← The Slate SheetSlatery

Slatery field notes / baseball / guided learning

What is xwOBA?

Slatery Research DeskOctober 12, 2026About 5–6 minutes
One contact, two ways to read itIllustrative
The outcome and the quality of contact tell different storiesA fictional batted ball is caught for an out. The expected lens considers the weighted outcomes of comparable contact, illustrated with a value of .695.Contact
OBSERVED RESULTCaught. An out.
EXPECTED LENS.695weighted value
A catch tells you what happened. xwOBA asks what comparable contact tends to produce. The .695 here comes from the invented probabilities below; it is not a player’s season statistic.
1

The idea

An out can still be good hitting.

A hitter smashes a line drive, but the outfielder catches it. The box score records an out, even though the hitter made great contact.

xwOBA helps you look beyond that result. It uses how hard the ball was hit and its angle off the bat to estimate what similar batted balls usually produce. Walks, hit-by-pitches and strikeouts count too.

The full name is expected weighted on-base average. “Weighted” means different outcomes contribute different amounts: a home run is worth more than a single.

What do exit velocity and launch angle mean?

Exit velocity is how fast the ball leaves the bat. Launch angle is its upward or downward angle. Running speed also matters for some types of contact.

2

Worked example

One batted ball. Several possible outcomes.

Imagine a fictional batted ball. A model gives it the possibilities below. We multiply each probability by that outcome’s weight, then add the contributions. The actual ball can still become an out.

This example is deliberately simplified. Its probabilities and weights are invented so you can follow the calculation; they do not reproduce Statcast’s model.

From possible outcomes to one valueInvented probabilities & weights
OutcomeIllustrative probabilityWeightContribution
Out
40%
0.000.00 × 40% =0.0000
Single
35%
0.900.90 × 35% =0.3150
Double
15%
1.251.25 × 15% =0.1875
Triple
2%
1.601.60 × 2% =0.0320
Home run
8%
2.002.00 × 8% =0.1600
Probability × weight, then add0.6945 ≈ .695
These probabilities add to 100%. The weights are invented teaching values; official wOBA weights vary by season. No real player or Statcast model output is represented.

The single contributes .35 × .90 = .3150. The home run contributes .08 × 2.00 = .1600 despite being much less likely. Add all five contributions and the expected weighted value is .6945, or about .695.

The hit probability in this toy example is 60%: 35% singles + 15% doubles + 2% triples + 8% home runs. The weighted value is .695. They are different quantities. That is why reading an xwOBA value as a percentage leads you astray.

3

The bigger picture

Great contact doesn’t erase strikeouts.

A hitter can crush the ball when they connect and still strike out a lot. Full xwOBA counts those strikeouts, plus walks and hit-by-pitches, so you see more than the contact alone.

Try this separate, invented sample: 20 plate appearances, with 4 walks, 1 hit-by-pitch, 5 strikeouts and 10 batted balls averaging .440. Those ten balls are different from the .695 example above.

Try it yourselfInvented example ¡ 20 plate appearances

Keep the contact quality. Change the strikeouts.

Move the slider. Each strikeout you remove becomes a batted ball averaging .440. Walks and hit-by-pitches stay the same.

Illustrative xwOBA0.397
7.93 total valueá 20 plate appearances
0 strikeouts15 strikeouts

10 batted balls, 5 strikeouts. Same contact quality; fewer strikeouts leave more chances to add value.

See the numbers behind the slider

10 batted balls × .440 + 4 walks × .70 + 1 hit-by-pitch × .73 = 7.93. Divide by 20 to get the result above.

Invented sample and weights. This shows how the calculation works, not what a player will do.

Now compare two hitters.

B makes better contact, but strikes out twice as often. Both have the same walks and hit-by-pitches. Once those strikeouts count, A comes out ahead.

Contact quality is one part of the storyTwo invented 100-PA samples
Contact onlyFull xwOBA: includes walks & strikeouts
Hitter A70 contact ¡ 20 K
0.451
0.386
Hitter B50 contact ¡ 40 K
0.504
0.323
B makes better contact. A has the higher xwOBA after strikeouts count.
Both samples also contain 8 unintentional walks and 2 hit-by-pitches. Toy weights: walk .70; HBP .73. A: (70 × .451 + 8 × .70 + 2 × .73) ÷ 100 = .3863. B: (50 × .504 + 8 × .70 + 2 × .73) ÷ 100 = .3226. Bars share a 0–.650 scale. This is a separate example from the 20-PA lab.

Check whether a number includes strikeouts. “xwOBA on contact” (or xwOBACon) only looks at batted balls. Full xwOBA includes the other outcomes too.

Show the calculation and which plate appearances count

Add the expected value from batted balls and the weighted value of walks and hit-by-pitches, then divide by the plate appearances that count.

Expected contact value + weighted walks + weighted HBP
At-bats + unintentional walks + sacrifice flies + hit-by-pitches

Strikeouts count as at-bats and add zero to the top of the fraction. Intentional walks and sacrifice bunts are excluded. The slider example has neither of those nor any sacrifice flies.

4

Read an example

What does the gap tell you?

Compare actual results (wOBA) with expected results (xwOBA). Both numbers must cover the same dates and plate appearances. Here’s an invented example.

Read an example research cardIllustrative ¡ anonymous hitter
Hitter A ¡ full xwOBA

Start with the sample.

80 plate appearances*
Actual results: wOBA.290What happened
Expected results: xwOBA.350What the contact & other outcomes suggest
.350 − .290+.060Expected minus observed
Read it this way

The hitter’s contact, walks and hit-by-pitches suggested better results than they actually got.

Invented example. *Both numbers use the same 80 plate appearances that count toward the calculation, including strikeouts. No real player or forecast.

The .060 gap suggests this hitter’s results fell short of what their contact and other outcomes supported. It gives you a reason to look closer.

It doesn’t guarantee a rebound. Health, opponents and the hitter’s swing can change. xwOBA can help inform a forecast, but it doesn’t tell you what will happen next game.

You can try these checks on the MLB lineup report.

5

Put it to work

Four checks before you use the number.

  1. What does it count?

    Full xwOBA or contact only? Higher is better for a hitter; lower xwOBA allowed is better for a pitcher.

  2. How much evidence is there?

    Check the dates and sample size. Eighteen batted balls tell you less than a full season. Avoid treating a brief hot streak as the whole story.

  3. What are you comparing it with?

    Use the same season and source for the league average. Match the dates and what counts when comparing players. Don’t average a hitter’s and pitcher’s numbers to predict a matchup.

  4. Does the matchup fit?

    Check pitch mix, handedness and likely playing time. A strong fastball split matters less if the pitcher rarely throws fastballs. xwOBA alone doesn’t give you a betting probability.