The Case to Replace Batting Average

Batting average. It’s a stat most of us learned early on in our baseball fandom as we scoured the backs of baseball cards to know who the best hitters were. It’s the first stat in the traditional “slash line” of batting average, on-base percentage, and slugging percentage (AVG/OBP/SLG). A .300 batting average is the unofficial benchmark for a high performing hitter. Rogers Hornsby is the Cubs all-time career leader in batting average at a .350 clip. We’ve often used batting average as the cornerstone statistic for evaluating how effective a hitter is. But what if there’s a better, more comprehensive stat by which to analyze hitting performance?

By the end of the Cubs’ world championship season of 2016, two of the most important Cubs hitters had nearly identical batting averages. NLCS co-MVP Javier Baez finished with a slightly higher batting average than eventual World Series MVP Ben Zobrist — by .001. But Zobrist had a higher OBP, higher SLG, and higher OPS than Baez. Based on batting average, we would think that Baez and Zobrist had essentially equal value as hitters in the Cubs lineup. Enter wOBA. Weighted on-base average is an advanced statistic that gives us the ability to combine all hitting outcomes together and assign value to them. It is more comprehensive and accurate than traditional baseball card statistics because it weighs each hitting outcome proportionally to the run value it provides. Said another way, wOBA will give you more clarity on how a hitter contributes to their team’s run scoring than AVG, OBP, SLG, or OPS. Neil Weinberg at Fangraphs referred to wOBA as a “gateway statistic” for fans who are looking to learn more about the analytic side of baseball (highly recommend you read his article here). Consider this post peer-pressure to get hooked on advanced statistics.

Let’s start with defining wOBA, or weighted on-base average. Fangraphs defines wOBA as “a rate statistic that attempts to credit a hitter for the value of each outcome (single, double, etc) rather than treating all hits or times on base equally.” So how is this more informative than batting average? Batting average is also a rate statistic, but only measures the rate of hits per at-bat. Batting average doesn’t assign any additional value to a triple versus a single, for example. This is where batting average falls short as a descriptive statistic. Let’s say two hitters both have three hits in their first ten at-bats of a season. That’s 3/10 = .300 batting average for both hitters. But let’s say their stat lines break down like this:

• Hitter A has one single, one double, and home run
• Hitter B has three singles

Because Hitter A has more hits that have higher probabilities of scoring runs, he has a higher wOBA of .409 whereas Hitter B has a wOBA of .270. Hitter A has contributed more to his teams run scoring probability than Hitter B, but the traditional batting average statistic doesn’t capture that. Batting average would lead us to believe both hitters have performed the same through their first ten at-bats when that isn’t the case. Not all hitters with the same batting average can be viewed equally on that statistic alone.

Let’s check out another Cubs example. In 2017, Cubs first baseman Anthony Rizzo and shortstop Javier Baez both had a batting average of .273. Looking at OBP, SLG, and OPS you can start to see where Rizzo’s contributions to the Cubs run scoring probabilities differed from Baez’s plate appearance outcomes. Though they had identical batting averages, Rizzo’s wOBA was 0.054 points higher than Baez’s who finished near league-average. They both achieved hits at the same rate, but their weighted on-base averages describe the differences in the outcomes of those hits and other plate appearance outcomes like walks and hit-by-pitch (both of which we know Rizzo achieved more often than Baez).

Name	AVG	OBP	SLG	OPS	wOBA
Anthony Rizzo	0.273	0.392	0.507	0.899	0.380
Javier Baez	0.273	0.317	0.480	0.796	0.326

But can’t I just look at OBP, SLG, and OPS to find the differences? On-base Percentage (OBP) has the same limitation as batting average. Though OBP includes hits + walks + hit-by-pitch, it does not assign additional value to different hit outcomes. Slugging Percentage (SLG) calculates the total number of bases per at-bat which means it does weight the type of hit outcome a hitter achieves, but it doesn’t weight them accurately nor does it include hitting outcomes other than hits. On-base Plus Slugging (OPS) combines OBP+SLG. OPS is a good statistic in theory but falls short because of the limitations and inaccuracies in its two parent stats. Plus if you can get all of that data baked into one statistic, isn’t that easier? wOBA is certainly more complex to calculate, but necessarily so. It is more informative and more descriptive as a result.

---

The next challenge in learning a new statistic is recognizing what a good wOBA is. Frank Schwindel led the Cubs in 2021 with a .420 wOBA. What the heck does that even mean? Fangraphs put together a guide that outlines weighted on-base averages. Yeah, #Schwinsanity. And what about Rogers Hornsby? He’s the Cubs all-time career leader in wOBA at .460.

Rating	wOBA
Excellent	.400
Great	.370
Above Average	.340
Average	.320
Below Average	.310
Poor	.300
Awful	.290

Here’s how the rest of the 2021 lineup fared from a wOBA perspective (check out Fangraphs for this and many other interesting statistics).

Name	wOBA
Frank Schwindel	0.420
Kris Bryant	0.368
Rafael Ortega	0.355
Patrick Wisdom	0.346
Anthony Rizzo	0.341
Willson Contreras	0.337
Robinson Chirinos	0.336
Nico Hoerner	0.330
Ian Happ	0.328
Matt Duffy	0.327
Javier Baez	0.326
Jake Marisnick	0.311
Joc Pederson	0.308
Sergio Alcantara	0.279
Jason Heyward	0.275
David Bote	0.268
Eric Sogard	0.259

wOBA helps us better understand how much a player contributes to their team’s run scoring. In general, advanced statistics require us to think about performance metrics differently than traditional statistics. Traditional statistics tend to be simpler, but also less informative. As we continue to explore advanced statistics, think of them with a different context than traditional statistics. How did that action contribute to a team’s runs or wins? If we start to think about performance metrics with that question in mind, advanced statistics will begin to seem less foreign and open us up to deeper insights into player performance at Clark & Addison.

---