the value of a player’s defensive abilities.

Rather than view each of these statistics in isolation, featured here is a broad view of how they have corresponded over the last ten seasons. Below is a scatterplot depicting the intersection of these two statistics for all 1,400+ qualifying player-seasons from 2010-2019. Following that scatterplot are two histograms, depicting the distribution of players’ Off and Def values, respectively. These visualizations motivate a couple interesting discussions about baseball’s standout players, seasons, and trends in the 2010s.

Takeaway 1: ‘Off’ Features More Variance, Looks ‘Normal‘

The Off statistic generally aims for every 10 points to equate to 1 WAR; given the breadth of Off distribution, there appears to be greater opportunity to rack up WAR offensively than defensively, even at premium defensive positions. The All-Decade high for offensive contributions came from Bryce Harper, with a 77.7 figure in his 2015 MVP campaign. The lowest total, -36.6, is represented by Chris Davis’ rough (to put it lightly) 2018 season. There is nearly 115 points of difference between those two totals, and the histogram above visually verifies that broad, normal-looking distribution.

Defensively, the best campaign of the decade came in 2010 by virtue of Yadier Molina. Framing is likely incorporated into the Def statistic, as the top 16 player-seasons, in terms of Def, are attributed to catchers. Andrelton Simmons, with 26.1 Def in 2018, is the highest-rated position player. There is visibly less variance in Def than in the case of Off. Its range, from Yadier’s 48.3 down to Carlos Santana’s 2011 season at 35.3 spans roughly 85 points. Upon viewing the Def histogram, one can see that it seems a bit less traditionally normal-shaped (almost uniform toward the center), and with long unlikely tails to either side.

Takeaway 2: Teams Still Prioritize Offensive Profiles

Until quite recently, statistics as comprehensive as Off or Def didn’t exist, particularly in relation to defensive capabilities. But in recent years as new methodologies have arisen to more comprehensively capture defensive performance, player defense has become far more scrutinized. Still, it is fairly clear that the majority of qualifying player-seasons had positive offensive value, despite only sometimes carrying their weight on the defensive end. 

Below is a crosstable which counts the number of player-seasons in each quadrant of Figure 1. As you’ll see, the most common type of player-season, easily, is one where any given player had a positive impact offensively and a negative impact defensively. 

This may not be an entirely fair comparison though, as it does appear harder on average to rate positively on the defensive end than offensively (given how the histograms above are centered). To address the incongruent nature of Off and Def, this next crosstable depicts player-seasons based on the percentile (over/under 50%) of Off and Def values.

To me, this crosstable highlights just how hard it is to be a proficient and well-rounded MLB player. The two most populous groups are those whose players are below the 50th percentile in one of the two statistics. The least likely outcome is for a player-season to be above average offensively and defensively. Put another way, just under 23% of qualifying player-seasons in the last decade were above average on both sides of the ball. 

Takeaway 3: There Are Outliers, Positive or Otherwise

Figure 1 highlights a handful of intriguing outliers. On the far righthand side, Bryce Harper leads all others with his 2015 season. But one should note that 4 of the next 5 rightmost points represent Mike Trout seasons. In fact, of the Top 10 Off seasons the past decade, 4 come just in the last 2 seasons from the trio of Trout, Mookie Betts (2018), and Christian Yelich (2019). 

One point in the upper righthand quadrant sticks out as representing an elite offensive and defensive season; it belongs to Buster Posey in his 2012 season. In fact, when taking the sum of both Off and Def statistics, that season ranks second in the entire decade, behind Mookie Betts’ 2018. The next 5 best combined campaigns belong to Mike Trout, of course. 

What about the most lopsided seasons? Based on the difference in percentiles between Off and Def statistics, Zack Cozart‘s 2014 ranks first overall. In that year, Cozart ranked in just the 1st percentile offensively, while also ranking in the 97th percentile defensively. Yadier Molina, Freddy Galvis, Andrelton Simmons, Alcides Escobar, and J.J. Hardy each had a season where they ranked in the sub-5th percentile offensively while simultaneously ranking north of the 95th percentile on defense; it seems that phenomenon is due, at least in part, to those players playing shortstop. 

On the other end of the spectrum, Paul Konerko, Miguel Cabrera, Victor Martinez, Edwin Encarnación, and Lance Berkman each had >95th percentile offensive seasons while hovering at or below the 5th percentile of rankings defensively.

Takeaway 4: It is Really, Really Tough to be a Two-Way Phenom

How hard is it to be an elite player on one end of the ball, while also being above average on the other? To address this question, instances where a player has been at, or above, the 90th percentile offensively while also being above the 50th percentile defensively (thus a campaign like Bryce Harper’s 2015, for all its offensive brilliance, wouldn’t qualify) have been counted. 

The total number of those player-seasons in the past decade: 60. Just over 4% of player-seasons in the last decade have been elite offensively and above-average defensively. In other words, for the top 10% (~140) of offensive campaigns, more than half of those campaigns were played by below average defensive players, across all positions.

As a final step, instances where a player has ranked at, or above, the 90th percentile defensively while also being above average offensively have also been counted. The total number of those players in the past decade: 60, again. And that is a good takeaway to end with: as it turns out, it was just as rare that a player in the last decade was elite offensively while being proficient defensively, as for it to be the other way around.