## Sabermetrics 101: RE24

RE24 stands for run expectancy based on 24 base outs. While it may not roll off the tongue it is a very useful stat. Lets dissect the stat so we can gain a full understanding of what the stat is measuring and how to best put this information to use.

**Base Out States**

Base out states are simply how many runners are on base, what bases they are on (often referred to as runner alignment), and the number of outs. For example, runners on 2nd and 3rd with one out is one base out state. Another would be bases empty and no outs. There are 24 possible combinations of base out states, thus, RE24.

Every batter that leads off an inning comes to the plate in the base out state of nobody on and nobody out. Say, the leadoff hitter doubles. Then the second batter comes to the plate in the base out state of 0 out and runner on 2nd. If the leadoff hitter grounds out then the second hitter starts his at bat in a base out state of 1 out and nobody on. Pretty intuitive so far.

**Run Expectancy**

Run expectancy is exactly what it says it is. It is the number of runs expected based on the 24 base out states. Below is a run expectancy matrix for a context neutral environment (each park and year have unique matrices based on their run scoring environments).

The base out state matrix is what you use to calculate the RE24 for each plate appearance. To reiterate the values for each base out state is the expected (or average) runs per inning based on the base out state in which the plate appearance begins.

**How To Calculate**

The calculation for each plate appearance requires 3 pieces of data. First, you need to know the run expectancy at the beginning of the plate appearance based on the base out state. You then subtract that from the run expectancy at the end of the play based on the base out state. Finally, you add in any runs that scored during the play.

Let’s look at two examples.

First, let’s use Mike Trout‘s first plate appearance on May 18, 2012 against the Padres. Trout came to the plate in the top of the 1st inning with nobody on and nobody out. By referencing our matrix we know that the run expectancy at the beginning of his plate appearance was 0.461. He walked on five pitches. So the run expectancy at the end of his plate appearance was 0.831. By subtracting 0.461 from 0.831 and adding 0 (since no runs scored) we get his RE24 for that plate appearance.

For the second example, let’s look at Bryce Harper‘s third plate appearance during a game on August 14, 2015. In the top of the 5th inning, Harper came to the plate with 2 outs and runners on 2nd and 3rd. Again, by referencing our matrix we know that the run expectancy at the beginning of Harper’s plate appearance was 0.570. On Matt Cain‘s first pitch Harper hit a home run to right center. So the run expectancy at the end of his plate appearance was 0.095. And because three runs scored on the play we need to add those.

That is how to calculate the RE24 for an individual at bat. To figure the RE24 for each game, season, or career you simply calculate the RE24 for each at bat and then add them all together.

**What Exactly Does This Tell Me & How Can I Use This**

There are a few things to remember when looking at this stat. First, it is expressed as runs (above average). Secondly, this is NOT a context neutral stat like WAR. RE24 measures what happened in the context of the inning and gives it a numerical (expressed as runs) value specific to that situation. Doubling with the bases empty and no outs has a different value than doubling with the bases loaded and 2 outs.

There are several applications for this statistic. You could use it to compare players to see who improved their teams expected runs over the course of a game, season, or career. I think this is a better stat than WAR, wOBA, or wRC+ because it is not context neutral. The higher a players RE24 the better they were at increasing their teams runs expectancy, thus increasing their team’s chances of winning.

You can also use this to evaluate all the potential outcomes of a given situation and decide what strategy is best. A manager could have a card with each possible situation and the likelihood of each outcome and based on the player at bat have a strategy specific to that exact situation. Say the Cubs leadoff hitter gets on first. If Kris Bryant comes to the dish the strategy is let Bryant do what he does; hit the ball hard. If, however, Jon Lester a career .180 hitter who strikes out in over 35% of his plate appearances, comes to the plate the odds that Lester is going to strike out, hit into a fielder’s choice or a double play is high. The smarter play in this situation might be to attempt to bunt the runner to second base. While this still results in a negative RE24 (-0.187) it is less of a loss than a strikeout (-0.342) or a double play (-0.736). While those runs expected obviously don’t count, baseball is about playing the percentages and smart managers like Joe Maddon and others whose organizations are on the cutting edge of integrating sabermetrics would no doubt arm themselves with this information.

RE24 is also a pitching stat but in reverse. So if a batter is credited with 0.289 RE24 then the pitcher is debited 0.289 RE24. Pitching RE24 has several interesting applications especially for relievers that we will dive into in another post.

There you have it, RE24 in a nutshell. It’s not a perfect stat but I believe it does a superior job of quantifying the contributions of each individual at bat to the teams chances of winning than any current stat.

**More From Sabermetrics 101**

Sabermetrics 101: A Brief Introduction

Pingback: Sabermetrics 101: A Brief Introduction | The Baseball Scholar

Pingback: Sabermetrics 101: WAR (Wins Above Replacement) | The Baseball Scholar

Pingback: Sabermetrics 101: wOBA – The Baseball Scholar

Pingback: I’m Not A Machine, I’m Just Albert – The Baseball Scholar

What year is this graph 2017?

LikeLike

Pingback: Expectativa de Carrera I: Calculando la Matriz de Expectativa de Carrera. | Juanito Banca