A few weeks ago, Dave Cameron wrote a piece on RE24, explaining that, because RE24 measures offensive production with respect to the specific base-out state, one could compare it to a context-neutral offensive metric, such as Batting runs, in order to measure the effects of situational hitting.
Situational hitting is a vague term often used to laud making outs as long as it moves the runner up a base, but as I see it, all the phrase means is hitting differently depending on the situation. That is, good “situational hitting” is distributing your hits and extra base hits into the times that you hit when runners are on base, and especially in scoring position.
Subtracting Batting Runs (or Bat) from RE24 works as a good measure of situational hitting because it compares the value of the context-neutral event (single, strikeout, home run, etc) with the value of the actual change in base-out state. A single is worth more in certain situations; that “more” is measured using this method.
However, while this subtraction is a simple measure in some ways, it is unnecessarily complex in others. Both Bat and RE24 are park-adjusted and league-adjusted, but they make these adjustments in very different ways, not to mention the fact that FanGraphs’ RE24 includes stolen bases and caughts stealings, while Bat does not, so the differences between the two is actually a measure of situational hitting with a little bit of baserunning mixed in for fun.
To fix this, I recalculated RE24 to not include a park adjustment or include SB/CS. I then compared this new version of RE24 not to Bat, but to wRAA, the non-park adjusted version of our context neutral linear weights metric. Let’s take a look at the leaders and laggards based on this comparison this season (as of Wednesday):
|347||Will Middlebrooks||Red Sox||355||-15.1||-3.6||-11.5|
|350||Rajai Davis||Blue Jays||355||-15.2||-1.5||-13.7|
Freddie Freeman, by context-neutral standards, is having a very good year. He’s been about 50% better than the average hitter; he walks, hits for moderate power, and makes a lot of contact. First basemen are supposed to be above average hitters, and Freddie Freeman, without considering context, has been better than the average first baseman.
Considering context, Freeman has been not just a very good hitter, but one of the best hitters in baseball. His RE24 is second in the league, whether you make a park adjustment and include stolen bases or not. This 2+ win jump based on the timing of Freeman’s hits is unsurprising when you consider the fact that he is hitting .440 with runners in scoring position this year. That’s right. .440. I’m almost surprised that the above difference isn’t greater given a number like that.
You may also notice that two Cardinals are in the top three, and three are in the top ten. This makes sense, since St. Louis has a remarkable 138 wRC+ with runners in scoring position this year, compared to a 105 wRC+ overall. Their batting average with RISP is .329, better than the next best Tigers by an incredible 47 points.
But why are we talking about batting average with RISP when we have this fancy new situational hitting tool! If we sum the new RE24 and wRAA for every team, and again take the difference between the two metrics, we see the following results:
70 runs. The Cardinals have seen a 70 run difference based on the timing of when they get their hits. That’s seven additional wins created only from sequencing. In other words, the Cardinals situational hitting was essentially the equivalent to adding Miguel Cabrera to their line-up.
Is this situational/timely hitting something that we can expect to continue? Probably not. Of the 30 teams, only 14 were on the same side of zero last year in this difference. The Cardinals, despite their ridiculous situational hitting this year, were almost exactly average with regards to their situational hitting last year. On an individual basis, there is only about a 0.1 correlation between this difference last year and this year. It’s not a large sample, but is some confirmation of what we already knew about the lack of a sustainable clutch hitting skill.
Not everything has to be predictive, though. This measure might not tell us much about what will happen, but it helps us understand what has happened in the past. When you look at Freeman’s RBI numbers or the Cardinals average with RISP, it’s difficult to know how much of an actual advantage their performances conferred to their teams. This gives us an idea of just how valuable those performances were, and reminds us again that sequencing can be a huge part of wins and losses.