Baseball Prospectus rolled out a new statistic this week called Deserved Runs Created, as well as a league-adjusted DRC+ that allows you to compare players against league average (100). Let’s take a look at the stat, and what it tells us about the Rays.
In case you’d rather do the reading yourself before I get into the summation, here’s the journey you should take:
- The Expected Contribution, by Jonathan Judge
- Introducing Deserved Runs Created Plus (DRC+), by Bryan Grosnick
- The Performance Case for DRC+, by Jonathan Judge
- Why DRC+, by Jonathan Judge
A hitter’s DRC+ of 100 is average, 120 is good, and 145 or higher is excellent.
So what is DRC+?
DRC+ measures all of a player’s contributions at the plate. The model then isolates how much of the outcome should be credited to the hitter, then weights those contributions on the value he provided. Then, the DRC+ model adjusts for context, which include factors like which park the hitter played in and the quality of the opposing pitcher.
I’m thinking of it as FIP for hitting. The approach is complicated in that it uses “mixed modeling” to apply that context, but the good news is that B-Pro is making those models publicly available for scrutiny.
My favorite aspect of this stat is the additional element of variability by prescribing standard deviation. I’ll let Judge explain that:
Mike Trout, for example, in 2018 had a DRC+ of 180, with a standard deviation (plus or minus) of 13 points. So, we are pretty confident Mike Trout was somewhere between 167 and 193, and skeptical about him being in the range beyond that, but are happy to admit he very well could deserve something other than exactly 180. Being honest about uncertainty is an important part of understanding what it means to be accurate.
Judge was the lead architect behind this statistic, and his intention was to limit variance and improve predictability in his metric by isolating hitter contributions.
Compared to statistics like wOBA, wRC+, or OPS+, Judge writes that he has achieved 0.73 reliability in his Performance Case linked above, as compared to 0.35 for the other three stats that lead public discourse, in terms of variance. The result is that predictiveness improved to 0.50 for DRC+ as compared to 0.37 for the other three mentioned.
Now you might be thinking that 0.50 doesn’t sound very predictive — it’s a coin flip as to whether DRC+ is repeatable year over year. But when compared to what was barely a one-in-three shot of predictability (Judge measured a span of 2010 to 2018, which includes the changing baseball), and the same was true when expanding the scope. Again, Judge:
DRC+ experiences no dropoff whatsoever in performance, despite the changing ball and volatile run environment. wRC+ and OPS+, on the other hand, experience notable declines in their reliability and predictive power, with descriptive power somewhat suffering also.
I particularly like that Judge has worked DRC+ is not a finished product, but it’s to the stage where Judge and company were comfortable releasing it to the public.
With the written explanation done, here’s something for the visual learners:
The Rays and DRC+
Baseball Prospectus released a leader board alongside these articles, and here are the full season results from players who wore a Rays jersey in 2018. Left handed hitting is indicated through an asterisk (*), and bold players are those who remain on the 40-man roster.
This leader board was filtered for at least 40 plate appearances.
Only 185 players performed at 100 or greater in 2018, and if those 185 players were distributed evenly, we’d see at least six on each roster. Here the Rays have seven, but that includes Denard Span who was traded shortly after his acquisition.
You’ll note standard deviation (SD) is included in the leader board. Among all players, 18 is the average SD; among qualified hitters (502 PA) the average is 13. Only 41 players has a deviation lower than Tommy Pham’s 12, with the lowest being Giancarlo Stanton with an 8 SD.
In reviewing the list above it’s interesting to note who performed at a “good” or “average” level in 2018, but it may be more interesting to also consider how this stat operates as a projection.
Acknowledging the coin-flip nature of these projections, a simple summary might say that Tommy Pham is the only “good” hitter above, and that it’s possible Ji-Man Choi will be “good” as well.
Meanwhile, based on 2018 performances, Joey Wendle, Daniel Robertson, Austin Meadows, Brandon Lowe, and Matt Duffy project to be “average” hitters, while Willy Adames, Michael Perez, and Jake Bauers could possibly be “average” as well in 2019.
It’s worthwhile to note that three Rays prospects fall in that possibly “average” category. New players often struggle when they reach the majors, but DRC+ did not give uniform treatment to prospects, as Wendle and Lowe rated above league average.
What observations of DRC+ do you have? Join us in the comments.