I’ve been a big proponent of the importance of walks and strikeouts to baseball, but I’m setting them aside for the first part of our analysis: the **Batted Ball Project**. Here, we are only looking at batted balls, and our goal is to zero in on* that part of offense that results from batted balls.*

What we’re all about here is trying (as with many other analysts) to identify the “non-random” portion of on-field production. That has led us, in the past, to focus on “plate skills” (drawing walks, avoiding strikeouts), and treating balls-in-play as relatively random (“random-y” is one made-up word I used).

The more I have come along, I came to realize that there are “batted-ball skills” that weren’t captured in the formulas I had been using, and that the statistics were available at the major-league level to analyze those aspects of performance.

So off we go with batted balls. Can we identify the “non-random” part of batted-ball statistics?

***

We have to start with a measurement of “offense” — since we’re trying to determine the “batted-ball” portion of “offense.”

The leading contenders would be OPS and wOBA, and we decided that’s “it’s OK” to use OPS, since we discovered — somewhat by accident — that its apparent successor, wOBA, didn’t bring much more value to the table (even though it is technically more precise).

But OPS has the great value of simplicity. So we’ll go with OPS.

On-base percentage plus slugging percentage:

OPS = OBP + SLG

I think that OPS “works” because it boils down to **the two essential elements of offense**, both of which are crucial, and which are often at odds with each other:

- Not making outs
- Producing runs

So when we talk about “offensive production,” we are going to measure it with OPS.

***

Now then, what are the batted-ball aspects of OPS?

First, OBP.

OBP is a “*yes-no”* or “*out-not out*” measurement. When we are isolating batted balls, it would be “*hit-not hit*.”

So we’ll look at hits as a percentage of batted balls.

Is that just BABIP, or batting average on balls in play? Almost … but home runs are batted balls but not balls in play. So we need a second measure that includes home runs.

It turns out this HR-encompassing concept has been named “BACON” — or “batting average on contact.” Love it!

BABIP = “Hits in play” (HIP) / “Balls in play” (BIP)

BACON = All hits (HIP + HR) / All batted balls (BIP + HR)

Before I discovered the “BACON” formulation, it was using “Bttd” as my abbreviation for “all batted balls” (balls in play plus HR), and that’s what I have in my spreadsheets, so you might see that pop up in places too. In other words “H / Bttd” (hits per batted ball) is the same as BACON (batting average on contact).

And when we’re looking at the just the HR part, it is HR-per-batted-ball (“HR / Bttd”).

***

As for the SLG portion, it is already mostly based on batted balls, since walks don’t enter into it.

But the denominator of the SLG formula is at-bats, which includes strikeouts.

We need a formula with “batted balls” [or “contact”] as the denominator.

At this point, let’s break down SLG, as we often do, into “batting average” and “isolated power.”

SLG = BA + ISO

For an equivalent of batting average based only on batted balls, we have the “BACON” stat we just discussed (hits per batted ball, which includes homers).

In order to narrow down the ISO component, we are essentially looking at extra-base hits per batted ball (XBH / Bttd) (although in ISO the XBH are weighted by total bases [2x doubles, 3x triples, 4x HR]).

But we will break it down a bit further and look at two stats:

- Doubles and triples per batted ball (“2b3b / Bttd”); and
- Home runs per batted ball (“HR / Bttd”).

Then by using the “weighting” that is used in SLG, we can create what we call “batted-ball isolated slugging” (BtISO):

- BtISO = (2x doubles + 3x triples + 4x HR) / Bttd

With BtISO, we’re eliminating the role of strikeouts in ISO, since strikeouts increase the denominator of ABs. So we’re only looking at power on batted balls.

***

We will add back in strikeouts and walks at the end, but for the **Batted Ball Project**, we’ve now identified the stats that we will consider to be the “batted-ball” portion of OPS (and thus offensive production):

- BABIP (hits per balls-in-play)
- 2b3b / Bttd (doubles & triples per batted ball)
- HR / Bttd (homers per batted ball)

These three stats make up the whole of “batted-ball” OPS, with HR/Bttd contributing to both the “on-base” and the “slugging” portions. In different combinations, we get:

- BACON (hits per batted-balls) [combines BABIP and HR/Bttd]
- BtISO (batted ball ISO) [weights 2b, 3b & HR per batted ball]

Now that we have our measuring sticks, let’s see how the different types of batted balls have an impact on offense.

Here’s part 3.