Yesterday, in his fantastic baseball newsletter, Joe Sheehan wrote about the historically poor 2015 offensive production of the Seattle Mariners' catchers. As a group, Mariners' catchers have an on-base percentage (OBP) below .200, and a slugging percentage (SLG) of .253. Slugging .253! That preposterously low mark is lower than the batting average of 15 teams this season. As Joe points out in his article, the Mariners' catchers are the worst-hitting position group in baseball history. Sheehan's work demonstrating the sadness of the Mariners' catching group led me to look into the offense production across the defensive spectrum for each team this season.
To do this I used the Baseball-Reference.com play index split finder for team batting production by defensive position. For each team's position group I collected their performance relative to the league's performance for that split. For example, comparing the Dodgers' shortstops OPS (.707) to the league-wide performance by shortstops (.690). This comparison is given by Baseball-Reference as sOPS+, a measure that is similar to tOPS+, which I have used in a few other posts this year. For those unfamiliar, an sOPS+ of 100 is baseline. On offense, which I will be looking at here, higher numbers mean the team is better in the split (defensive position) than the average for that split, and lower numbers mean they have been worse than the league average. Referring to the Dodgers example, they have an sOPS+ of 105, meaning that as a group the Dodgers' shortstops' OPS has been 5 percent better than the league mark for shortstops. As Bryan Grosnick outlined the other day, OPS is not the best statistic to use, but in this case its availability makes it a useful measure.
In the figure below you can see the sOPS+ results for the 2015 season for each team, as a function of position. Cells in red represent positions that are above average (100) for the given position, cells in blue are below average, and cells in white are around the average. The darkness of the cell is an indication of how far from average the production was for that group for that position. For example, you can see the dark blue sadness representing the lowly Mariners' catchers in the first column.
There are a few interesting things to notice in this chart. First, you can see that most teams have a mix of positive and negative production across the defensive positions. There is only one team, the Dodgers, that is without a position with below league average offensive production. The Blue Jays incredible offense and the Rockies in the thin air of Denver have only had two positions with below average production. On the other side of things, there is no team without a position with above league average offensive production, but there are two teams that only have one; the Angels (thanks Mike Trout!) and the Padres (thanks Justin Upton!).
In addition to showing how teams have distributed their offense around the diamond, this chart is also interesting because it shows the great players. The dark red cells represent players like Mike Trout, Josh Donaldson, Joey Votto, Bryce Harper, Andrew McCutchen and Paul Goldschmidt. Those guys get almost all of the plate appearances for their team at their position, so we are not talking about a group being platooned together to provide the production. They are elite hitters.
I suspected there might be something to having a balanced lineup. To measure offensive balance across defensive positions I calculated each team's coefficient of variation (CV); an approach that has been used before when examining lineup balance. CV is simply the standard deviation of a team's sOPS+ divided by their mean sOPS+. Lower values indicate more balance (e.g., Mets and Dodgers), and higher values indicate less balance - more of a stars and scrubs approach (e.g., Angels, Reds). The correlation between CV and runs scored for each team was basically zero (r = .01). It would appear that, based on this crude measure, lineup balance is not all that important. What does matter is getting production. The teams with a high average sOPS+ (e.g., Blue Jays, Diamondbacks, Nationals, and Rockies) tended to score more runs this season (r = 0.57). Now, I should note that this is a quick and dirty way of looking at this relation, and obviously excludes the important factor of a designated hitter in the American League, but it does make sense.
Related to this investigation, although not really possible to directly extract from the chart above, is the offensive production that teams have received from the typically defense-first positions (i.e., catcher, second base, shortstop, center field). After years of toiling around an offensive production level that was five percent below league average (94-96 tOPS+ from 1997 to 2010), teams received production from players at these defense-first positions that was only two percent below league average for each of the last four seasons (98 tOPS+ 2011 to 2014). But, in 2015 production from these positions dropped back down to four percent below average (96 tOPS+), around where it was for much of the previous decade. This could be just a brief blip in the trend, we will have to wait and see what 2016 brings.
How we view good offensive players (and how teams view them as well) is at least partly a function of their position. A shortstop does not need to outhit a first baseman because he is contributing more on defense. But, with that said, it is certainly nice for a team if that shortstop outhits his fellow shortstops. Ideally he also defends well. The balance between offense and defense around the diamond is a fascinating aspect of roster construction, especially during a time when offense is down - largely as a result of more strikeouts and better understanding of defensive positioning. In the current climate it may be worth exchanging nifty glove-work for pop in the bat.
. . .