Sabermetrics 101: BABIP and OPS+

This is the second installment of a series on sabermetrics by Josh Strohman, an MLB contributor for The Walk On Blog.

Last week we went over the commonly used player value statistics of WAR and WPA. This week we are going to plunge into some of the meatier sabermetric hitting metrics, BABIP and OPS+. These statistics serve different purposes, but both are useful their own ways.

Batting Average on Balls in Play (BABIP)

BABIP is an extremely simple statistic that allows us to compare a player against themselves. BABIP is exactly what the name implies, the percentage of balls a player puts in play that drop for hits. BABIP can also be used for pitchers, but for the purpose of our discussion we will stick to the offensive aspect of the metric.

How BABIP is determined

As I mentioned, BABIP is very simple, so much so, that you can figure any player’s BABIP with only a simple stat line. BABIP values can be found at FanGraphs; however, if you want to calculate BABIP on your own, the equation is:

BABIP: Things to keep in mind

Maybe the greatest use of BABIP is that it can tell us how we should expect a hitter to perform in the future. We can expect a player to maintain a similar BABIP to their career total. So, a player who is overperforming or underperforming will regress to their career BABIP.

For example, if 40 games into the season, Bryce Harper (career BABIP of .317) has a BABIP of .380, we can expect Harper to regress back to a number similar to his career BABIP. Likewise, if Harper had a BABIP of .270 40 games into the season, we could expect his BABIP to return back to around .317.

There are three factors that affect a player’s BABIP. The first is the opposing defense. If a player hits a ball deep into the outfield gap, an elite center fielder may be able to catch the ball for an out; whereas, an average center fielder won’t make the play. A batter has no control over the defense he is up against, nonetheless, it will affect his BABIP.

The second factor affecting BABIP is luck. A batter could hit a line drive right at an outfielder that is caught for an out, but another player could hit a blooper that drops for a hit. The luck involved is completely random, so again, luck is out of the player’s control.

The last factor that affects BABIP is the player’s ability. A player who hits the ball harder and has more line drives is more likely to have more balls fall for hits.

Because a batter has no control over the defense and luck involved with BABIP, a large sample size is needed to stabilize the statistic. Research has shown that a batter needs to have 820 balls in play in order to be confident in a player’s BABIP. That is roughly 3-4 seasons worth of balls in play for an MLB regular.

Tiers of BABIP Values

BABIP is largely dependent on the type of hitter. For example, a fly ball hitter like Mike Trout will have a comparatively lower BABIP because fly balls are less likely to become hits as ground balls and line drives (line drives are most likely to become hits). Obviously, Mike Trout is not a bad hitter, but he will not have the highest BABIP because he is a fly ball hitter.

Despite this, I will still attempt to put BABIP into tiers, in order to, add context to the statistic.

On-base Plus Slugging Plus (OPS+)

OPS+ is a hitting statistic found on Baseball References and is commonly cited, so it will be important for us to better understand it. OPS+ is the first of four major hitting metrics we will look at and will serve as the base for our understanding.

OPS+ aims to encapsulate all the essential factors of hitting and put them into one simple to understand statistic. OPS+ does not, however, come without its flaws. Ready? Sweet, let’s jump in.

How OPS+ is determined

OPS+ is the normalized version of OPS. So, to begin let’s break down OPS. OPS is simple the sum of On-Base Percentage (OBP) and Slugging Percentage (SLG).

OBP tells us the rate that a player reaches base and SLG measures the quality of a player’s hits. Here are how OBP and SLG are determined:

Alright, so now that we have a good grasp on OPS, let’s hop back into OPS+. I previously alluded to the fact that OPS+ was the league normalized version of OPS, let’s talk about what that means. It will be crucial to understand for several statistics we will learn in the coming weeks.

A league normalized statistic (any statistic with a + or – at the end) is one that sets the league average to 100. So, a hitter that has an OPS+ of 100, hit at the league average for that season.

Another nice thing about normalized statistics (like OPS+) is that every tick above (or below) 100 is a percentage point better (or worse) than the league average. For example, in 2019, Christian Yelich had an OPS+ of 179, this means that Yelich was 79% better at the plate than the league average hitter. The opposite is also true, a player with an OPS+ of 80 was 20% worse than the league average.

The last thing we need to discuss is that OPS+ is park and league adjusted. I very briefly described park adjustments last week (link above). Park and league adjustments allow us to better compare players from different teams and different leagues.

OPS+: Things to keep in mind

The first thing that you need to keep in mind when using OPS+ is that it is not adjusted for player position. This may sound trivial, but it is very important to know so we don’t use OPS+ as a measure of overall player value to a team.

It is unfair to compare a first baseman to a shortstop in terms player value using OPS+ because shortstop is far more difficult defensively. A great shortstop with an average bat may be equally valuable to a team as a first baseman with an elite bat. OPS+ is strictly an offensive statistic and should be used as such.

Because OPS+ is normalized, park adjusted and position adjusted, it allows us to compare players more accurately across not only teams and leagues, but also seasons. Want to know how 2019 Mike Trout compared to 1941 Ted Williams? OPS+ can do that for you.

Up to this point, OPS+ sounds too good to be true, and you would be right. The statistics has two major flaws that should be considered when using it.

We need to get one thing across before we look at the issues with OPS+ and that is that producing runs is the most important thing a hitter can do to help their team win. Because of this, our statistics need to reflect and correctly reward actions involved with run production, this is the goal of advanced metrics. Knowing that, let’s jump into the problems with OPS+.

The first major problem with OPS+ is that weighs SLG and OBP equally in terms of run production. The truth is that OBP is nearly 2x more important (1.8x to be exact) in run scoring than SLG.

The second problem is that the value of each type of hit (1B, 2B, 3B and HR) are incorrectly weighed in SLG. SLG would leave us to believe that a home run is 4x more valuable to run production than a single, when in reality a home run is only ~2.4x more valuable to run scoring than a single.

Next week we will take a look at some statistics that compensate for the shortcomings of OPS+.

Tiers of OPS+ values

Despite the problems with OPS+, it can still be useful in determining player production. This along with how commonly the statistic is cited, makes it an important statistic to understand.

Here is a table that should supply some context for OPS+ values.

Stay tuned for next week when we talk about the holy grail of hitting statistics, wOBA, wRAA and wRC+.