Viewing the Astros through the DRC+ lens

Although I am not an expert in any regard towards statistical analysis in baseball, the numbers and the process behind obtaining it are fascinating. As a fun(?) exercise, I would like to view the Astros through the lens of DRC+, a relatively new metric from Baseball Prospectus.

Back in early December 2018, Baseball Prospectus unveiled a new metric called Deserved Runs Created Plus, abbreviated simply as DRC+. Here is a brief explanation from Baseball Prospectus noting the difference between this newly released metric and its predecessors that are currently available to the public.

"DRC+ differs from other (public) hitting metrics in that it focuses on each hitter’s expected contribution, rather than merely averaging the result of hitter PAs."

Based solely on the name and abbreviation, DRC+ may appear like an updated version of the commonly used wRC+ (Weight Runs Created Plus) from FanGraphs. Personally, I’ve mostly used wRC+ in my analysis, although it can be interchanged with OPS+ depending on my mood. But DRC+ interests me quite a bit, especially due to the above claim about “expected contribution.” And, no, it isn’t simply an upgrade for wRC+ or OPS+. It is an attempt to allocate proper credit to a hitter from any particular outcome.

Jonathan Judge, one of the inventors of the metric, has an article here on BP that breaks down the case for DRC+. From the get-go, Judge contends that there are “two serious problems” with the currently available hitting metrics.

Reason no.1:

"they purport to offer summaries of player contributions, when in fact they merely average play outcomes in which the players participated; and"

Reason no. 2:

"they treat all outcomes, whether it be a walk or a single, as equally likely to be driven by the player’s skill, even though no one believes that is actually true."

Judge doesn’t go into the specifics of the DRC+ calculation in his article, but he does contend that DRC+ addresses both issues and leaves us with a “more reliable” and “more accurate” metric for hitter evaluation. It is also advised as a better predictor of future performance compared to other hitting metrics. The metric itself doesn’t assumes that the result of a play equals player contributions, which makes plenty of sense. And DRC+ is also designed to reward hitters who provide similar outcomes on a consistent basis.

To my understanding, DRC+ also attempts to account for context behind an outcome. For example, a misplayed batted ball from an outfielder turns a double into a triple, although the fielder is responsible but not given an error. Was it a hitter’s skill alone that led to the triple? Nope as there are other factors in play. The opposing pitcher is another key variable in how a hitter may perform. There is a difference in playing in the same division with Max Scherzer or Lucas Giolito. The metric, in my opinion, is a reasonable effort to give a hitter credit where credit’s due with an emphasis on context.

Bryan Grosnick, also with BP, describes the metric’s approach quite well here:

"DRC+ uses a mixed-model approach to deal with several contextual variables that affect the hitter’s performance, and assign an expected value to the player’s performance that neutralizes those factors."

It is noted by Judge, though, that DRC+ doesn’t prove the other metrics are necessarily “wrong.” He sums it up here:

"There is no “right” or “wrong” when it comes to player metrics; rather, it is a continuous process of trying to get things “more right” than we had them before."

I am curious to see how much traction the metric receives in the coming months, especially as the upcoming 2019 season unfolds. I do admire BP’s attempt to properly allocate credit to a player when appropriate, although I’d like to see further research into the stat, which I am not qualified to give. By the way, DRC+ also addresses some concerns about park effects, especially in Colorado’s Coors Field. Remember that park effects are important part of statistical analysis in baseball today. I do tend to like this aspect of the stat a lot, but further research on the impact would be nice.

So, let’s take a look at some examples involving the Astros, and how different DRC+ is compared to wRC+ and OPS+. Like wRC+ and OPS+, DRC+ is based on a 100 being league average. Anything over 100 is considered above-league average while anything less than 100 is considered below-league average.

Jeff Bagwell, 1996

• 155.3 DRC+
• 173 wRC+
• 178 OPS+

Jimmy Wynn, 1969

• 164.6 DRC+
• 167 wRC+
• 166 OPS+

Jose Altuve, 2017

• 134.8 DRC+
• 160 wRC+
• 164 OPS+

Richard Hidalgo, 2000

• 148.8 DRC+
• 149 wRC+
• 147 OPS+

Michael Bourn, 2008

• 62.4 DRC+
• 58 wRC+
• 57 OPS+

I pulled five examples throughout Astros’ history with a breakdown by DRC+, wRC+, and OPS+ for each player. In the cases of Bagwell and Altuve, there is a bit of a difference between DRC+ and the other metrics. Wynn’s incredible 1969 season is quite similar across the board along with Hidalgo’s 2000 season. I also used Bourn’s 2008 season with the Astros as an example on the end of the spectrum, but the results are still relatively close.

Without knowing the exact formula, it is difficult to pinpoint the exact reason(s) why Bagwell’s and Altuve’s DRC+ is quite different than their respective wRC+ and OPS+ figures. At the same time, I do plan to incorporate more of DRC+ into my analysis for the upcoming 2019 season as there aspects of it that I quite like. Going forward, I’d highly recommend reading the words about DRC+ from Baseball Prospectus directly. It is well worth the time.